![[Valid RSS]](images/valid-rss-rogers.png "Valid RSS")
This is a valid RSS feed.
This feed is valid, but interoperability with the widest range of feed readers could be improved by implementing the following recommendations.
rel="self" type="application/rss+xml" />
^
</p>
<?xml version='1.0' encoding='UTF-8' ?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"> <channel><title>Greg's diary</title><link>http://www.lemis.com/grog/diary.php</link><atom:link href="http://www.lemis.com/grog/diary.xml.php"rel="self" type="application/rss+xml" /><description>Greg Lehey's online diary</description><dc:creator /><dc:date>2024-09-28T14:02:58+10:00</dc:date> <dc:rights>Copyright 2015, 2016, 2017, 2018 Greg Lehey</dc:rights><!-- for Emacs, this is a -*- mode: fundamental; coding: utf-8 -*- document --><!-- temporary file, entries for last 7 days of previous month, used for RSS version --> <!-- topic po not selected <p> It's no secret that I have lots of cameras: </p> <p> But I hardly use any of them. In principle I could get by with the <a href="https://en.wikipedia.org/wiki/Olympus_OM-D_E-M1_Mark_II">Olympus OM-D E-M1 Mark II</a>, but since I have others, it's convenient to use them too. I use the E-M1 Mark II for more complicated things, including the weekly <a href="http://www.lemis.com/grog/stones-road/exterior.php">house photos</a>, while I keep the <a href="https://en.wikipedia.org/wiki/Olympus_OM-D_E-M1">OM-D E-M1 Mark I</a> with the <a href="https://www.dpreview.com/products/olympus/lenses/olympus_12-200_3p5-6p3">M.Zuiko 12-200 mm f/3.5-6.3</a> zoom lens for less demanding photos. And then there's the <a href="https://en.wikipedia.org/wiki/Olympus_PEN_E-PM2">E-PM2</a> that I carry in my handbag for unexpected shots, and <a href="http://www.lemis.com/yvonne/">Yvonne's</a> <a href="https://en.wikipedia.org/wiki/Olympus_OM-D_E-M5_Mark_III">OM-D E-M5 Mark III</a>. </p> <p> Enough? Well, of course, but things could be improved. I have been keeping the E-M1 (Mark I) in the lounge room for cat photos like this one: </p> <p> Not the best of results. It was relatively dark, and the lens limited maximum aperture to f/6.1, requiring 1/5 s at 3200/36° ISO, and I haven't been able to get rid of the colour artefacts. I could have got another lens, but then I could have missed the shot, as this attempt (with flash, only 2 minutes later) shows: </p> <p> What I need is a wide aperture lens. And of course I have at least two, the <a href="https://www.panasonic.com/au/support/product-archives/lumix-cameras-video-cameras/lumix-g-lenses/h-x025e.html">Leica Summilux 25 mm f/1.4</a> and the <a href="https://www.voigtlaender.de/lenses/mft/25-mm-10-95-nokton-ii/">Voigtländer Nokton 25 mm f/0.95</a>. The Voigtländer has the disadvantage of manual focus, so I chose the Summilux. I just need another camera, and for that I chose the E-PM2, which I replaced in my handbag with an <a href="https://en.wikipedia.org/wiki/Olympus_PEN_E-PM1">E-PM1</a>. And with the Summilux I could have taken the same image (still at 1/6 s, but f/1.4) at 200/24° ISO. </p> <p> OK, do that, test: </p> <p> Yes, that works better, but look at that depth of field! </p> <div align="left"> </div> <p> Yvonne's face is in focus, but even the power point, only slightly closer, is clearly fuzzy. I'll have to pay particular attention to the focus point. </p> End deselected topic po (Yet another camera!) --> <!-- topic ko not selected <p> Another attempt today at “jing mian” (capital noodles), an attempt to merge <a href="http://www.lemis.com/grog/recipes/Beijing-fried-sauce-noodles.php">Beijing “fried sauce” noodles</a> and <a href="http://www.lemis.com/grog/recipes/KL-Hokkien-Mee.php">KL Hokkien Mee</a>: </p> <p> Not much difference: forget the <a href="https://en.wikipedia.org/wiki/Michiu">michiu</a>, use less water. And I'm still not convinced. </p> End deselected topic ko (Jing mian again) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240828-011029</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240828-011029</link><category>technology</category><category>opinion</category><title>VirtualBox again</title> <description> <![CDATA[ <div align="justify"> <p> It's been a while since I last tried to use my <a href="https://en.wikipedia.org/wiki/VirtualBox">VirtualBox</a> VMs. But I have this <a href="https://en.wikipedia.org/wiki/Ubuntu">Ubuntu</a> system that I installed <a href="http://www.lemis.com/grog/diary-jun2024.php?subtitle=Ubuntu%20again&article=D-20240623-021811#D-20240623-021811">a couple of months ago</a>, and I wanted to look at the <a href="https://www.x.org/">X</a> fonts. Fired it up and... no networking! </p> <p> What went wrong there? Did I forget to load the <a href="https://en.wikipedia.org/wiki/Loadable_kernel_module">KLDs</a>? No: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@hydra (/dev/pts/12)</font> <font color="red">~</font> <font color="blue">23</font> -> <b><tt>kldstat | grep vbox</tt></b></code><br /> 6 3 0xffffffff82c14000 590d0 vboxdrv.ko<br /> 7 2 0xffffffff82c6e000 4210 vboxnetflt.ko<br />10 1 0xffffffff82c82000 55b0 vboxnetadp.ko<br /> </tt></div></blockquote> <p> That's all I need. Try again with <i>tweedledum</i>, a <a href="https://www.freebsd.org/">FreeBSD</a> machine. No problems. And Microsoft? Fired up <i>disaster</i> and found: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/10)</font> <font color="red">~</font> <font color="blue">278</font> -> <b><tt>ping disaster</tt></b></code><br />PING disaster.lemis.com (192.109.197.168): 56 data bytes<br />64 bytes from 192.109.197.168: icmp_seq=0 ttl=128 time=0.181 ms<br />64 bytes from 192.109.197.168: icmp_seq=1 ttl=128 time=0.325 ms<br />...<br /> </tt></div></blockquote> <p> That was the good news. But somehow it had decided to perform an update, something that I definitely didn't want. And when that was done, networking had gone away! </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/10)</font> <font color="red">~</font> <font color="blue">281</font> -> <b><tt>ping disaster</tt></b></code><br />PING disaster.lemis.com (192.109.197.168): 56 data bytes<br />^C<br />--- disaster.lemis.com ping statistics ---<br />3 packets transmitted, 0 packets received, 100.0% packet loss<br /><code><font color="blue">=== grog@hydra (/dev/pts/10)</font> <font color="red">~</font> <font color="blue">282</font> -> <b><tt>arp disaster</tt></b></code><br />disaster.lemis.com (192.109.197.168) at 08:00:27:d4:23:94 on re0 expires in 1199 seconds [ethernet]<br /> </tt></div></blockquote> <p> And somehow <i>arp(8)</i> kept showing expiry times of 1200 or 1199 seconds, suggesting that there had been some activity. Yes: <i>tcpdump(8)</i> showed continuous </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>15:36:07.573601 ARP, Request who-has eureka.lemis.com tell disaster.lemis.com, length 46<br />15:36:07.573605 ARP, Reply eureka.lemis.com is-at bc:5f:f4:c9:9b:bf (oui Unknown), length 28<br /> </tt></div></blockquote> <p> Clearly <i>disaster</i> isn't receiving the reply from <i>eureka</i>. I've <a href="http://www.lemis.com/grog/diary-jun2024.php?subtitle=More%20Microsoft%20on%20VirtualBox%20investigation&article=D-20240626-012332#D-20240626-012332">seen this before</a>. What did I do then? Gave up. </p> <p> Somehow there's something very strange about VirtualBox networking. All the problems I have are also network-related. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-28T01:10:29+00:00</dc:date></item> <!-- topic ao not selected <p> Clearly this horribly expensive “natural rat bait” doesn't kill rats. There are plenty of people who specialize in that, so did a web search and called up <a href="https://www.southernandbeyondpc.com/">Southern and Beyond pest control</a> in <a href="https://en.wikipedia.org/wiki/Sebastopol,_Victoria">Sebastopol</a>, mainly because they're the closest. The prognosis is not good: they'd just put in rat bait, though presumably the more effective poisonous kind. </p> <p> I can do that too, of course. But yes, they confirmed that the poison can kill cats who eat the dead rats. But they can't get up into the ceiling. Never mind, maybe the rats come down from time to time. That would explain the droppings in the garage, and of course the cats' great interest in the garage. </p> <p> So: I'm no further. How about a rat version of the cage that I have somewhere, and in which I once caught a mouse? Oh. I also once had a rat cage and <a href="http://www.lemis.com/grog/diary-jul2011.php?subtitle=Caught%20a%20rat&article=D-20110727-010407#D-20110727-010407">caught a rat</a> in it: </p> <p> Where is the cage? That was 13 years ago, and I can't recall having seen it since. Still, I should be able to find a new one. </p> End deselected topic ao (Begone foul rat) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240828-024028</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240828-024028</link><category>multimedia</category><category>technology</category><category>opinion</category><title>Chasing down the video flashing</title> <description> <![CDATA[ <div align="justify"> <p> Another video with “flashing” this evening, this time not from <a href="https://en.wikipedia.org/wiki/ZDF">ZDF</a> but from <a href="https://en.wikipedia.org/wiki/ARD_(broadcaster)">ARD</a>. Just what <a href="http://www.lemis.com/grog/diary-aug2024.php?topics=tc&subtitle=More%20system%20upgrades&article=D-20240814-012416#D-20240814-012416">I've been waiting for</a>. Switch over to server <tt>1:</tt>, running the <i>nv</i> driver and not the <i>nvidia</i> driver, view the same video at the same position. It flashes! </p> <p> Oh. That's not what I had expected. But at least I have eliminated one potential culprit. What am I left with? It only ever happens on my TV and with some (mainly older) German videos. So we're left with three potential culprits: my <a href="https://hisense.com.au/product/55-uhd-4k-tv-series-a7g/">Hisense A7G</a> TV, the German videos and <a href="https://en.wikipedia.org/wiki/mpv_(media_player)">mpv</a>, the video player. At least two of them must be in collusion, but which? I suppose I should install an unrelated video player (so <a href="https://en.wikipedia.org/wiki/MPlayer">MPlayer</a> is eliminated) and see if that helps. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-28T02:40:28+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240829-023448</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240829-023448</link><category>technology</category><category>opinion</category><title>More VirtualBox fun</title> <description> <![CDATA[ <div align="justify"> <p> <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=VirtualBox%20again&article=D-20240828-011029#D-20240828-011029">Yesterday</a> I was left with puzzles about the networking of my <a href="https://en.wikipedia.org/wiki/VirtualBox">VirtualBox</a> VMs. Today I tried to track it down. </p> <p> I failed. Everything just worked. So what went wrong yesterday? I discovered that today I had booted from the “<a href="https://en.wikipedia.org/wiki/DVD">DVD</a>”, but that shouldn't have made any difference. OK, reinstall. </p> <p> And how about that, it installed a second copy of Ubuntu with differences that I don't understand. But the networking continued to be available, and with a bit of messing around I was able to install such basics as <i>ifconfig</i>, <i>rwhod</i> and <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a>. Now I can look at those <a href="https://www.x.org/">X</a> fonts. </p> <p> Oh. Once <a href="https://en.wikipedia.org/wiki/Linux">Linux</a> and <a href="https://www.freebsd.org/">FreeBSD</a> had relatively similar file hierarchies, but no longer. I wasn't even able to establish if the X server logged anywhere; certainly the way it was started looked very different: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt> 1 1651 1651 1651 ? -1 Ssl 0 0:00 /usr/sbin/gdm3<br /> 1651 1657 1651 1651 ? -1 Sl 0 0:00 \_ gdm-session-worker [pam/gdm-autologin]<br /> 1657 1700 1700 1700 tty2 1700 Ssl+ 1000 0:00 \_ /usr/libexec/gdm-x-session --run-script env GNOME_SHELL_SESSION_MODE=ubuntu /usr/bin/gnome-session --session=ubuntu<br /> 1700 1704 1700 1700 tty2 1700 Sl+ 1000 0:02 \_ /usr/lib/xorg/Xorg vt2 -displayfd 3 -auth /run/user/1000/gdm/Xauthority -nolisten tcp -background none -noreset -keeptty -novtswitch -verbose 3<br /> 1700 1754 1700 1700 tty2 1700 Sl+ 1000 0:00 \_ /usr/libexec/gnome-session-binary --session=ubuntu<br /> </tt></div></blockquote> <p> But after some searching found a directory <i>/usr/share/fonts</i>, which appears to have the kind of fonts I was looking for. Where do I put them? <i>/usr/local/share/fonts/Ubuntu/</i> for the time being. Then to review the current way of making them visible to X, another thing that I have forgotten. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-29T02:34:48+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240830-004507</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240830-004507</link><category>technology</category><category>opinion</category><title>X font progress</title> <description> <![CDATA[ <div align="justify"> <p> So now I have the <a href="https://en.wikipedia.org/wiki/Ubuntu">Ubuntu</a> <a href="https://www.x.org/">X</a> fonts in <i>/usr/local/share/fonts/Ubuntu/</i>. How do I tell X about them? There used to be something you needed to do, but I forget. </p> <p> What I <a href="http://www.lemis.com/grog/diary-oct2023.php?subtitle=Fonts:%20Microsoft%20beats%20X!&article=D-20231029-012114#D-20231029-012114">did recall</a> was the <a href="https://www.gnu.org/software/emacs/"><i>Emacs</i></a> function <tt>view-hello-file</tt>, which shows the expression for “Hello” in many languages and scripts. Last October it showed, inter alia, </p> <a id="Photo-0" name="Photo-0" href="diary-aug2024.php?dirdate=20231028&imagesizes=2&topics=c#Photo-0"> <img alt="This should be Emacs-fonts.png. Is it missing?" border="0" id="Photo_0" title="Photo Emacs-fonts.png. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20231028/small/Emacs-fonts.png" width="480" height="141" /></a> <p> Check up. The fonts are all there! Well, nearly all of them. <a href="https://en.wikipedia.org/wiki/Makasar_script">Makasar</a> is still not there, possibly also not on Ubuntu. In fact, it's not even there in older versions of <i>Emacs</i>. Nothing to worry about right yet. </p> <p> It almost seems that I just needed to put the fonts in the directory hierarchy and the clients find them automatically. Even <a href="https://de.quora.com/Den-deutschen-Substantiven-H%C3%B6he-F%C3%A4che-Tiefe-entspricht-im-Niederl%C3%A4ndischen-hoogte-vlakte-diepte-Ist-hier-im-Deutschen-ein-t-ausgefallen/answer/Stefan-Speck-1">Stefan Speck's Gothic examples</a> are represented. And the saved email messages that I commented about <a href="http://www.lemis.com/grog/diary-oct2023.php?subtitle=More%20font%20pain&article=D-20231028-041128#D-20231028-041128">earlier</a> also render correctly. Only <a href="https://www.mozilla.org/firefox"><i>firefox</i></a> didn't. Oh. Restart it, and that worked too. </p> <p> Great. Copy the fonts to my other machines and continue. But at this point, <i>Emacs</i> on <i>eureka</i> crashed: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@eureka (/dev/pts/8)</font> <font color="red">~</font> <font color="blue">3</font> -> <b><tt></tt></b></code><br />Backtrace:<br />0x51126e <???> at /usr/local/bin/emacs<br />0x50fc33 <???> at /usr/local/bin/emacs<br />0x802a38b6a <pthread_sigmask+0x51a> at /lib/libthr.so.3<br />0x802a3824c <pthread_getspecific+0xe1c> at /lib/libthr.so.3<br />emacs -display hydra:0.2 &<br />[3] 15695<br />You have new mail in /var/mail/grog<br />[1] Segmentation fault (core dumped) emacs public_html/diary.php<br /><code><font color="blue">=== grog@eureka (/dev/pts/8)</font> <font color="red">~</font> <font color="blue">4</font> -> <b><tt></tt></b></code><br /> </tt></div></blockquote> <p> It kept on crashing. After some examination, it seems that the fonts were to blame. Remove them and all was well. It didn't happen on the other systems, so I assume that the system on <i>eureka</i> was just too old. Still, a great improvement. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-30T00:45:07+00:00</dc:date></item> <!-- topic g not selected <p> Seen while walking the dogs today: </p> <p> Particularly the first one, with a <a href="https://en.wikipedia.org/wiki/Grevillea_rosmarinifolia">Grevillea rosmarinifolia</a> and some unspecified <a href="https://en.wikipedia.org/wiki/Acacia">Acacia</a>, caught my eye. </p> End deselected topic g (Spring on its way) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240830-021500</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240830-021500</link><category>technology</category><title>hydra crash</title> <description> <![CDATA[ <div align="justify"> <p> For no particularly good reason, I ran <i>nmap</i> against my /24 today. <i>hydra</i> crashed! It's not clear why; there was no evidence of a dump or anything. But somehow it's another indication that all is not well with the networking, and this time <a href="https://en.wikipedia.org/wiki/VirtualBox">VirtualBox</a> wasn't involved. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-30T02:15:00+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240830-021914</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240830-021914</link><category>technology</category><category>general</category><category>opinion</category><title>PoS pain</title> <description> <![CDATA[ <div align="justify"> <p> <a href="http://www.lemis.com/yvonne/">Yvonne</a> forgot some things on her shopping trip yesterday, and she was going to go in again today to pick up the remainder. But some of the purchases, at the <a href="https://www.facebook.com/Fruits.shack.Wendouree/">Fruit Shack</a>, were things that I was interested in choosing, so I went instead. </p> <p> First to <a href="https://en.wikipedia.org/wiki/Bunnings">Bunnings</a> in <a href="https://en.wikipedia.org/wiki/Delacombe">Delacombe</a> to look at rat traps, and only remembered that I had looked online yesterday and decided that they were far too expensive (by a factor of 2 to 3). On to the Fruit Shack, passing <a href="https://en.wikipedia.org/wiki/ALDI">ALDI</a> on the way, so did the remainder of the shopping there. </p> <p> ALDI has self-serve checkout, and I've used <a href="http://www.lemis.com/grog/diary-sep2023.php?subtitle=ALDI%20self-checkout&article=D-20230907-021810#D-20230907-021810">it in the past</a>. But this time it really showed itself from its worst side. I tried to pay with “savings”, and it went through the whole rigmarole, including <a href="https://en.wikipedia.org/wiki/Personal_identification_number">PIN</a> entry and card removal, before declining the transaction: no linked “savings” account. And as punishment I couldn't choose anything else. I needed an assistant to come and reset something. </p> <p> OK, start paying. “INACTIVITY TIMEOUT”. Another assistant, who stayed there long enough to see a second timeout. </p> <a id="Photo-1" name="Photo-1" href="diary-aug2024.php?dirdate=20240829&imagesizes=12&topics=c#Photo-1"> <img alt="This should be ALDI-receipt.png. Is it missing?" border="0" id="Photo_1" title="Photo ALDI-receipt.png. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240829/small/ALDI-receipt.png" width="276" height="245" /></a> <p> How long were the timeouts? 30 seconds at the most, possibly only 10 or 15. Finally I paid by credit card, accepting their 0.5% surcharge. And as a bonus security measure, the printout included my PIN in plain text, displayed as “Card number”. My guess is that they're still trying to debug the device. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-30T02:19:14+00:00</dc:date></item> <!-- topic aho not selected <p> We've had <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> for nearly 1½ years now, and <a href="http://www.lemis.com/grog/Albums/Animals/Mona.php">Mona</a> for over 4 months. They've been particularly in our sights because they're not allowed outside (though Bruno keeps escaping for up to a minute or two). </p> <p> I think we made a mistake when we got Mona: we put both cats in the lounge room overnight to accustomize them to each other. That works, but how do we now move them to the laundry, where we really want to keep them? If we do that, it won't be in the immediate future. And in particular <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> seems unhappy. Though they lie together at lot of the time, he wakes up frequently and meows loudly. </p> <p> What should we do? He used to sleep in <a href="http://www.lemis.com/yvonne/">Yvonne's</a> bedroom, in a cage to keep him from annoying her. Would he be happy in my bedroom, without a cage? Maybe, but there's still the annoyance factor. I tried it with the expected outcome: he was happy for a while, but then started investigating things. After a while I decided against it and put him outside again. </p> <p> That didn't last long. Round 4:00 he was meowing again. OK, into the garage? Yes, please. For another couple of hours, and then he meowed again, proving in the process that the garage isn't overly sound-proof. </p> <p> We're still not there, but it would be interesting to understand <i>why</i> he's so unhappy. Clearly the cats wake relatively early, 5:00 or so, when I frequently find them in front of the bedroom door. And their interest in the garage suggests that there could be mice in there. Can I leave them in there overnight? </p> End deselected topic aho (Cat insight) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240831-020026</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240831-020026</link><category>technology</category><category>photography</category><title>More Microsoft insights</title> <description> <![CDATA[ <div align="justify"> <p> Did some photo processing today. Try to fire up <i>distress</i>. Nothing happened. </p> <p> That's from an <a href="https://en.wikipedia.org/wiki/fvwm"><i>fvwm</i></a> menu, which, like all such things, is too polite to complain. Try starting the underlying script: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/11)</font> <font color="red">~</font> <font color="blue">8</font> -> <b><tt>/home/local/bin/dordesktop distress 3790x2110+0</tt></b></code><br />Connecting to distress<br /><br />ATTENTION! Found a certificate stored for host 'distress', but it does not match the certificate<br />received from server.<br />Review the following certificate info before you trust it to be added as an exception.<br />If you do not trust the certificate the connection atempt will be aborted:<br /><br /> Subject: CN=distress<br /> Issuer: CN=distress<br /> Valid From: Thu Aug 29 09:44:00 2024<br /> To: Fri Feb 28 10:44:00 2025<br /><br /> Certificate fingerprints:<br /><br /> sha1: 6997782949cdf6f8e1db8a8056ebb46ae4506d11<br /> sha256: 09225dcd1ca45eba670dfd3ac99f3271e4ea23bc4bce9b737f77a7109f16723a<br /><br />Do you trust this certificate (yes/no)? <b>yes</b><br /> </tt></div></blockquote> <p> And after that things ran normally. That must have something to do with <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=hydra%20crash&article=D-20240830-021500#D-20240830-021500">yesterday's crash</a>. But where does the certificate come from? Where is it stored? Why is it a problem when <i>hydra</i> reboots? Why, <i>is</i> it a problem when <i>hydra</i> reboots? Yet another borderline issue. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-08-31T02:00:26+00:00</dc:date></item> <!-- topic k not selected <p> I had planned to cook some chicken with this spice paste today: </p> ` <p> It requires broccoli, so <a href="http://www.lemis.com/yvonne/">Yvonne</a> had bought some on Wednesday. But not enough. So we needed to do something—almost anything—else. But I had already prepared the raw chicken, so it had to be something compatible. </p> <p> To my own surprise, I managed to make <a href="http://www.lemis.com/grog/recipes/chicken-with-cashews.php">chicken with cashew nuts</a>, which I once considered one of my most complicated dishes: </p> <p> About the only thing of interest was a detail in the recipe: </p> <blockquote class="fullwidth"> <p class="listitemwidth"> Start 45 minutes before serving. </p> <p class="listitemwidth"> Put the shiitake in boiling water and soak for an hour or more. </p> </blockquote> <p> But even that wasn't an issue in my timing. </p> End deselected topic k (Recipe refinement) --> <!-- topic aho not selected <p> Over to visit Pene Kirk this morning, and I took the fuzz from the ceiling for a vet's expert opinion: </p> <p> Her verdict: polyester. No animal content. And the dirt in the fuzz was just that, dirt, not mouse or rat droppings. </p> <p> Oh. I suppose that makes sense, though, and it fits with my hypothesis that the rodents chewed their way through the ducting. The hole that they would have left is hardly worth worrying about. And Pene still didn't have a good solution to how to get rid of them (“good luck then”). She had used rat bait, but only because she knows that her cats wouldn't get near them. </p> End deselected topic aho (Rats in the roof: Pene's verdict) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240901-023712</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240901-023712</link><category>technology</category><category>opinion</category><title>Skittish rodents?</title> <description> <![CDATA[ <div align="justify"> <p> We've had enough issues with mice and rats in the house lately, but there are more. One of the many outstanding issues in my transition to <i>hydra</i> is the mouse: for reasons I don't understand, it seems to work intermittently. When marking text (left button down, move the mouse) it seems to reset the start point from time to time. </p> <p> My fault? I was beginning to think so, but lately there has been a second issue: suddenly the mouse pointer will jump <i>to another screen</i>. That's not my fault, but whose fault is it? The mouse? <a href="https://www.x.org/">X</a>? The <a href="https://en.wikipedia.org/wiki/Nvidia">Nvidia</a> driver? Even potentially <a href="https://en.wikipedia.org/wiki/fvwm">fvwm</a>? </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-01T02:37:12+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-aug2024.php#D-20240901-024029</guid><link>http://www.lemis.com/grog/diary-aug2024.php?topics=c#D-20240901-024029</link><category>technology</category><category>photography</category><category>opinion</category><title>distress despair</title> <description> <![CDATA[ <div align="justify"> <p> <a href="http://www.lemis.com/grog/stones-road/exterior.php">House photo</a> day today, involving lots of processing. And <i>distress</i>, my Microsoft 10 box, decided that it wanted to reboot. OK, what the hell, though I hate being told what to do. But after rebooting, things didn't work normally. <a href="https://www.dxo.com/dxo-photolab/">DxO PhotoLab</a> isn't a ball of fire at the best of times, but now it just hung. And other photo software wasn't much better. </p> <p> Bloody Microsoft! While waiting for that, start processing on <i>dischord</i>, a Microsoft 7 box. It was also glacially slow! What's wrong with these things? <a href="https://en.wikipedia.org/wiki/Samba_(software)">Samba</a>? How do you connect these things? Wasn't there some program I could use to test connectivity? <i>smbclient</i>? Is <i>eureka</i> even active? Check from <i>hydra</i>. Hang: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@hydra (/dev/pts/4)</font> <font color="red">~</font> <font color="blue">20</font> -> <b><tt>ps lt11</tt></b></code><br /> UID PID PPID C PRI NI VSZ RSS MWCHAN STAT TT TIME COMMAND<br />1004 70313 18824 25 20 0 13344 2956 <b>nfs</b> D+ 11 0:00.00 ls C<br /> </tt></div></blockquote> <p> Oh. Whatever it is, it's not Microsoft's fault. But clearly <i>eureka</i> was up, so I checked there. Hang: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@eureka (/dev/pts/16)</font> <font color="red">~</font> <font color="blue">1</font> -> <b><tt>ps lt6</tt></b></code><br /> UID PID PPID CPU PRI NI VSZ RSS MWCHAN STAT TT TIME COMMAND<br />1004 5266 43366 0 20 0 16888 2476 <b>suspfs</b> D+ 6 0:00.00 ls -lbL, /Photos/<br /> </tt></div></blockquote> <p> Oh. More and more, it became clear that the disk was not accessible. No error messages, just a hang. Force unmount? That hung the entire system. Big red button time, and the third reboot in <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=hydra%20crash&article=D-20240830-021500#D-20240830-021500">3 days</a> </p> <p> And fortunately the disk came up again. But it's clear that it's time to replace the disk. I've had it up and running for <a href="http://www.lemis.com/grog/diary-jun2017.php?subtitle=New%20disk%20at%20last&article=D-20170626-000457#D-20170626-000457">over 7 years</a>. I had wanted to wait until the price of big <a href="https://en.wikipedia.org/wiki/SSD">SSDs</a> dropped, but I don't think I can wait that long, so it's a new magnetic disk for me. 12 TB? 16 TB? The prices aren't bad, but strangely backup disks are more expensive. I need to think how to handle that. </p> <p> And apart from that, most things came up well. Only my weather station software didn't. It kept reporting “MySQL server has gone away”, which was no longer the case. But it seems it doesn't handle that situation, so I needed to restart it. </p> <p> And then <i>distress</i> told me that “drive” <b>P:</b> was disconnected. How do you reconnect it? I still don't know: surprisingly, it recovered by itself. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-01T02:40:29+00:00</dc:date></item> <!-- topic ko not selected <p> <a href="http://www.lemis.com/grog/recipes/beef-with-orange.php">Beef with orange and broccoli</a> for dinner today. I had forgotten how much work it is, and in the process a number of improvements occurred to me. </p> <p> But the result was disappointing. Not because of the work, but because of new sources of ingredients. <a href="http://www.lemis.com/grog/diary-jan2022.php?subtitle=Beef%20with%20broccoli%20and%20orange&article=D-20220123-012016#D-20220123-012016">Previously</a> I had made it with this orange peel, which I had found surprisingly small: </p> <p> But we had run out of that, and the new stuff looks like this: </p> <p> In addition, I had bought some “sizzle steak” (and not “sizzling steak”, the term that <a href="https://en.wikipedia.org/wiki/Woolworths_Supermarkets">Woolworths</a> use) from <a href="https://en.wikipedia.org/wiki/ALDI">ALDI</a>: </p> <p> It looks nice, marbled and natural, unlike the steak from Woolworths, and it's also more expensive than the Woolworths' offering. But it's <i>tough</i>! And the orange peel is also “tough”: it clearly needs soaking before use. The result was a somewhat disappointing meal. </p> End deselected topic ko (Dinner gone wrong) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240903-015753</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240903-015753</link><category>technology</category><category>opinion</category><title>Where did all my disk space go?</title> <description> <![CDATA[ <div align="justify"> <p> My nightly backups have another surprise in store: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Filesystem 1M-blocks Used Avail Capacity Mounted on<br />/dev/da0p1 7,630,093 7,630,043 -76,250 101% /videobackup<br />tiwi:/spool 7,567,870 7,423,744 68,447 99% /spool<br /> </tt></div></blockquote> <p> <i>videobackup</i> is a copy of <i>tiwi:/spool</i>, made with <a href="https://en.wikipedia.org/wiki/rsync">rsync</a>. Why the discrepancy? Off to find out, and, not surprisingly, discovered a number of files on <i>/videobackup</i> that didn't get removed when they went away on <i>/spool</i>. OK, remove them: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/0)</font> <font color="red">~</font> <font color="blue">55</font> -> <b><tt>df</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity Mounted on<br />/dev/da0p1 7,630,093 7,563,450 -9,658 100% /videobackup<br />tiwi:/spool 7,567,870 7,420,523 71,668 99% /spool<br /> </tt></div></blockquote> <p> How can that happen? What does <i>du</i> say? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/0)</font> <font color="red">~</font> <font color="blue">56</font> -> <b><tt>du -s /videobackup</tt></b></code><br />7563451 /videobackup<br /><br /><code><font color="blue">=== root@tiwi (/dev/pts/14)</font> <font color="red">~</font> <font color="blue">3</font> -> <b><tt>du -s /spool</tt></b></code><br />7573170 /spool<br /> </tt></div></blockquote> <p> <i>videobackup</i> usage is as expected, but <i>/spool</i> now has <i>more</i> storage in use than <i>/videobackup</i>, in fact more than the size of the disk and over 150 GB more than <i>df</i> shows! How can that happen? <i>du</i> doesn't follow <a href="https://en.wikipedia.org/wiki/Symbolic_link">symbolic links</a>. Is there something hidden under a mount point? But nothing is mounted on <i>/spool</i>, and even if it were, <i>df</i> should have shown the correct values. </p> <p> What happens if I <i>umount</i> and remount <i>/spool</i>? Not what I expected: the values are unchanged, but from all other machines I get: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@lagoon (/dev/pts/7)</font> <font color="red">~</font> <font color="blue">12</font> -> <b><tt>mount tiwi:/spool /spool</tt></b></code><br />mount_nfs: nmount: /spool: Stale NFS file handle<br /> </tt></div></blockquote> <p> At first I thought it was an issue of processes having the old handle open, but <i>tiwi:/spool</i> wasn't mounted on <i>lagoon</i> when I remounted it on <i>tiwi</i>. I wasn't able to solve the problem, and I ended up rebooting <i>tiwi</i>—4 machines rebooted in 4 days! How that grates. And for reasons that I don't understand, <i>tiwi</i> started <a href="https://www.x.org/">X</a>, but came up without a window manager. That's low priority, though: <i>tiwi</i> should go away in the near future, and I don't use the display. More to the point, though, the backup disk was <i>still</i> full. Let's see what happens with the next backup. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-03T01:57:53+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240903-024741</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240903-024741</link><category>technology</category><category>opinion</category><title>Another disk fail?</title> <description> <![CDATA[ <div align="justify"> <p> Seen in this morning's logs: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 2 01:12:21 eureka kernel: (da1:umass-sim1:1:0:0): READ(16). CDB: 88 00 00 00 00 01 22 76 29 a8 00 00 00 08 00 00<br />Sep 2 01:12:21 eureka kernel: (da1:umass-sim1:1:0:0): CAM status: SCSI Status Error<br />Sep 2 01:12:21 eureka kernel: (da1:umass-sim1:1:0:0): SCSI status: Check Condition<br />Sep 2 01:12:21 eureka kernel: (da1:umass-sim1:1:0:0): SCSI sense: ABORTED COMMAND asc:47,3 (Information unit iuCRC error detected)<br />Sep 2 01:12:21 eureka kernel: (da1:umass-sim1:1:0:0): Retrying command (per sense data)<br />Sep 2 01:12:29 eureka kernel: (da1:umass-sim1:1:0:0): READ(16). CDB: 88 00 00 00 00 01 22 76 29 a8 00 00 00 08 00 00<br />Sep 2 01:12:29 eureka kernel: (da1:umass-sim1:1:0:0): CAM status: SCSI Status Error<br />Sep 2 01:12:29 eureka kernel: (da1:umass-sim1:1:0:0): SCSI status: Check Condition<br />Sep 2 01:12:29 eureka kernel: (da1:umass-sim1:1:0:0): SCSI sense: NOT READY asc:4,1 (Logical unit is in process of becoming ready)<br />Sep 2 01:12:29 eureka kernel: (da1:umass-sim1:1:0:0): Polling device for readiness<br /> </tt></div></blockquote> <p> That's my “new” 10 GB backup disk, the one I bought <a href="http://www.lemis.com/grog/diary-jun2023.php?subtitle=New%20backup%20disk&article=D-20230610-031120#D-20230610-031120">only a little over a year ago</a>. Is it failing already? Presumably it has a warranty of at least 36 months. But is it defective? What does “Information unit iuCRC error detected” mean? And the time is suspicious. From the inverter: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt> <br />tstamp Pac Status SOC VBat PacGrid PacBat PacPV FromPV W1<br />2024-09-02 01:12:54 1046 Waiting to connect to On-grid 22 254 NULL 1100 0 0 0<br />2024-09-02 01:12:55 873 Off-grid 22 254 NULL 910 0 0 0<br />2024-09-02 01:13:00 515 Off-grid 22 255 NULL 540 0 0 0<br />2024-09-02 01:13:01 520 Off-grid 22 255 NULL 544 0 0 0<br />...<br />2024-09-02 01:13:21 512 Off-grid 21 255 NULL 536 0 0 0<br />2024-09-02 01:13:22 570 Waiting to connect to On-grid 21 255 NULL 594 0 0 0<br /> </tt></div></blockquote> <p> That first line is strange; normally it comes <i>after</i> a grid outage, like the last line. Presumably this was a transient. </p> <p> That's quite possible, but there are two issues: firstly, the disk error message was reported nearly a minute before the outage, and then it repeated at 2:25, when the inverter had no issues. On the other hand, <a href="http://www.lemis.com/yvonne/">Yvonne</a> reported two failures of her bedside clock, one of them round 2:30. So for the moment it looks as if the errors are due to flaky power. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-03T02:47:41+00:00</dc:date></item> <!-- topic P not selected <p> <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=Trump%20running%20scared&article=D-20240805-021238#D-20240805-021238">When I signed up</a> for <a href="https://en.wikipedia.org/wiki/Truth_Social">“Truth” Social</a>, I had to choose 3 people to “follow”. <a href="https://en.wikipedia.org/wiki/Donald_Trump">Donald Trump</a> was a clear choice, and to my surprise <a href="https://en.wikipedia.org/wiki/Kamala_Harris">Kamala Harris</a> is there too, so I “followed” her. </p> <p> And the third? I was given a list, and the name <tt>catturd2</tt> stuck out, so I followed him/her. So far he/she seems to have been the most amusing. Today I received <a href="https://truthsocial.com/@catturd2/posts/113064580002579108">this image</a>: </p> <p> </p> End deselected topic P (Trump's hiding place discovered) --> <!-- topic Jho not selected <p> Under the shower this morning, the water seemed cool. I was able to adjust it to normal shower temperature, but why did I need to? </p> <p> Outside to take a look at the hot water system. Yes, it displayed an error code, <b>E9</b>, and power cycling didn't make it go away. What's that? I've had issues with this unit <a href="http://www.lemis.com/grog/diary-dec2022.php?subtitle=No%20hot%20water!&article=D-20221215-020200#D-20221215-020200">before</a>, <b>E11</b>, and on that occasion I found the instructions online. But this time I couldn't find it. After all, the unit (apparently a Bosch Compress 3000 DW FOO, or maybe HP 270-2E0 FOO; why do these things always have two names?) is <a href="http://www.lemis.com/grog/diary-jul2019.php?subtitle=Repairing%20the%20hot%20water%20system&article=D-20190731-023954#Bosch-hot-water">over 5 years old</a>! But I had the URL of the document, and based on that I was able to find it locally. </p> <p> So: <b>E9</b> means “Incorrect system purging or Water pump blocked (water not recirculating) or Hydraulic circuit blocked (water not recirculating)". In each case, call a manufacturer's representative. OK, call Bosch (1300 30 70 37). Andrew answered immediately: it seems that their phone system was down, and he couldn't find the representative immediately, but he would call back. </p> <p> In the meantime, how about Paul of <a href="http://ballaratplumbingservices.com.au/">Ballarat Plumbing Services</a> (0418 527 576), the bloke who installed it. Yes, could come this afternoon. Please <a href="https://en.wikipedia.org/wiki/SMS">SMS</a> details. </p> <p> Did that, and how about that, Paul showed up in the afternoon as promised. He suspected a defective pump, but it seems that fortunately it was only clogged with some black gunge, which he says is typical of systems run off tank water (why?). Here the pump in the second photo, still bleeding now-clear water: </p> <p> But the thing takes hours to get back to temperature, and we won't know for sure until tomorrow morning. </p> End deselected topic Jho (No hot water!) --> <!-- topic go not selected <p> I had noted with some disappointment that the wildflower bush that we planted last year flowered briefly and then died. Here a year ago and today: </p> <p> That's the collection of twigs in the middle, not the healthy plants around it. But things are looking up: </p> <p> Will it come back? What happened in the first place? </p> End deselected topic go (Wildflower recovery?) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240904-021441</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240904-021441</link><category>technology</category><category>multimedia</category><category>opinion</category><title>Finding the lost disk space</title> <description> <![CDATA[ <div align="justify"> <p> After my concerns about disk space <a href="http://www.lemis.com/grog/diary-sep2024.php?subtitle=Where%20did%20all%20my%20disk%20space%20go?&article=D-20240903-015753#D-20240903-015753">yesterday</a>, the backup ran normally today, and for some reason there was much more space available. Here yesterday, then today, seen from <i>eureka</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Filesystem 1M-blocks Used Avail Capacity Mounted on<br />/dev/da0p1 7,630,093 7,630,043 -76,250 101% /videobackup<br />tiwi:/spool 7,567,870 7,423,744 68,447 99% /spool<br />...<br />/dev/da0p1 7,630,093 7,372,189 181,603 98% /videobackup<br />tiwi:/spool 7,567,870 7,374,421 117,770 98% /spool<br /> </tt></div></blockquote> <p> Why? A look around showed me that <i>everything</i> was gone from <i>/spool/Videos</i>, my collection of films. Ah, just a viewpoint. The directories are on <i>hydra</i>, and they're not visible from <i>tiwi</i>. On <i>teevee</i> I really do have other file systems mounted on <i>/spool</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>tiwi:/spool 7,567,870 7,374,406 117,785 98% /spool<br />hydra:/VB3/spool/Videos 3,815,019 1,762,948 2,013,920 47% /spool/Videos<br />hydra:/VB3/spool/Images 3,815,019 1,762,948 2,013,920 47% /spool/Images<br /> </tt></div></blockquote> <p> Theoretically that would explain why the sum of the file sizes on <i>/spool</i> would exceed the size of the disk. But that doesn't affect the views from <i>tiwi</i> or <i>eureka</i>, as shown above. I should examine the backup logs more carefully. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-04T02:14:41+00:00</dc:date></item> <!-- topic ko not selected <p> I'm out of deep-frozen cooked rice, time for a new batch. </p> <p> I've been cooking rice for well over 50 years, and until recently I had no issues. Put the rice in the saucepan (I've found rice cookers <a href="http://www.lemis.com/grog/diary-jan2009.php?subtitle=Rice%20cookers:%20why%20bother?&article=D30-14#D30-14">not worth the trouble</a>) and cover with water by <i>that much</i>. How much? About 2 cm. </p> <p> Problem: that's not very repeatable. Use the same amount of rice in a larger-diameter saucepan and the water ratio will increase. So I started measuring and found that the water was about 1.5 times the amount of rice. But then I heard of people using as much as twice the amount of water, and recently I have increased it to up to 1.8 times. It seems to work. What about 1.9 times? That's what I tried today. <i>Definitely</i> too much, at least for this rice (<a href="https://en.wikipedia.org/wiki/ALDI">ALDI's</a> best long grain). </p> <p> Now I should check less water again, and maybe give rice cookers another chance. </p> End deselected topic ko (Cooking rice, yet again) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240905-004907</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240905-004907</link><category>technology</category><category>opinion</category><title>New disk</title> <description> <![CDATA[ <div align="justify"> <p> My new 16 TB Seagate Exos disk has arrived already. How do I install it? One issue with having <i>hydra</i> built commercially is that I don't understand it as well as the machines I built myself. Took off the covers and compared with the motherboard documentation (at <a href="https://download.msi.com/archive/mnu_exe/mb/MAGX670ETOMAHAWKWIFI.pdf">https://download.msi.com/archive/mnu_exe/mb/MAGX670ETOMAHAWKWIFI.pdf</a>), and that's straightforward enough. But where do I mount the disk? <a href="http://www.lemis.com/grog/diary-oct2023.php?topics=c&subtitle=Looking%20at%20hydra&article=D-20231018-231622#D-20231018-231622">When I picked up the machine</a>, Seth pointed out disk mounting positions, but they're for <a href="https://en.wikipedia.org/wiki/SSD">SSDs</a>. There seems to be no provision for 3½" magnetic disks. Should I put it on the base? Or remove the <a href="https://en.wikipedia.org/wiki/DVD">DVD</a>, find an adapter frame and put it in there? Or just put it in <i>eureka</i>? That would clearly be the simplest solution. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-05T00:49:07+00:00</dc:date></item> <!-- topic Jh not selected <p> Andrew from Bosch called back after only a little more than 24 hours. All done and dusted, of course, but he was able to give me another number to call: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>If you need further service work for the heat pump then consider contacting Craig smith. 0434460771<br /> </tt></div></blockquote> End deselected topic Jh (Bosch service) --> <!-- topic Jho not selected <p> <a href="http://www.lemis.com/yvonne/">Yvonne</a> has been having fun getting a quotation for replacing our defective gate with a sliding one like we had in <a href="http://www.lemis.com/wantadilla/">Wantadilla</a>. Things don't look good: we had one quote round $10,000. The current one, from BOE Constructions, was $6,050—<i>without</i> installation. Today Chris of Complete Fencing Solutions Ballarat came along to look at the installation. He later sent a quote—<i>another</i> $5,445! That's a total of $11,495! I can't recall what we paid at Wantadilla, but I think it was under $3,000. We can resist that. </p> <p> In fact, it seems that he has quoted for things that were already in the other quote, the motor, controller and solar power kit, a total of $4,235. And that corresponds pretty directly to the $403.43 (including postage) that we paid for the current actuator! </p> End deselected topic Jho (Installing a sliding gate) --> <!-- topic ko not selected <p> I have fond memories of pepper-based foods from my time in <a href="https://en.wikipedia.org/wiki/Sarawak">Sarawak</a>, so a while back I bought this spice mix: </p> <p> I've been meaning to use it for ever, but I was concerned that <a href="http://www.lemis.com/yvonne/">Yvonne</a> might find it too overpoweringly <i>pedas</i>. It expired a few months ago, and today we had some other Chinese leftovers, so I got round to cooking it. </p> <p> I had my doubts, and I should have looked at the recommended recipe more carefully: 450 g (isn't that a familiar number?) chicken, 150 g broccoli and 25 g carrot. And both vegetables should be cooked for only 2 minutes! </p> <p> Despite all that, it didn't turn out too badly: </p> <p> But clearly I should read the instructions more carefully before buying these things. </p> End deselected topic ko (Sarawak pepper chicken) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240905-022430</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240905-022430</link><category>technology</category><category>opinion</category><title>More NFS pain</title> <description> <![CDATA[ <div align="justify"> <p> Watching TV in the evening, I wanted to watch a film. They're on <i>/spool/Videos/</i>, <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a> mounted on an external drive from <i>hydra</i>. And I had an access error. </p> <p> Oh. Yes, <i>hydra</i> agreed. Disk not accessible. I must have accidentally disconnected it while looking at the machine this afternoon, and <i>/var/log/messages</i> agreed: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 4 14:25:03 hydra kernel: ugen3.2: <Western Digital My Passport 2627> at usbus3 (disconnected)<br />Sep 4 14:25:03 hydra kernel: umass0: at uhub2, port 4, addr 1 (disconnected)<br />Sep 4 14:25:03 hydra kernel: da0 at umass-sim0 bus 0 scbus8 target 0 lun 0<br />Sep 4 14:25:03 hydra kernel: da0: <WD My Passport 2627 4008> s/n 575856324537304557343143 detached<br />Sep 4 14:25:03 hydra kernel: ses0 at umass-sim0 bus 0 scbus8 target 0 lun 1<br />Sep 4 14:25:03 hydra kernel: ses0: <WD SES Device 4008> s/n 575856324537304557343143 detached<br />Sep 4 14:25:03 hydra kernel: (ses0:umass-sim0:0:0:1): Periph destroyed<br />Sep 4 14:25:03 hydra kernel: (da0:umass-sim0:0:0:0): Periph destroyed<br />Sep 4 14:25:03 hydra kernel: umass0: detached<br />Sep 4 14:25:06 hydra kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device Western Digital My Passport 2627 (0x1058:0x2627)<br />Sep 4 14:25:06 hydra kernel: ugen3.2: <Western Digital My Passport 2627> at usbus3<br />Sep 4 14:25:06 hydra kernel: umass0 on uhub2<br />Sep 4 14:25:06 hydra kernel: umass0: <Western Digital My Passport 2627, class 0/0, rev 2.10/40.08, addr 1> on usbus3<br />Sep 4 14:25:06 hydra kernel: umass0: SCSI over Bulk-Only; quirks = 0x8000<br />Sep 4 14:25:06 hydra kernel: umass0:8:0: Attached to scbus8<br />Sep 4 14:25:06 hydra kernel: da0 at umass-sim0 bus 0 scbus8 target 0 lun 0<br />Sep 4 14:25:06 hydra kernel: da0: <WD My Passport 2627 4008> Fixed Direct Access SPC-4 SCSI device<br />Sep 4 14:25:06 hydra kernel: da0: Serial Number 575856324537304557343143<br />Sep 4 14:25:06 hydra kernel: da0: 40.000MB/s transfers<br />Sep 4 14:25:06 hydra kernel: da0: 3815415MB (7813969920 512 byte sectors)<br />Sep 4 14:25:06 hydra kernel: da0: quirks=0x2<NO_6_BYTE><br />Sep 4 14:25:06 hydra kernel: ses0 at umass-sim0 bus 0 scbus8 target 0 lun 1<br />Sep 4 14:25:06 hydra kernel: ses0: <WD SES Device 4008> Fixed Enclosure Services SPC-4 SCSI device<br />Sep 4 14:25:06 hydra kernel: ses0: Serial Number 575856324537304557343143<br />Sep 4 14:25:06 hydra kernel: ses0: 40.000MB/s transfers<br />Sep 4 14:25:06 hydra kernel: ses0: SES Device<br /> </tt></div></blockquote> <p> OK, force <i>umount</i>, <i>fsck</i>, <i>mount</i>. All OK. Mount from <i>teevee</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>mount_nfs: nmount: /VB3: Stale NFS file handle<br /> </tt></div></blockquote> <p> I've <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=c&subtitle=Where%20did%20all%20my%20disk%20space%20go?&article=D-20240903-015753#D-20240903-015753">seen that before</a>, but the first time was only a couple of days ago. Why is this happening now? And I didn't find a solution this evening; we watched series episodes instead: they're on a different disk, on <i>tiwi</i>. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-05T02:24:30+00:00</dc:date></item> <!-- topic hH not selected <p> When I got up this morning I discovered a message on <i>albo</i>, my mobile phone: <a href="https://en.wikipedia.org/wiki/Powercor_Australia">Powercor</a> had become aware of a <a href="http://www.lemis.com/grog/powercor/power-failures.php#bottom">grid power failure</a> at 5:52, along with another message telling me that power had been restored 41 minutes later. The air conditioner was running, but fortunately we didn't run out of power. </p> <p> Than another at 11:27, while we were having breakfast, this time only 22 minutes. The inverter didn't handle it well, causing various screams from the electronics, but the only thing really affected was the radio. </p> <p> Clearly this was another <a href="http://www.lemis.com/grog/diary-sep2019.php?subtitle=Another%20power%20failure&article=D-20190905-234500#D-20190905-234500">fifth anniversary gift</a>. But to be fair to Powercor, things have looked up a lot since we moved to <a href="https://en.wikipedia.org/wiki/Dereel">Dereel</a>. Let's hope they stay that way. </p> End deselected topic hH (Anniversary power fail) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240906-021724</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240906-021724</link><category>technology</category><category>multimedia</category><category>opinion</category><title>NFS recovery</title> <description> <![CDATA[ <div align="justify"> <p> So why am I getting these stale <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a> handles? Discovered a number of these messages in <i>/var/log/messages</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 4 18:40:48 teevee kernel: newnfs: server 'hydra' error: fileid changed. fsid 0:0: expected fileid 0x31700, got 0x2. (BROKEN NFS SERVER OR MIDDLEWARE)<br /> </tt></div></blockquote> <p> What does that mean? The result of trying to remount a new mount on <i>hydra</i>? What does <i>hydra</i> say? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 4 18:44:50 hydra kernel: UFS: forcibly unmounting /dev/da0p1 from /VB3 </tt></div></blockquote> <p> But that was a few minutes later. Nothing obvious. A web search brought up various stuff, suggesting that it's new, but nothing that helps much. What I did establish is that the issue is with <i>hydra</i>, not with <i>teevee</i>, and that it's not related to existing mounts: the same thing happens on <i>lagoon</i>, which didn't have the file system mounted. </p> <p> So: what happens if I mount the same device on a different mount point on <i>hydra</i>, and then export that and mount it from <i>teevee</i>? That's a thing that I wanted to do anyway, so tried that. Instead of mounting and exporting <i>/VB3</i>, use the mount point name <i>/Videos</i>. Success! </p> <p> So somewhere it seems that there's an issue in the NFS stack. I suppose I should follow up. </p> <p> To confuse the issue, looked at the file systems mounted on <i>teevee</i> after that and found, inter alia </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Filesystem 1048576-blocks Used Avail Capacity Mounted on<br />tiwi:/VB2 5,722,572 5,182,906 482,440 91% /VB2<br />tiwi:/spool 7,567,870 7,374,516 117,675 98% /spool<br />hydra:/VB3/spool/Images 3,815,019 1,762,948 2,013,920 47% /spool/Images<br />tiwi:/VB2 39,664 23,880 12,610 65% /VB2<br />hydra:/Video 3,815,019 1,762,948 2,013,920 47% /Video<br />hydra:/Video/spool/Videos 3,815,019 1,762,948 2,013,920 47% /spool/Videos<br /> </tt></div></blockquote> <p> There are a number of issues here: </p> <ul> <li class="fullwidth"> <div class="listitemwidth"> I have mounted a NFS file system from <i>hydra</i> (<i>hydra:/VB3/spool/Images</i> on top of <i>/spool</i>, an NFS file system mounted from <i>tiwi</i>. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> Possibly the problem is related to the fact that I still had <i>hydra:/VB3/spool/Images</i> mounted. I seem to have mised that. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> <i>/VB3</i> isn't mounted on <i>hydra</i>, but it's still visible on <i>teevee</i>, and the contents are correct, and I can create and remove files. Where is it getting this information from? </div> </li> <li class="fullwidth"> <div class="listitemwidth"> <i>tiwi:/VB2</i> is mounted twice, accessing two different file systems (the second appears to be <i>tiwi:/</i>). </div> </li> </ul> <p> I've run into the “double mount” problem before. It doesn't appear to be limited to NFS. Why should it ever be useful to mount a file system twice, especially if two different file systems get mounted on the same mount point? I've already established that a <i>umount</i> will umount the last instance, but I still needed to force <i>umount</i> it </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/spool</font> <font color="blue">55</font> -> <b><tt>umount /VB2</tt></b></code><br />umount: unmount of /VB2 failed: Device busy<br /><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/spool</font> <font color="blue">56</font> -> <b><tt>umount -f /VB2</tt></b></code><br /><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/spool</font> <font color="blue">57</font> -> <b><tt>df</tt></b></code><br />...<br />tiwi:/VB2 5,722,572 5,182,906 482,440 91% /VB2<br /><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/spool</font> <font color="blue">58</font> -> <b><tt>l /VB2</tt></b></code><br />total 19428<br />drwxrwxr-x 2 grog home 512 12 Feb 2018 .snap<br />-r-------- 1 root wheel 33,554,432 1 May 2019 .sujournal<br />drwxr-xr-x 8 grog home 512 25 Jun 11:43 spool<br />drwxr-xr-x 23 grog wheel 512 29 Apr 2023 spool-old </tt></div></blockquote> <p> This last <i>ls</i> shows what I expect, so I have the correct file system mounted. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-06T02:17:24+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240906-024555</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240906-024555</link><category>technology</category><category>opinion</category><title>Installing the new disk</title> <description> <![CDATA[ <div align="justify"> <p> So what do I do with my new 16 TB Seagate Exos disk? On the whole it seems to be a better idea to put it in <i>eureka</i>, where the old disk is. That would save issues with <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a> and <a href="https://en.wikipedia.org/wiki/Server_Message_Block">SMB</a>, and there's plenty of space for it. So reluctantly shut <i>eureka</i> down for the second time in a few days, put the disk in, cabled it up, rebooted into the BIOS setup menu. No new disk. </p> <p> What happened there? I know I've had cable problems in <i>eureka</i>—after all, it <i>is</i> <a href="http://www.lemis.com/grog/diary-jan2014.php?subtitle=New%20computer,%20finally&article=D-20140201-022258#D-20140201-022258">over 10 years old</a>—but I thought that this would work. </p> <p> What are the alternatives? <i>hydra</i> after all? It seems a messy way to go, since the web server files are on <i>/Photos</i>, and the web server runs on <i>eureka</i>. I already have an issue with <a href="http://www.lemis.com/yvonne/">Yvonne's</a> photos on <i>lagoon:/Photos</i>, for which I need to copy the photos to <i>eureka</i> (yes, potentially <a href="https://en.wikipedia.org/wiki/Symbolic_link">symlinks</a> would do the job, but so far I haven't looked at it; one way or another it's an issue). So: how about putting it in a test machine, copying the data and then moving the disk to the physical position of <i>/Photos</i> in <i>eureka</i>? But which machine? In principle I have three spare <a href="https://en.wikipedia.org/wiki/ThinkCentre">ThinkCentres</a>, but two are probably defective, and the third is still running as <i>tiwi</i> until I get my weather station software sorted out. </p> <p> But there are still the two machines from Bruce Evans. I've decided that I don't like the <a href="https://support.hp.com/au-en/document/c01709726">HP Z800</a>, which seems particularly hard to work with, but what about the other one, the one he called <i>zetaplex</i>? I haven't really looked at it very carefully, but it seems normal enough, if you ignore the silly lights: </p> <a id="Photo-2" name="Photo-2" href="diary-sep2024.php?dirdate=20240906&imagesizes=112&topics=c#Photo-2"> <img alt="This should be quartet-7.jpeg. Is it missing?" border="0" id="Photo_2" title="Photo quartet-7.jpeg. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240906/small/quartet-7.jpeg" width="300" height="225" /></a> <p> I'm sure that wasn't Bruce's choice. It's comparable to but about 11% faster than <i>eureka</i> (<a href="https://www.cpubenchmark.net/compare/2027vs2275/Intel-i7-4771-vs-Intel-i7-4790K">Intel i7 4790K instead of Intel i7 4771</a>). It has enough space for all the disks on <i>eureka</i>, but only 16 GB of memory instead of 32 GB. Still, a good backup to have in case <i>eureka</i> dies. </p> <p> The disk is in a tray with cables too short to remove it, possibly an issue with the assembly. I'll look at that later. But to my surprise I found 2 <a href="https://en.wikipedia.org/wiki/SSD">SSDs</a> almost completely hidden in other trays, both too polite to mention their capacity; I'll look at that when the copying is done. Put in my <i>quartet</i> SSD, the one I set up <a href="http://www.lemis.com/grog/diary-jun2024.php?subtitle=More%20VirtualBox%20fun&article=D-20240629-005229#D-20240629-005229">a couple of months ago</a>, and of course the new disk. Power up, into <a href="https://en.wikipedia.org/wiki/BIOS">BIOS</a> setup. Not seen! Of course, modern BIOS/<a href="https://en.wikipedia.org/wiki/UEFI">UEFI</a> screens are really hard to understand. Booted into <a href="https://www.freebsd.org/">FreeBSD</a> anyway. It saw the disk. Reboot, into the BIOS screen. It saw the disk. </p> <p> Is there something about the disk that wants somebody to throw the first stone? Would it have worked in <i>eureka</i>? I couldn't be bothered to recable all over again, so I'll never know. </p> <p> So, time to integrate into the LAN. But I couldn't access the net! After some examination, I found that the box has three Ethernet interfaces, like the Z800 (why?). But there seem to be only two connections. My startup script had assigned the same IP address to both <i>em0</i> and <i>re0</i>, while <i>em1</i> remained unconfigured. Connect the cable to <i>em0</i> (on the motherboard) and all was well. </p> <p> Then a custom <i>newfs</i> for a disk with large files, something like I decided upon <a href="http://www.lemis.com/grog/diary-jun2016.php?subtitle=Backup%20disk%20hell&article=D-20160628-002411#D-20160628-002411">years ago</a>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/10)</font> <font color="red">~</font> <font color="blue">44</font> -> <b><tt>newfs -L Photos -O 2 -U -a 64 -b 16384 -d 16384 -e 2048 -f 2048 -g 3145728 -h 64 -i 2752512 -m 1 -o space /dev/ada1p1</tt></b></code> </tt></div></blockquote> <p> After writing what seem to be far too many cylinder groups, it appeared to hang creating the journal, about 90 seconds. But it was apparently accessing the disk the whole time. </p> <p> And that gave me the biggest file system I have ever seen: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@quartet (/dev/pts/2)</font> <font color="red">~</font> <font color="blue">23</font> -> <b><tt>df /Photos /newphotos/</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity Mounted on<br />eureka:/Photos 7,629,565 6,806,391 746,877 90% /Photos<br />/dev/ada1p1 15,257,008 0 15,104,437 0% /newphotos<br /> </tt></div></blockquote> <a name="IBM-3330" id="IBM-3330"></a> <p> That corresponds to nearly 10,000 <a href="https://en.wikipedia.org/wiki/IBM_3330">IBM 3330</a> subsystems, my old yardstick for disk capacity. If I interpret <a href="https://gemini.google.com/app/12f6ded0cb11427e">this incorrect answer</a> correctly, a 3330 cabinet has a footprint of 7.65 m², so we're looking at about 7.6 <a href="https://en.wikipedia.org/wiki/Hactare">ha</a> of cabinets, on top of which you'd need probably the same amount to move round the things. </p> <p> Then the copy, using <i>tar</i> across <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a>, running close to wire speed at between 90 and 110 MB/s. At that speed it should take about a day, filling the first IBM 3330 equivalent in about 20 seconds. But while looking at the results, I discovered that a lot of the directories had a modification timestamp 1 January 2009. The system clock had lost its memory, and that must have been the <a href="https://en.wikipedia.org/wiki/Epoch_(computing)">epoch</a> for the BIOS. OK, run an <i>ntpdate</i> and fire up <i>ntpd</i>. </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@quartet (/dev/pts/3)</font> <font color="red">~</font> <font color="blue">19</font> -> <b><tt>date</tt></b></code><br />Thu 1 Jan 2009 14:21:09 AEDT<br /><code><font color="blue">=== root@quartet (/dev/pts/3)</font> <font color="red">/home/grog</font> <font color="blue">1</font> -> <b><tt>ntpdate eureka</tt></b></code><br /> 5 Sep 18:25:31 ntpdate[2069]: step time server 192.109.197.137 offset +494744636.627861 sec<br />You have mail in /var/mail/grog<br /><code><font color="blue">=== root@quartet (/dev/pts/3)</font> <font color="red">/home/grog</font> <font color="blue">2</font> -> <b><tt>date</tt></b></code><br />Thu 5 Sep 2024 18:26:46 AEST<br /><code><font color="blue">=== root@quartet (/dev/pts/3)</font> <font color="red">/home/grog</font> <font color="blue">3</font> -> <b><tt>ntpd</tt></b></code><br />daemon control: got EOF<br /> </tt></div></blockquote> <p> Huh? What does that mean? <i>ntpd</i> wasn't running. More web surfing brought me to <a href="https://www.linuxquestions.org/questions/showthread.php?p=6505352">this article</a>, which at least confirmed that it's not a <a href="https://www.freebsd.org/">FreeBSD</a> issue. Use the <tt>-4</tt> to avoid trying <a href="https://en.wikipedia.org/wiki/IPv6">IPv6</a>. But that didn't help. On the other hand, <i>/var/log/messages</i> showed: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 5 18:31:15 quartet ntpd[2088]: ntpd 4.2.8p18-a (1): Starting<br />Sep 5 18:31:15 quartet ntpd[2088]: Command line: ntpd<br />Sep 5 18:31:15 quartet ntpd[2089]: bind(23) AF_INET 192.109.197.151:123 flags 0x19 failed: Address already in use<br />Sep 5 18:31:15 quartet ntpd[2089]: unable to create socket on re0 (3) for 192.109.197.151:123<br />Sep 5 18:31:15 quartet ntpd[2088]: daemon child died with signal 11<br />Sep 5 18:31:15 quartet kernel: pid 2089 (ntpd), jid 0, uid 0: exited on signal 11 (core dumped)<br /> </tt></div></blockquote> <p> More web surfing, no insights. But one thing stood out: “unable to create socket on re0”. That's what comes of having multiple interfaces with the same IP address. Reset <i>re0</i> and it found <i>em0</i> and ran. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-06T02:45:55+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240907-015305</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240907-015305</link><category>technology</category><category>opinion</category><title>Disk copy, continued</title> <description> <![CDATA[ <div align="justify"> <p> Into the office this morning to discover that my disk copy had failed after transferring 4 TB: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>...<br />x ./grog/20160115/orig/P1153725.jpg<br />x ./grog/20160115/orig/P1153728.jpg<br />x ./grog/20160115/orig/P1153729.jpgtar: (null)<br />: Truncated tar archive: Unknown error: -1<br />tar: Error exit delayed from previous errors.<br /> </tt></div></blockquote> <p> That wasn't completely unexpected. OK, continue with <a href="https://en.wikipedia.org/wiki/rsync">rsync</a>, something that I'll have to do anyway a couple of times. In the process, discovered that the <tt>--delete-after</tt> option changes <i>rsync's</i> behaviour: it first builds a list of files, which takes some time. I wanted immediate results, so I restarted without: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/var/tmp</font> <font color="blue">34</font> -> <b><tt>Log rsync -av --delete-after /Photos/ /newphotos</tt></b></code><br />===== Fri 6 Sep 2024 09:30:09 AEST on quartet.lemis.com: rsync -av --delete-after /Photos/ /newphotos<br />building file list ... ^C<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/var/tmp</font> <font color="blue">35</font> -> <b><tt>Log rsync -av /Photos/ /newphotos</tt></b></code><br />===== Fri 6 Sep 2024 09:32:22 AEST on quartet.lemis.com: rsync -av /Photos/ /newphotos<br />sending incremental file list<br />.sujournal<br />rsync: [receiver] rename "/newphotos/.sujournal.8MXurP" -> ".sujournal": Operation not permitted (1)<br />grog/19640828/orig/hide/<br />grog/19640828/orig/hide/img466.tif<br /> </tt></div></blockquote> <p> And of course it ran all day. But when it was done, I had: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,806,544 746,725 90% 1,764,224 1,479,422 54% /Photos<br />/dev/ada1p1 15,257,008 7,237,056 7,867,381 48% 1,877,934 4,608,208 29% /newphotos<br /> </tt></div></blockquote> <p> Why the discrepancy? The used space can depend on the <i>newfs</i> parameters, which I should compare. But more obvious is the number of <a href="https://en.wikipedia.org/wiki/inode">inodes</a> used, 115,000 more than <i>/Photos</i>. That looks like link breakage. Where? Mañana. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-07T01:53:05+00:00</dc:date></item> <!-- topic po not selected <p> How do I take a photo of <i>quartet</i>, previously <i>zetaplex</i>? A photo of a computer is straightforward enough, but in this case I wanted to highlight this silly blue illumination, so I didn't want much extraneous light, and the light levels were low. OK, no worries, pull down the blinds to minimize external illumination and take at high <a href="https://en.wikipedia.org/wiki/Film_speed#Digital_camera_ISO_speed_and_exposure_index">ISO</a>: </p> <p> Acceptable, but not good. In particular, the aperture f/1.4 means that the back is out of focus, and the noise could be better: </p> <p> OK, mount on a tripod and try a longer exposure at 200/24° ISO: </p> <p> That's an improvement, but there's not much detail in the computer below <i>quartet</i> (<i>distress</i>). </p> <p> How about <a href="https://en.wikipedia.org/wiki/High-dynamic-range_imaging">HDR</a>? Two different ways, one with <a href="http://hugin.sourceforge.net"><i>Hugin</i></a> tools and one with <a href="http://www.hdrsoft.com/"><i>Photomatix</i></a>: </p> <p> There's surprisingly little difference from the normal shots. </p> <p> OK, mobile phones are supposed to be as good as real cameras. Try that: </p> <p> Oh. Of course, they have fixed focal lengths. To be fair, I didn't think to try the “telephoto lens”, which I think doubles the focal length. But never mind, I can crop: </p> <p> But that has terrible detail: </p> <p> Or I can come closer and accept the perspective distortion: </p> <p> That doesn't help much either. Detail is terrible, and there's clear evidence of camera shake: </p> <p> So mobile phones are really not worth the trouble. It looks as if my best choice is the old camera on tripod and more aggressive postprocessing. </p> <p> In passing, the <a href="https://en.wikipedia.org/wiki/Exif">Exif</a> data for the phone claims +5 <a href="https://en.wikipedia.org/wiki/Exposure_value">EV</a> exposure compensation. That looks like a misunderstanding on the part of the firmware programmers. There was no compensation. </p> End deselected topic po (Computer photo challenge) --> <!-- topic aho not selected <p> Found <a href="http://www.lemis.com/grog/Albums/Animals/Mona.php">Mona</a> looking interestedly at the air conditioner return air filter. And indeed, I had heard some noise in there. Opened the filter holder, and in went Mona. Within 10 seconds she had caught <i>and killed</i> a mouse: </p> <p> Has she done that before? My guess is no. I think the killing at this point was accidental. Cats play with mice before killing them, and Mona showed interest in playing with it. But the mouse wasn't interested: </p> <div align="left"> </div> <p> After a bit of messing around, Mona walked off, while <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> spent another 10 minutes or so playing with it. It's interesting to note that he didn't interfere with Mona at all. Ultimately he lost interest too, and I gave it to <a href="http://www.lemis.com/grog/Albums/Animals/Larissa.php">Larissa</a>, who carried it off and presumably ate it. </p> <blockquote class="fullwidth"> <div class="listitemwidth"> <i>So, was it a mouse? I had looked at the tail and thought a young rat. And <a href="http://www.lemis.com/grog/diary-sep2024.php?subtitle=Dinner%20with%20Pene%20Kirk&article=D-20240908-020503#D-20240908-020503">on the following day</a> Pene Kirk agreed with me.</i> </div> </blockquote> End deselected topic aho (A mouse? Rats!) --> <!-- topic ko not selected <p> We have a lot of carrots that have to be eaten quickly. What do I do with them? A fantasy stir-fry, in the process trying this spice mix that is also approaching its use-by date: </p> <p> The result: </p> <p> Edible, not much more. The paste is sticky, and the 80 g I used were barely enough. Next time I'll use a whole sachet (120 g). </p> End deselected topic ko (Another ALDI spice mix) --> <!-- topic go not selected <p> Gradually the weather is getting warmer, and more flowers are appearing: </p> <div align="left"> </div> <p> Jesse Walsh came along to tidy out the “trough” round the terrace, and <a href="http://www.lemis.com/yvonne/">Yvonne</a> was able to find at short notice somebody known only as “Eugene” to come and collect the leftovers: </p> <p> Eugene tells me that he had a friend in <a href="https://en.wikipedia.org/wiki/Geelong">Geelong</a> called Phil Lehey. That would be interesting, but I suspect that he had misunderstood the spelling of the surname. </p> End deselected topic go (Spring on its way) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240908-020443</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240908-020443</link><category>technology</category><category>opinion</category><title>Understanding the disk copy issues</title> <description> <![CDATA[ <div align="justify"> <p> So why is the copy of my <i>/Photos</i> disk so much bigger? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@quartet (/dev/pts/2)</font> <font color="red">~</font> <font color="blue">31</font> -> <b><tt>df -i /Photos /newphotos/</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,806,976 746,292 90% 1,764,291 1,479,355 54% /Photos<br />/dev/ada1p1 15,257,008 7,237,056 7,867,381 48% 1,877,931 4,608,211 29% /newphotos<br /> </tt></div></blockquote> <p> It strongly suggests <a href="https://en.wikipedia.org/wiki/Symbolic_link">symlinks</a> changed to separate files. Off for a look, and discovered—not surprisingly—the main discrepancy in <i>/newphotos/grog</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/1)</font> <font color="red">/Photos</font> <font color="blue">83</font> -> <b><tt>du -s /Photos/grog</tt></b></code><br />6603970 /Photos/grog<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/newphotos</font> <font color="blue">97</font> -> <b><tt>du -s /newphotos/grog</tt></b></code><br />7011999 /newphotos/grog<br /> </tt></div></blockquote> <p> Those are sizes in megabytes! 408 <b>GB</b> size difference. OK, go down a level. The big difference was: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>eureka: 93834 www<br />quartet: 383635 www<br /> </tt></div></blockquote> <p> Really? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/1)</font> <font color="red">~</font> <font color="blue">85</font> -> <b><tt>du -s /Photos/grog/www</tt></b></code><br />383632 /Photos/grog/www<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">~</font> <font color="blue">99</font> -> <b><tt>du -s /newphotos/grog/www</tt></b></code><br />383635 /newphotos/grog/www<br /> </tt></div></blockquote> <p> Yes, this must be a link issue. </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/1)</font> <font color="red">~</font> <font color="blue">86</font> -> <b><tt>find /Photos/grog/www -type l | wc -l</tt></b></code><br /> 1099<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">~</font> <font color="blue">100</font> -> <b><tt>find /newphotos/grog/www -type l | wc -l</tt></b></code><br /> 1099<br /> </tt></div></blockquote> <p> Oh. Exactly the same number of symlinks. In fact, both disks have exactly 4005 symlinks. Not the issue. </p> <p> “Hard” links? You'd see that from the link count. But the directories that I looked at all had exactly the same content and link counts. Is it possible that two issues are ganging up on me? The “big” files in the <i>www</i> directories are supposed to be linked to the corresponding file in two levels above in the hierarchy. For example: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@eureka (/dev/pts/1)</font> <font color="red">~</font> <font color="blue">89</font> -> <b><tt>ls -l /Photos/grog/20240903/Hot-water-system-1.jpeg /Photos/grog/www/20240903/big/Hot-water-system-1.jpeg</tt></b></code><br />-rwxr--r-- 1 grog wheel 3,270,177 3 Sep 15:37 /Photos/grog/20240903/Hot-water-system-1.jpeg<br />-rwxr--r-- 1 grog wheel 3,270,177 3 Sep 15:37 /Photos/grog/www/20240903/big/Hot-water-system-1.jpeg<br /> </tt></div></blockquote> <p> But they're not! I need to see whether I have broken something, and whether older images are correctly linked. In that case there's more to be repaired. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-08T02:04:43+00:00</dc:date></item> <!-- topic hk not selected <p> It's been some time since we have had guests for dinner. Ten years ago Chris Bahlo was a regular visitor, but she has now found a life, and we no longer see her very often. But today we were able to host Pene Kirk, and <a href="http://www.lemis.com/yvonne/">Yvonne</a> went to some trouble: bisque de fruits de mer, <a href="http://www.lemis.com/grog/recipes/bef-stroganoff.php">bef Strogranoff</a> and poires gourmandes: </p> <p> Somehow we spent more time talking about animals and health than I expected. Pene was one of the people who recommended <a href="https://en.wikipedia.org/wiki/Cataract">cataract</a> surgery when I was thinking about it, and she confirmed <a href="http://www.lemis.com/grog/diary-mar2024.php?subtitle=Simulating%20cataract%20surgery%20improvement&article=D-20240324-004313#D-20240324-004313">my experience</a> that the cataract diminishes colour recognition and makes everything look a bit yellowish. And looking at yesterday's “mouse”, she agreed that it looked more like a juvenile rat. My assumption was based on the tail, which was hairless, but Pene based her assessment on the shape of the head. </p> End deselected topic hk (Dinner with Pene Kirk) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240909-015601</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240909-015601</link><category>general</category><category>technology</category><category>opinion</category><title>Unwise actions</title> <description> <![CDATA[ <div align="justify"> <p> Received in the mail today: </p> <a id="Photo-3" name="Photo-3" href="diary-sep2024.php?dirdate=20240908&imagesizes=1112&topics=c#Photo-3"> <img alt="This should be Unwise-detail.png. Is it missing?" border="0" id="Photo_3" title="Photo Unwise-detail.png. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240908/small/Unwise-detail.png" width="400" height="443.72469635628" /></a> <p> Yes, I'm still with <a href="https://www.wise.com/">“Wise”</a>, despite my annoyance last month. My searches show that they're the best of a bad lot. But now they're asking me to migrate my secure desktop environment to my mobile phone. Or are they? This is just an email, asking me to log in. Most emails of that nature are scams. What do they say on their web site? </p> <p> How do I access the web site? Once again these <b><i>HORRIBLE</i></b> <a href="https://en.wikipedia.org/wiki/CAPTCHA">CAPTCHAs</a>, asking me to identify fire hydrants, whatever they may be. A quick Google search <a href="https://www.bundabergnow.com/2021/04/06/fire-hydrant-inspection-program-underway/">shows</a>: </p> <a href="https://www.bundabergnow.com/wp-content/uploads/2021/04/hydrant.jpg"> <img alt="" title=". Click to see in original size." src="https://www.bundabergnow.com/wp-content/uploads/2021/04/hydrant.jpg" width="300" /></a> <p> But there was nothing like that on the CAPTCHA. OK, “Wise” are a British company. <a href="https://www.rbfrs.co.uk/fire-hydrants/what-is-a-fire-hydrant/">What do British fire hydrants look like?</a> Pretty much the same: </p> <a href="http://www.lemis.com/grog/Day/20240908/hydrants1.jpeg"> <img alt="" title=". Click to see in original size." src="http://www.lemis.com/grog/Day/20240908/hydrants1.jpeg" width="300" /></a> <p> But it seems that the CAPTCHA wants me to identify <a href="https://en.wikipedia.org/wiki/USA">US</a> fire hydrants! How can they expect me to know what they look like? That reminds me of other things they want to know, like identifying “crosswalks”, whatever they may do. My guess is that <a href="https://en.wikipedia.org/wiki/AI">AI</a> can do better than I can. </p> <p> Finally fought my way in and checked on the app. <b><i>NOTHING AT ALL!</i></b> No warnings, no mention of an app. What kind of security is that? About the only thing that looks kosher is that the app is available in the <a href="https://play.google.com/store/apps/details?id=com.transferwise.android">Google app store</a>. Signed in and got mail from Wise confirming it, another marginal proof that it's legitimate. And it allows me to sign in with a thumbprint. </p> <p> Am I really overly critical? There are so many warning bells, and even the idea of using a mobile phone for anything related to security scares me. The app description includes: </p> <blockquote class="fullwidth"> <div class="listitemwidth"> Safety starts with understanding how developers collect and share your data. Data privacy and security practices may vary based on your use, region, and age. The developer provided this information and may update it over time. </div> </blockquote> <p> Well, for me safety starts with understanding how malicious people can abuse the system. On the whole I trust the developers. Since they asked me to review it, I gave it 2 stars with the comment: </p> <blockquote class="fullwidth"> <div class="listitemwidth"> This app scares me. "Wise" send me email asking me to install it by logging in from the email message. Their web site doesn't mention it. This has all the hallmarks of an exploit. Hopefully I'm wrong. </div> </blockquote> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-09T01:56:01+00:00</dc:date></item> <!-- topic ko not selected <p> Jane Ashhurst is a fan of Japanese food, something that I still barely understand. We had planned to make <a href="https://en.wikipedia.org/wiki/Sushi">sushi</a>, something I have never made before, but I decided that it might be a good idea to try it first. I have a sushi kit that I bought from <a href="https://en.wikipedia.org/wiki/ALDI">ALDI</a> years ago—so long that it has passed its use-by date—and today I tried making some. </p> <p> The instructions for the pack were short but helpful. 300 g of rice. Cook with (only!) 400 ml water, add vinegar (which proved to be 40 ml) to the mess and spread on the <a href="https://en.wikipedia.org/wiki/Nori">nori</a>. Add filling ingredients. </p> <p> After preparing the filling, back to look at the <a href="https://youtu.be/nIoOv6lWYnk">video</a> that I located <a href="http://www.lemis.com/grog/diary-jun2024.php?subtitle=Sushi?&article=D-20240623-032843#D-20240623-032843">a couple of months ago</a>: </p> <iframe width="400" height="300" src="http://www.youtube.com/embed/nIoOv6lWYnk" allowfullscreen frameborder="0" > </iframe> <p> A little late. I had done a number of things differently, including cutting the ingredients finely. But it was of help, and I followed it as best I could. </p> <p> That wasn't very best: </p> <p> Oh. Not quite the way it looked in the video. Try again: </p> <p> At first sight that looked better, but only at first sight. As soon as I touched it, it self-destructed: </p> <p> OK, the video shows the cook using half sheets of nori. How about whole sheets? </p> <p> That looks marginally better, but still nothing like good. After a bit more experimentation, ended up with: </p> <p> What went wrong? The biggest issue is wrapping. How should the ends of the nori overlap? Our expert cook didn't have an issue, of course, but I still don't know how to do it right. Yes, leave the ends free of rice, but how do I get them to line up? I need to find a video “how to ruin sushi”. </p> <p> In passing, it seems that a full width of nori takes about 100 g of rice, much more than I expected. And it was good to have the vinegar measured out in the kit. states to use 20% <i>by volume</i> vinegar to <i>dry</i> rice (which, of course, is already cooked). But the instructions seem to be right: 40 g vinegar to 300 g dry rice, or 13⅓%. </p> End deselected topic ko (Sushi?) --> <!-- topic ko not selected <p> What goes with <a href="https://en.wikipedia.org/wiki/Sushi">sushi</a>? <a href="https://en.wikipedia.org/wiki/Miso_soup">Miso soup</a>, it seems. And I have plenty of shiro <a href="https://en.wikipedia.org/wiki/Miso">miso</a>. Just find a recipe. It seems that a typical recipe is with <a href="https://en.wikipedia.org/wiki/Tofu">doufu</a> and spring onions. But I need <a href="https://en.wikipedia.org/wiki/Dashi">dashi</a>. I can make that in various ways. One is with seaweed. </p> <p> <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=Jane%20Ashhurst%20visits&article=D-20240814-023243#D-20240814-023243">Last month</a> Jane Ashhurst brought some ingredients with her, including this: </p> <p> What is it? There was nothing written on the bag she put it in. My guess is that it's some kind of <a href="https://en.wikipedia.org/wiki/Kelp">kelp</a>, but what? I'll have to ask. </p> <p> But then there was this: </p> <p> Jane can't have bought that. About the only thing that I could read, apart from the pretty useless “Nutrition Facts”, was that it expired nearly 10 years ago. And I have a number of bags of it. What is it? With some searching I found the word <a href="https://en.wikipedia.org/wiki/Wakame">wakame</a> hidden on the side. And the rest? <a href="https://en.wikipedia.org/wiki/Google_Translate">Google Translate</a> to the rescue: </p> <p> Use in miso soup! 1 g for who knows how much soup? That seemed a little, so I put in 2 g: </p> <p> I don't think I've ever seen anything swell up that much. In the end I threw half of it out again. </p> <p> Next, how much miso? One recipe called for 2 tablespoons in 4 cups, whatever that may mean. It seems that there are 16 US tablespoons to a US cup, so that's 1/32 of the volume. While searching, came up with another ratio of miso to water, ¼ teaspoon to 1 cup. If I have that right, it's a ratio of 1:192, almost nothing. To be generous, I took a 1:20 ratio: 500 g water, 25 g miso, 30 g spring onion and 65 g (the size of the block) of doufu. </p> <p> The result? Boring. Certainly not too much miso. After some investigation, it seems that the wakame is not a substitute for dashi, as this photo from the <a href="https://en.wikipedia.org/wiki/Miso_soup">Wikipedia page</a> shows </p> <a href="https://upload.wikimedia.org/wikipedia/commons/thumb/e/e8/Miso_Soup_001.jpg/2560px-Miso_Soup_001.jpg"> <img alt="" title=". Click to see in original size." src="https://upload.wikimedia.org/wikipedia/commons/thumb/e/e8/Miso_Soup_001.jpg/2560px-Miso_Soup_001.jpg" width="300" /></a> <p> So for next time we need to find some good dashi. </p> End deselected topic ko (Miso soup) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240910-021044</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240910-021044</link><category>technology</category><category>opinion</category><title>NBN advice for morons</title> <description> <![CDATA[ <div align="justify"> <p> The <a href="https://www.nbnco.com.au/">National Broadband Network</a> sends information emails from time to time, maybe monthly. I've only just found out about them and signed up. I don't know why I bothered. </p> <p> Today I got a newsletter offering <a href="https://www.nbnco.com.au/corporate-information/about-nbn-co/newsroom/articles/customer/avoid-the-wi-fi-gap-get-the-internet-you-need">Avoid the ‘Wi-Fi gap’ and get the internet you need</a> (capitalization original). It also offers a <a href="https://www.nbnco.com.au/corporate-information/about-nbn-co/newsroom/articles/customer/avoid-the-wi-fi-gap-get-the-internet-you-need#too-long-didnt-read">TL:DR</a> (punctuation also original), which can be summarized: buy new equipment and a faster Internet link. No mention of troubleshooting of any kind. Thank you, NBN, for reinforcing my negative impressions. </p> <p> To be fair, though, there was more information about the upcoming outages: <a href="https://www.nbnco.com.au/support/network-status/nbn-fixed-wireless-outages">this page</a> explains that they're upgrading the system to new “5G mm Wave technology” (yet again original typography). What's that? I hadn't heard of it, and there's <a href="https://www.celona.io/5g-lan/5g-mmwave">little to be found on the web</a> so far. It's spelt “mmWave” and refers to frequencies of, well, millimetre wavelengths, specifically above 24 GHz. OK, well and good. The bandwidths are more interesting. </p> <p> And there's more about that <a href="https://www.nbnco.com.au/residential/upgrades/more-fixed-wireless#accordion-17dd10920d-item-9a853f7e1a">here</a>: </p> <blockquote class="fullwidth"> <p class="listitemwidth"> nbn Fixed Wireless Home Fast offers peak wholesale download speeds between 200-250 and upload speeds between 8-20 Mbps. </p> <p class="listitemwidth"> nbn Fixed Wireless Superfast offers peak wholesale download speeds of 400Mbps and upload speeds between 10-40 Mbps. </p> <p class="listitemwidth"> Plus, we’ve boosted speeds on our most popular plan; Fixed Wireless Plus. Wholesale speed capability has now been lifted from up to 75/10Mbps to up to 100/20Mbps. </p> </blockquote> <p> That does sound interesting. What does <a href="https://aussiebroadband.com.au/">Aussie Broadband</a> have to say about that? Oh, another of these stupid <a href="https://en.wikipedia.org/wiki/2FA">2FA</a> things, and after accepting it, the page hangs. Aussie, can you do anything right any more? </p> <p> And the outages? <a href="https://www.nbnco.com.au/support/network-status/nbn-fixed-wireless-outages">The first page</a> contains information about the upcoming outages, which <a href="http://www.lemis.com/grog/diary-aug2024.php?subtitle=Bloody%20NBN%20again!&article=D-20240823-003119#D-20240823-003119">I calculated</a> as being 37 hours </p> <blockquote class="fullwidth"> <p class="listitemwidth"> During these upgrade works, you may experience reductions in service and periodic outages as we roll out the upgrades in your area over several weeks. In most cases, the outages may range from 3-4 minutes. However, for others, some periodic outages may occur throughout the day and last for up to 12 hours. While the upgrade works are being completed in your area, you may experience more than one outage. We’ll notify your provider of any outages ahead of time, so they can keep you informed. We understand how frustrating outages can be, which is why we’re working hard to keep these to a minimum and reduce the occurrence of outages wherever we can. </p> <p class="listitemwidth"> How we’re minimising the impact of potential outages: </p> <ul> <li class="fullwidth"> <div class="list2width"> We’re using temporary antennas to help maintain service for many customers. </div> </li> <li class="fullwidth"> <div class="list2width"> We’re migrating services to other local towers, where viable. </div> </li> <li class="fullwidth"> <div class="list2width"> We’re preparing equipment before arriving on site to minimise downtime. </div> </li> </ul> </blockquote> <p> “In most cases, the outages may range from 3-4 minutes.” Admittedly, it can be longer. And I particularly appreciate the fact that they're preparing in advance. </p> <p> And of course there's the other shoe: if I get connected to 5G mmWave, they'll have to change the <del>antenna</del> ODU. More downtime. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-10T02:10:44+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240910-030520</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240910-030520</link><category>technology</category><category>general</category><category>opinion</category><title>Academia surpasses itself</title> <description> <![CDATA[ <div align="justify"> <p> For years I've been getting email from <a href="https://en.wikipedia.org/wiki/Academia">Academia</a> with questions like “Greg Lehey ✏️ Did you write "Treasurer"?”. To answer I first need to pay them money! I can do without that. </p> <p> But today I got the ultimate question: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>131 ND 08-09-2024 To academia@lem ( 903) Mentioned by Greg Le $1, 30 day trial. “G. Lehey” mentioned by “Greg Lehey” </tt></div></blockquote> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-10T03:05:20+00:00</dc:date></item> <!-- topic ko not selected <p> <a href="http://www.lemis.com/grog/diary-apr2024.php?subtitle=Finally%20a%20chili%20sauce?&article=D-20240429-002628#D-20240429-002628">A few months back</a> I found a <a href="http://www.lemis.com/grog/recipes/chili-sauce.php">chili sauce mix</a> that I liked, at least for <a href="http://www.lemis.com/grog/recipes/huevos-rancheros.php"><i>huevos rancheros</i></a>. But I've finished the batch, time for a new batch. </p> <p> Why so much sesame oil? Why sesame oil at all? Despite the origins, this isn't intended for Korean food. So I left it out, and while I was at it, I added even more garlic. </p> <p> And the result? Tasted at least as good, but it was so watery. Why? My guess is that the <a href="https://en.wikipedia.org/wiki/Gochujang">gochujang</a> that I used last time wasn't the freshest, and it had dried out. </p> End deselected topic ko (Chili sauce revisited) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240911-021507</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240911-021507</link><category>technology</category><category>opinion</category><title>Understanding the photo size discrepancy</title> <description> <![CDATA[ <div align="justify"> <p> It's been <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=c&subtitle=Understanding%20the%20disk%20copy%20issues&article=D-20240908-020443#D-20240908-020443">a few days</a> since I investigated the cause of my excess data usage on the new <i>/Photos</i> disk. I had suspected issues with links between my photos processing directory and the “big” versions in the web hierarchy. Both images are the same, so they should be links, not copies. But preliminary investigations showed that they were copies on the original as well. </p> <p> That was for photos taken on <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=c#3">3 September 2024</a>. Was it always that way? Off looking. No! </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/Photos/grog</font> <font color="blue">119</font> -> <b><tt>l -i $DATE/*jpeg www/$DATE/big</tt></b></code><br />1480997 -rwxr--r-- 2 grog lemis 3,544,601 1 Jan 2015 20150101/Ceiling.jpeg<br /><br />www/20150101/big:<br />1480997 -rwxr--r-- 2 grog lemis 3,544,601 1 Jan 2015 Ceiling.jpeg<br /> </tt></div></blockquote> <p> That's the same file, as intended. Why did it stop? It seems that things changed in mid-April this year. But why? My scripts haven't changed in a couple of years. </p> <p> More investigation shows, in <i>/Photos/Tools/docopy</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>if [ `hostname -s` = "eureka" ]; then<br /> CP=ln<br />else<br /> CP="cp -p"<br />fi<br /> </tt></div></blockquote> <p> That makes sense, up to a point. At the the time, <a href="http://www.lemis.com/grog/diary-sep2020.php?subtitle=Integrating%20disks,%20day%203&article=D-20200906-034855#D-20200906-034855">4 years ago</a>, I moved <a href="http://www.lemis.com/yvonne/">Yvonne's</a> photos to a disk on her machine, <i>lagoon</i>. But the web files still needed to be on <i>eureka</i>. </p> <p> Now, however, I'm doing my processing on <i>hydra</i>, but accessing <i>eureka:/Photos</i> via <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a>. And I'm still copying! Yes, I could check for <i>hydra</i> too, but it makes more sense to check whether the source and destination are on the same file system. </p> <p> How do you do that? It seems that there should be a simple way, but I didn't find it. Instead I used the output of <i>df(1)</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>SRCDISK=`df .`<br />DSTDISK=`df ${WEBDIR}`<br />if [ "$SRCDISK" = "$DSTDISK" ]; then<br /> # Yes: link<br /> CP=ln<br />else<br /> # No: copy<br /> CP="cp -p"<br />fi<br /> </tt></div></blockquote> <p> And that works. A simple <tt>make web</tt> should do the trick: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/23)</font> <font color="red">/Photos/grog/20240831</font> <font color="blue">189</font> -> <b><tt>du -sc . ../www/20240831/</tt></b></code><br />21203 .<br />149 ../www/20240831/<br />21352 total<br /><code><font color="blue">=== grog@hydra (/dev/pts/23)</font> <font color="red">/Photos/grog/20240831</font> <font color="blue">190</font> -> <b><tt>df -i .</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,828,390 724,879 90% 1,765,566 1,478,080 54% /Photos<br /><code><font color="blue">=== grog@hydra (/dev/pts/23)</font> <font color="red">/Photos/grog/20240831</font> <font color="blue">191</font> -> <b><tt>make web/</tt></b></code><br />...<br /><code><font color="blue">=== grog@hydra (/dev/pts/23)</font> <font color="red">/Photos/grog/20240831</font> <font color="blue">192</font> -> <b><tt>df -i .</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,828,244 725,024 90% 1,765,588 1,478,058 54% /Photos<br /><code><font color="blue">=== grog@hydra (/dev/pts/23)</font> <font color="red">/Photos/grog/20240831</font> <font color="blue">193</font> -> <b><tt>du -sc . ../www/20240831/</tt></b></code><br />21203 .<br />3 ../www/20240831/<br />21206 total<br /> </tt></div></blockquote> <p> The 3 MB left in the <i>www</i> directory were the small and tiny images. </p> <p> Now to fix up <i>/Photos</i>: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/18)</font> <font color="red">/Photos/grog/www</font> <font color="blue">4</font> -> <b><tt>for i in 20240[4-9]*; do (cd ../$i && make web); done</tt></b></code> </tt></div></blockquote> <p> The result, before and after: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@hydra (/dev/pts/24)</font> <font color="red">~</font> <font color="blue">8</font> -> <b><tt>df -i /Photos/</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,828,244 725,024 90% 1,765,554 1,478,092 54% /Photos<br />...<br /><code><font color="blue">=== grog@hydra (/dev/pts/24)</font> <font color="red">~</font> <font color="blue">12</font> -> <b><tt>df -i /Photos/</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,820,747 732,522 90% 1,764,078 1,479,568 54% /Photos<br /> </tt></div></blockquote> <p> It doesn't look like much, but it's still a saving of 7.5 GB. </p> <p> So that explains the first issue. The second is the fact that apparently <i>all</i> the directories that I have looked at on <i>hydra</i> don't have linked files. Is that an issue with <a href="https://en.wikipedia.org/wiki/Samba_(software)">Samba</a> (unlikely) or because I started with <i>tar</i> and completed with <i>Samba</i>? The way either works would be to copy the date files separately from the <i>www</i> hierarchy. To be investigated. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-11T02:15:07+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240911-030504</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240911-030504</link><category>technology</category><category>opinion</category><title>Aussie web site hangs</title> <description> <![CDATA[ <div align="justify"> <p> Why can't I access the tariff information on the <a href="https://aussiebroadband.com.au/">Aussie Broadband</a> web site? Started writing an email and looked on the web site for information. Pop! Up came a “chat” window. OK, ask Dominic. No, site working fine. Will he report that it isn't for me? No, not worth the trouble. </p> <p> With a bit of insistence, along with my possibly visible negative response to the “are you happy with this response?” popup, he went off and investigated, something that took a total of round an hour, including a restart because of a timeout (7 minutes, he says) on his part. I ticked “send me the transcript” box for both chats, but I never saw one. </p> <p> The startling answer: it's because of my “plan”, only 25/5 Mb/s. It's no longer in the Aussie offerings, and the web code hangs because of some unexpected issue. Will they fix it? No. </p> <p> One thing that Dominic did give me was the <a href="https://www.aussiebroadband.com.au/internet/nbn-plans/">correct page to find the “plans”</a>, starting at $69 per month—they say. But all it offers is “Fixed wireless plus” (100/20) for $89 per month unlimited. It seems that that's all they currently offer. That's $14 more than I'm currently paying, so I'll pass. Presumably they'll force my hand some time. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-11T03:05:04+00:00</dc:date></item> <!-- topic Hko not selected <p> When I was a lad, I used to enjoy eating <a href="http://www.lemis.com/grog/recipes/penang-laksa.php">Penang laksa</a> on frequent occasions. </p> <p> Or did I? Looking back at my diary, I find only two mentions, both in (<a href="http://www.lemis.com/grog/diary-aug1966.php#24">August</a> and <a href="http://www.lemis.com/grog/diary-nov1966.php#24">November</a>) 1966. Then I ate much more in <a href="https://en.wikipedia.org/wiki/Sarawak">Sarawak</a> in 1969. But that was probably different, as noted by <a href="diary-people.php#Sonny_Rajah">Sonny Rajah's</a> rejection of it on <a href="http://www.lemis.com/grog/diary-aug1969.php?topics=c#26">26 August 1969</a>. Even then I only ate it 19 times. </p> <p> So it seems that my love of Penang Laksa is more a fond memory than anything that I really liked at the time. Still, when I found a way to make it myself <a href="http://www.lemis.com/grog/diary-jan2017.php?subtitle=Laksa%20again&article=D-20170118-004228#D-20170118-004228">nearly 8 years ago</a>. And since then I've eaten it on a regular basis, though it's no longer such a favourite, and I've found other things that I prefer for breakfast. </p> End deselected topic Hko (Laksa over the millennia) --> <!-- topic ko not selected <p> Richmond's <a href="http://www.lemis.com/grog/recipes/penang-laksa.php">Penang laksa</a> isn't really as sour as I like, so I add <a href="https://en.wikipedia.org/wiki/Tamarind">tamarind</a> to the mixture. That's easy enough—I have tamarind paste that I just mix in—but what's this? </p> <p> I've had it for ever, though it's too polite to say how old it is. It was a 375 g pack of “wet tamarind”, and now it only weighs 322 g and isn't wet at all. But it doesn't look as if it has gone bad, so I cut off some slices and tried to mix them with water: </p> <div align="left"> </div> <p> My guess is that it should have been strained, but I didn't do that. Despite everything, it tasted OK. </p> End deselected topic ko () --> <!-- topic Po not selected <p> Today was the highlight of the US election campaign, the debate between <a href="https://en.wikipedia.org/wiki/Donald_Trump">Donald Trump</a> and <a href="https://en.wikipedia.org/wiki/Kamala_Harris">Kamala Harris</a>. </p> <iframe width="400" height="300" src="http://www.youtube.com/embed/VgsC_aBquUE" allowfullscreen frameborder="0" > </iframe> <p> I watched a little of it, but clearly I wasn't the only one. And once again, the commentaries didn't match my own observations. But I can spend more time than I have reading the commentaries. The bottom line seems to be that Harris ran Trump into the ground, but that there's no particular reason that it will change the outcome. </p> <p> But the <a href="https://en.wikipedia.org/wiki/New_York_Times">New York Times</a> advertising department found it a good reason to subscribe. </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>From nytimes@e.newyorktimes.com Wed Sep 11 14:04:16 AEST 2024<br />Date: Wed, 11 Sep 2024 04:04:13 +0000<br />Subject: Sale: Save now before the U.S. debate: A$0.50 a week.<br /> </tt></div></blockquote> <p> That was clearly sent 1½ hours after the end of the debate. Is that a good advertisement for the timeliness of their reporting? </p> End deselected topic Po (Towards president Kamala) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240912-044332</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240912-044332</link><category>technology</category><category>opinion</category><title>Understanding shell syntax</title> <description> <![CDATA[ <div align="justify"> <p> While investigating my lost space on my new photo disk, entered this: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@eureka (/dev/pts/6)</font> <font color="red">/Photos/grog</font> <font color="blue">17</font> -> <b><tt>DATE=20101009 ls -li $DATE/*jpeg www/$DATE/big</tt></b></code><br />ls: /*jpeg: No such file or directory<br />ls: www//big: No such file or directory<br /> </tt></div></blockquote> <p> Huh? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@eureka (/dev/pts/6)</font> <font color="red">/Photos/grog</font> <font color="blue">18</font> -> <b><tt>DATE=20101009</tt></b></code><br /><code><font color="blue">=== grog@eureka (/dev/pts/6)</font> <font color="red">/Photos/grog</font> <font color="blue">19</font> -> <b><tt>ls -li $DATE/*jpeg www/$DATE/big</tt></b></code><br /><code><font color="blue">=== grog@hydra (/dev/pts/28)</font> <font color="red">/Photos/grog</font> <font color="blue">14</font> -> <b><tt>ls -li $DATE/*jpeg www/$DATE/big</tt></b></code><br /> 863295 -rwxrw-r-- 2 grog lemis 3099465 9 Oct 2010 20101009/dam-dup-panorama.jpeg<br />...<br />www/20101009/big:<br /> 863295 -rwxrw-r-- 2 grog lemis 3099465 9 Oct 2010 dam-dup-panorama.jpeg<br />...<br /> </tt></div></blockquote> <p> Why didn't it work the first time round? I do this sort of thing all the time. Some discussion on IRC, culminating in: yes, you can set environment variables like that, but they only get inserted into the environment of processes started by this command. They <i>don't</i> apply to the command itself. </p> <p> Somehow that seems counterintuitive. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-12T04:43:32+00:00</dc:date></item> <!-- topic ko not selected <p> I had some broccoli left over. What should I do with it? Build a breakfast around it: </p> <div align="left"> </div> <p> Nothing special, but for reference: </p> <table summary="Ingredients"> <tr> <td align="right"><b>quantity</b></td> <td> </td> <td><b>ingredient</b> </td> <td> </td> <td><b>step</b> </td> </tr> <tr> <td valign="top" align="right">67 g</td> <td> </td> <td valign="top" align="left">doufu</td> <td> </td> <td valign="top" align="right">1</td> </tr> <tr> <td valign="top" align="right">50 g</td> <td> </td> <td valign="top" align="left">beef</td> <td> </td> <td valign="top" align="right">1</td> </tr> <tr> <td> </td> </tr> <tr> <td valign="top" align="right">15 g</td> <td> </td> <td valign="top" align="left">spring onions</td> <td> </td> <td valign="top" align="right">2</td> </tr> <tr> <td valign="top" align="right">2</td> <td> </td> <td valign="top" align="left">shiitake mushrooms</td> <td> </td> <td valign="top" align="right">2</td> </tr> <tr> <td valign="top" align="right">38 g</td> <td> </td> <td valign="top" align="left">broccoli</td> <td> </td> <td valign="top" align="right">2</td> </tr> <tr> <td valign="top" align="right">17 g</td> <td> </td> <td valign="top" align="left">red capsicum</td> <td> </td> <td valign="top" align="right">2</td> </tr> <tr> <td> </td> </tr> <tr> <td valign="top" align="right">30 g</td> <td> </td> <td valign="top" align="left">light soya sauce</td> <td> </td> <td valign="top" align="right">3</td> </tr> <tr> <td valign="top" align="right">12.5 g</td> <td> </td> <td valign="top" align="left">dark soya sauce</td> <td> </td> <td valign="top" align="right">3</td> </tr> <tr> <td valign="top" align="right">30 g</td> <td> </td> <td valign="top" align="left">shiro miso</td> <td> </td> <td valign="top" align="right">3</td> </tr> <tr> <td> </td> </tr> <tr> <td valign="top" align="right">150 g</td> <td> </td> <td valign="top" align="left">Shandong ramen noodles</td> <td> </td> <td valign="top" align="right">4</td> </tr> <tr> <td> </td> </tr> </table> <p> About the only thing of interest is how to fry the doufu. It sticks to the pan: </p> <p> There must be a way of preventing that. Heat in a microwave oven first? </p> End deselected topic ko (More breakfast experiments) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240913-023625</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240913-023625</link><category>technology</category><category>opinion</category><title>Disk copy pain</title> <description> <![CDATA[ <div align="justify"> <p> Copying my photo disk to the new 16 TB drive still isn't done! My first attempt ended up with far more <a href="https://en.wikipedia.org/wiki/inode">inode</a> and data use than the old disk. Part of that was because I used the wrong invocation: I omitted the <b>H</b> option, meaning that <i>rsync</i> created new files for each copy of a link. OK, start again with <i>/Photos/grog</i> to see if it works. </p> <p> But it put the result in <i>/newphotos/grog/grog</i>, and I couldn't find a way to stop it. Next time I'll try something like <i>cd</i> to the source directory and referring to it as <b>.</b> (dot), specifying the destination accordingly. But while I looked at it, it occurred to me that my file system parameters were significantly different from the old <i>/Photos</i>. After removing size parameters, I had: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><i>/Photos</i>: newfs -L Photos -O 2 -U -a 4 -b 32768 -d 32768 -e 4096 -f 4096 -g 3145728 -h 64 -i 2469888 -j -k 1152 -m 1 -o space /dev/ada1p1<br /><i>/newphotos</i>: newfs -L Photos -O 2 -U -a 64 -b 16384 -d 16384 -e 2048 -f 2048 -g 3145728 -h 64 -i 2752512 -m 1 -o space /dev/ada1p1<br /> </tt></div></blockquote> <p> OK, what the hell, let's start again from scratch. <i>umount</i> <i>/newphotos</i>, <i>newfs</i> it again with the new parameters, and start again. </p> <p> <i>umount</i> hung! I couldn't stop it, but it was running and consuming CPU time: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== grog@quartet (/dev/pts/2)</font> <font color="red">~</font> <font color="blue">121</font> -> <b><tt>date; ps aux | grep umount</tt></b></code><br />Thu 12 Sep 2024 13:33:10 AEST<br />root 23962 2.4 0.0 12728 1820 0 DL+ 13:19 0:15.09 umount /newphotos<br />...<br /><code><font color="blue">=== grog@quartet (/dev/pts/2)</font> <font color="red">~</font> <font color="blue">122</font> -> <b><tt>date; ps aux | grep umount</tt></b></code><br />Thu 12 Sep 2024 13:38:15 AEST<br />root 23962 2.4 0.0 12728 1820 0 DL+ 13:19 0:25.72 umount /newphotos<br /> </tt></div></blockquote> <p> That's round 2 seconds of CPU time per minute. It was also performing about 200 I/Os per second. It finally finished at 13:48. Extrapolating, that's round 45 seconds of CPU time and 27,000 I/O transfers, at a guess 250 MB. What was it doing? </p> <p> Finally ran <i>newfs</i> and started the transfer again. But how? <i>tar</i> seems to be the right way to go, but last time it failed in mid-transfer. For no reason that still seems good I decided to write a <i>tar</i> archive on the new disk. And that set off merrily with a transfer rate of up to 111 MB/s, pretty much exactly the bandwidth of the connection, and seldom dropping below 100 MB/s. </p> <p> OK, how about a compressed <i>tar</i>? That way I might miss the bottleneck of the wire. But no, the transfer rate dropped to round 35 MB/s: the compression process maxed out a CPU. And there's no way I could get a 3-fold compression with my photos. So plain tar it is. 6.8 TB at 100 MB/s would take 19 hours, so no more work today. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-13T02:36:25+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240913-031316</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240913-031316</link><category>technology</category><category>opinion</category><title>NBN outages: yes, no, maybe?</title> <description> <![CDATA[ <div align="justify"> <p> Email from <a href="https://aussiebroadband.com.au/">Aussie Broadband</a> today: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Date: Thu, 12 Sep 2024 12:41:14 +1000<br />From: Aussie Broadband <no-reply@team.aussiebroadband.com.au><br />Subject: nbn Unscheduled Outage<br /><br />NBN has let us know that your service/s may currently be affected by an outage.<br /> </tt></div></blockquote> <p> That's one of these “impossible” messages. How can I receive it if my link is down? Checked, and of course it was still up. But my phone (connected via my <a href="https://www.nbnco.com.au/">NBN</a> link) agreed: </p> <a id="Photo-4" name="Photo-4" href="diary-sep2024.php?dirdate=20240912&imagesizes=11112&topics=c#Photo-4"> <img alt="This should be NBN-outage-2-detail.jpeg. Is it missing?" border="0" id="Photo_4" title="Photo NBN-outage-2-detail.jpeg. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240912/small/NBN-outage-2-detail.jpeg" width="229" height="295" /></a> <p> Had there been an outage? Checked my <a href="http://www.lemis.com/grog/nbnstats.php">NBN stats page</a>. Yes: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Start time End time Duration Badness from to<br /> (seconds)<br /><br />1726103107 1726103311 204 0.001 # 12 September 2024 11:05:07 12 September 2024 11:08:31<br /> </tt></div></blockquote> <p> But that was over and done with over 1½ hours before they sent the message. </p> <p> Oh, sorry about that, try again: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Start time End time Duration Badness from to<br /> (seconds)<br /><br />1726103107 1726103311 204 0.001 # 12 September 2024 11:05:07 12 September 2024 11:08:31<br />1726109505 1726109741 236 0.581 # 12 September 2024 12:51:45 12 September 2024 12:55:41<br /> </tt></div></blockquote> <p> That almost looks as if the second outage happened because of their message. And at 16:12, only 3¼ hours later, they acknowledged that it was over. </p> <p> Done? No, more work needed: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>NBN are doing **emergency network maintenance** between **Fri 13th September 2024 00:00 AEST** and **Fri 13th September 2024 06:00 AEST**, for **180 min**. </tt></div></blockquote> <p> And yes, that's Aussie's inimitable markup, apparently my fault according to their “support” people. I suppose the emergency maintenance makes sense, and at least it's in the middle of the night. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-13T03:13:16+00:00</dc:date></item> <!-- topic aho not selected <p> In the evening <a href="http://www.lemis.com/yvonne/">Yvonne</a> feeds the animals: </p> <p> Those are pellets for the dogs. But today things looked different. What's that thing sticking out of the jug? </p> <p> Well, in this photo, taken after the event. it's a piece of insulation. But at the time it was a mouse tail! Where was the mouse? At the other end, of course, and very much alive. It escaped behind the fridge before any animal could catch it, so we'll have the fun of catching it later. </p> <p> Looking in the food tun, I found two more! It seems that they had chewed a hole in the lid of the tun, fallen in and couldn't get out again. I'm reminded of the story of the Irishman falling into a vat of beer. </p> <p> OK, <a href="http://www.lemis.com/grog/Albums/Animals/Mona.php">Mona</a> showed a lot of interest <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=a&subtitle=A%20mouse?%20%20Rats!&article=D-20240907-025631#D-20240907-025631">last week</a>. Brought her to the tun, where the mice were happily running around. Not interested. <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> was marginally more interested, but not enough. Took the tun outside and let the dogs take a look. Yes, they were interested and chased each of the mice as I let them out. And as far as I can see, they killed them within seconds, but they didn't really know what to do then. I only found one, but I don't think they ate the other one. We still have an uphill battle. </p> End deselected topic aho (Do-it-yourself mousetrap) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240914-020930</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240914-020930</link><category>technology</category><category>opinion</category><title>NBN outage?</title> <description> <![CDATA[ <div align="justify"> <p> So how long did the announced overnight <a href="https://www.nbnco.com.au/">NBN</a> outage last? It didn't. It seems that it didn't happen, and both <a href="https://aussiebroadband.com.au/">Aussie Broadband</a> and NBN web sites deny all knowledge of it. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-14T02:09:30+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240914-020936</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240914-020936</link><category>technology</category><category>opinion</category><title>Disk copy, day 9</title> <description> <![CDATA[ <div align="justify"> <p> Into the office this morning thinking that I hadn't been overly clever by copying an archive to <i>quartet</i>. It had completed, after only about 24 hours, but now I had to extract it, and that could take as long again. </p> <p> But it didn't. It was over in about 10 seconds: truncated archive. Now it was no longer 6.8 TB in size, only 2048 bytes! How did that happen? </p> <p> Oh. In another <i>xterm</i> I had accidentally entered: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/newphotos</font> <font color="blue">274</font> -> <b><tt>tar cvf Photos.tar /Photosiostat 1</tt></b></code><br />tar: /Photosiostat: Cannot stat: No such file or directory<br />tar: 1: Cannot stat: No such file or directory<br />tar: Error exit delayed from previous errors.<br /> </tt></div></blockquote> <p> Damn. Start again. Another day to go. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-14T02:09:36+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240914-021758</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240914-021758</link><category>technology</category><title>Ethernet issues?</title> <description> <![CDATA[ <div align="justify"> <p> Seen in my daily report: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 13 01:32:47 eureka kernel: em0: Watchdog timeout Queue[0]-- resetting<br />Sep 13 01:32:47 eureka kernel: Interface is RUNNING and ACTIVE<br />Sep 13 01:32:47 eureka kernel: em0: TX Queue 0 ------<br />Sep 13 01:32:47 eureka kernel: em0: hw tdh = 785, hw tdt = 903<br />Sep 13 01:32:47 eureka kernel: em0: Tx Queue Status = -2147483648<br />Sep 13 01:32:47 eureka kernel: em0: TX descriptors avail = 903<br />Sep 13 01:32:47 eureka kernel: em0: Tx Descriptors avail failure = 5<br />Sep 13 01:32:47 eureka kernel: em0: RX Queue 0 ------<br />Sep 13 01:32:47 eureka kernel: em0: hw rdh = 330, hw rdt = 329<br />Sep 13 01:32:47 eureka kernel: em0: RX discarded packets = 0<br />Sep 13 01:32:47 eureka kernel: em0: RX Next to Check = 330<br />Sep 13 01:32:47 eureka kernel: em0: RX Next to Refresh = 329<br />Sep 13 01:32:47 eureka kernel: em0: link state changed to DOWN<br />Sep 13 01:32:51 eureka kernel: em0: link state changed to UP<br /> </tt></div></blockquote> <p> What's that? Is it serious? It repeated later in the morning. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-14T02:17:58+00:00</dc:date></item> <!-- topic pHo not selected <p> One of the most significant things about the evolution of cameras is the viewfinder. You want to be able to see what you are going to take, and ensure that it is in focus. Over the years, technology improved things. I've been through this in some detail <a href="http://www.lemis.com/grog/diary-jul2021.php?subtitle=More%20camera%20history&article=D-20210716-023320#D-20210716-023320">a few years ago</a>. TL;DR: </p> <ul> <li class="fullwidth"> <div class="listitemwidth"> Very first cameras: simple viewfinder that matches the view of some lens. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> 1932: Camera with integrated rangefinder (<a href="https://en.wikipedia.org/wiki/Leica_II">Leica II</a>). Requires to be adjusted to the film plane and lens focal length, suffers from parallax at close distances. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> 1934: Single-lens reflex camera (<a href="https://en.wikipedia.org/wiki/Kine_Exakta">Kine Exakta</a>). No more parallax or focal length issues. Still requires alignment of viewfinder screen with film plane. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> 1981: First SLR with autofocus (<a href="https://en.wikipedia.org/wiki/Pentax_ME_F">Pentax ME F</a>). Less danger of focus problems. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> 1995: First digital SLR camera (<a href="https://en.wikipedia.org/wiki/Nikon_E_series">Nikon E</a>). No viewfinder advantage. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> 2008: First interchangeable lens camera with direct sensor data viewfinder (<a href="https://en.wikipedia.org/wiki/Panasonic_Lumix_DMC-G1">Panasonic Lumix DMC-G1</a>). No more issues with focus alignment. </div> </li> </ul> <p> All of these were progress. What's next? I think we're done. More recent mirrorless cameras allow magnifying the viewfinder image for more exact focus, but that's about it. </p> <p> So what do we do now? If you're <a href="https://en.wikipedia.org/wiki/Leica">Leica</a>, start all over again. I've just discovered that they have a range of cameras, most recently the <a href="https://www.dpreview.com/news/2945561017/leica-announces-the-screen-less-m11-d">M11</a>, which can be supplied without all mod cons, only the same kind of rangefinder that they introduced 92 years ago! And that not only at Leica's normal eye-watering prices (US $9,395, $400 <i>more</i> than the standard model, despite it having fewer components). To quote the article, </p> <blockquote class="fullwidth"> <div class="listitemwidth"> The company says the lack of a display will help users focus on "the elementary aspects of pictorial design such as composition, aperture, shutter speed and ISO.” </div> </blockquote> <p> I'm amazed. Of course, they should go the whole hog and offer only a specific sensor sensitivity (ISO). </p> End deselected topic pHo (Leica: back to the past) --> <!-- topic go not selected <p> While walking the dogs, saw this at the Nyary's place: </p> <p> They're removing all the vines! Why? </p> <p> Also saw this mystery flower again, the one that we planted in the south garden and then died. Or at least I think it's the same flower. Here two different bushes in different state of bloom: </p> <p> I'll keep an eye on them over the next few days. </p> End deselected topic go (No more wine in Dereel?) --> <!-- topic Jh not selected <p> Fiona Drayton brought us a new front gate a couple of days ago, and this evening Rod DeGroot and mate showed up to install it. Done! </p> <p> <a href="http://www.lemis.com/yvonne/">Yvonne</a> is happy, but somehow it doesn't quite seem to fit: </p> <p> It also closes too “closed”, in other words further than the post. But it's minor, and Rod put in a bar to limit the travel, so now it closes cleanly. </p> <p> Hopefully it'll stay that way. </p> End deselected topic Jh (A new gate) --> <!-- topic h not selected <p> Another <a href="http://www.lemis.com/grog/powercor/power-failures.php#bottom">grid power failure</a> this morning at 4:04, which for some reason causes <i>something</i> to beep. Damn! How long is it going to last? Turned off the air conditioner and tried to get back to sleep. Another beep a little later. </p> <p> The result? Only one outage, lasting 90 seconds. </p> End deselected topic h (Another power failure) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240915-023629</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240915-023629</link><category>technology</category><category>general</category><category>opinion</category><title>Disk copy, day 10</title> <description> <![CDATA[ <div align="justify"> <p> Into my office in the morning to find that <i>quartet</i> had rebooted as a result of the <a href="http://www.lemis.com/grog/powercor/power-failures.php#bottom">grid power outage</a>. Had it finished the copy? No. Mounted <i>/newphotos</i> (not in <i>/etc/fstab</i>) and discovered, to my horror: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/home/grog</font> <font color="blue">8</font> -> <b><tt>df -i /newphotos</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />/dev/ada1p1 15,257,008 903 15,103,534 0% 3 6,486,139 0% /newphotos<br /> </tt></div></blockquote> <p> Only 3 <a href="https://en.wikipedia.org/wiki/inode">inodes</a> and 903 MB in use. What's that? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/home/grog</font> <font color="blue">10</font> -> <b><tt>l /newphotos|less</tt></b></code><br />total 1279<br />-rwxrwxrwx 1 root wheel 82 3 Apr 2012 ._.TemporaryItems<br />drwxrwx--- 2 root operator 512 14 Jan 2016 .snap<br />-r-------- 1 root wheel 947,322,880 12 Sep 13:59 .sujournal<br />-rwxr--r-- 1 root wheel 754,482 28 Apr 2022 #foo#<br />drwxrwxr-x 2 grog wheel 512 13 Sep 12:07 0-grog<br />drwxr-xr-x 2 grog wheel 1,536 13 Sep 12:04 1-Skylum<br />drwxrwxrwx 2 grog wheel 1,024 13 Sep 12:14 2-grog<br /> </tt></div></blockquote> <p> Well, the 903 MB are clearly the journal. But what about those other files? They don't match the 3 <a href="https://en.wikipedia.org/wiki/inode">inodes</a>. Why did the system allow me to mount a dirty file system? OK, </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/home/grog</font> <font color="blue">11</font> -> <b><tt>fsck -y /dev/ada1p1</tt></b></code><br />** /dev/ada1p1<br />** SU+J Recovering /dev/ada1p1<br /><br />USE JOURNAL? yes<br /><br />** Reading 947322880 byte journal from inode 4.<br /><br />RECOVER? yes<br /><br />** Building recovery table.<br />** Resolving unreferenced inode list.<br />** Processing journal entries.<br /><br />***** FILE SYSTEM IS CLEAN *****<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/home/grog</font> <font color="blue">12</font> -> <b><tt>mount /dev/ada1p1 /newphotos/</tt></b></code><br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/home/grog</font> <font color="blue">13</font> -> <b><tt>df -i /newphotos</tt></b></code><br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />/dev/ada1p1 15,257,008 5,271,466 9,832,971 35% 889,919 5,596,223 14% /newphotos </tt></div></blockquote> <p> That looks better. Continued the copy, but after a while <i>quartet</i> paniced! I didn't find out until it had rebooted, but <i>/var/crash/core.txt.5</i> shows: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>panic: ffs_blkfree_cg: freeing free block<br /> </tt></div></blockquote> <p> That sounds like file system corruption. Journalling not good enough? Run a full <i>fsck</i>, which found many problems. After 45 minutes and 28.827 lines, it asked me if I really wanted to continue. Yes, why not. Finally, after nearly 90 minutes and 70,769 lines of output (roughly 1,250 pages in print), I got what I didn't want to see: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>884600 files, 1342191538 used, 2563602610 free (613818 frags, 320373599 blocks, 0.0% fragmentation)<br /><br />***** FILE SYSTEM MARKED DIRTY *****<br />***** FILE SYSTEM WAS MODIFIED *****<br />***** PLEASE RERUN FSCK *****<br /> </tt></div></blockquote> <p> Clearly there's no recovery for this. In particular, only 884,600 files remained. The rest must have been “recovered”. And I want perfection for a backup disk. So, start all over again with another <i>newfs</i>. </p> <p> While it was running, it occurred to me that I have seen this before. I should have disabled soft updates. But by then it was too late. Let's see what happens by tomorrow. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-15T02:36:29+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240916-014313</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240916-014313</link><category>technology</category><category>opinion</category><title>Disk copy, day 11</title> <description> <![CDATA[ <div align="justify"> <p> Into the office today to find: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>x ./grog/20110827/Components/dam-panorama-CPL-6-1EV.jpeg<br />x ./grog/20110827/Components/dam-panorama-CPL-6.jpegtar: (null)<br />: Truncated tar archive: Unknown error: -1<br />tar: Error exit delayed from previous errors.<br /> </tt></div></blockquote> <p> <i>Another</i> <a href="https://en.wikipedia.org/wiki/Network_File_System">NFS</a> issue! OK, that's enough. Try the rest with <a href="https://en.wikipedia.org/wiki/rsync">rsync</a>, this time with the <b>H</b> option. And how about that, after hardly more than 5 hours: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>grog/www/test/<br />grog/www/test/big/<br />grog/www/Photos/small/white-background.gif => grog/www/Photos/big/white-background.gif<br /><br />sent 1,193,207,971,130 bytes received 18,071,175 bytes 63,800,349.81 bytes/sec<br />total size is 7,886,580,015,513 speedup is 6.61<br />rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1359) [sender=3.3.0]<br /> 18702.08 real 433.88 user 1307.03 sys<br /> </tt></div></blockquote> <p> What's that error? OK, try again: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/1)</font> <font color="red">/Photos</font> <font color="blue">20</font> -> <b><tt>Log rsync -Hav . /newphotos</tt></b></code><br />===== Sun 15 Sep 2024 15:05:02 AEST on quartet.lemis.com: rsync -Hav . /newphotos<br />sending incremental file list<br />./<br />.sujournal<br />Log.log.quartet<br />Log.log.quartet.0<br />rsync: [receiver] rename "/newphotos/.sujournal.3wUiHl" -> ".sujournal": Operation not permitted (1)<br />Log.log.quartet.1<br />Log.log.quartet.2<br />Log.log.quartet.3<br />Log.log.quartet.4<br />Log.log.quartet.5<br />Log.log.quartet.6<br /><br />sent 174,640,554 bytes received 88,974 bytes 146,893.26 bytes/sec<br />total size is 7,886,623,224,434 speedup is 45,136.18<br />rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1359) [sender=3.3.0]<br /> 1188.83 real 4.47 user 28.22 sys<br /><br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/newphotos/grog</font> <font color="blue">65</font> -> <b><tt>date; df -i /Photos/ /newphotos/</tt></b></code><br />Sun 15 Sep 2024 15:28:12 AEST<br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,841,077 712,191 91% 1,764,773 1,478,873 54% /Photos<br />/dev/ada1p1 15,257,008 6,845,610 8,258,827 45% 1,764,814 4,721,328 27% /newphotos<br /> </tt></div></blockquote> <p> That looks better. And yes, the error was the attempt to copy <i>.sujournal</i>, the soft updates journal file. But we still have more space in <i>/newphotos</i> than in <i>/Photos</i>. Not surprising, given that I didn't delete files that weren't on <i>/Photos</i>. OK, a third time: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@quartet (/dev/pts/1)</font> <font color="red">/Photos</font> <font color="blue">21</font> -> <b><tt>Log rsync -Hav --exclude=.sujournal --delete-after . /newphotos</tt></b></code><br />===== Sun 15 Sep 2024 15:28:46 AEST on quartet.lemis.com: rsync -Hav --exclude=.sujournal --delete-after . /newphotos<br />building file list ... done<br />Log.log.quartet<br />deleting 0-grog/Wildflower-2.jpeg<br />deleting 0-grog/Wildflower-1.jpeg<br />...<br />deleting 6-HDR/PhotomatixResults01/e-from-house-0+0EV_+1EV_-1EV.tif<br />deleting grog/20240914/multishot~<br />deleting grog/20240914/magic-pto-now<br /><br />sent 36,804,908 bytes received 34 bytes 52,540.96 bytes/sec<br />total size is 7,886,589,671,198 speedup is 214,280.73<br /><code><font color="blue">=== root@quartet (/dev/pts/0)</font> <font color="red">/var/tmp</font> <font color="blue">79</font> -> <b><tt>date; df -i /Photos/ /newphotos/</tt></b></code><br />Sun 15 Sep 2024 15:41:19 AEST<br />Filesystem 1048576-blocks Used Avail Capacity iused ifree %iused Mounted on<br />eureka:/Photos 7,629,565 6,841,077 712,191 91% 1,764,773 1,478,873 54% /Photos<br />/dev/ada1p1 15,257,008 6,841,899 8,262,538 45% 1,764,815 4,721,327 27% /newphotos<br /> </tt></div></blockquote> <p> That looks better. But for some reason the <tt>iused</tt> field hasn't changed. I suspect that there's an issue with soft updates there. Still, <i>finally</i> I have a valid copy and an incantation to perform the sync. Only 11 days! </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-16T01:43:13+00:00</dc:date></item> <!-- topic ko not selected <p> Discovered this in a kitchen cabinet today: </p> <p> Where did it come from? I didn't buy it, and <a href="http://www.lemis.com/yvonne/">Yvonne</a> denies all knowledge. What it is is clearer: a splatter guard for a frying pan. But what's that plastic foil? It's the remains of the packaging, thoughtfully applied before the handle was attached. To remove it, you need a screwdriver to remove the handle: </p> <p> And then I saw this: </p> <p> If there's one thing that needs a machine wash, it's this kind of device that catches fat in the fabric. What do the instructions on the package say? “Wash before use”. Nothing else at all. I'm amazed. My guess is that the advice is misplaced. I'll put it in the dishwasher, and if it falls apart, I'll throw it out </p> End deselected topic ko (Quality kitchen utensils) --> <!-- topic Pho not selected <p> <a href="https://en.wikipedia.org/wiki/Ayşenur_Ezgi_Eygi">Ayşenur Eygi</a> was buried today. The world is outraged: an innocent civilian was <a href="https://www.washingtonpost.com/world/2024/09/11/american-activist-aysenur-eygi-killed-idf-west-bank/">shot dead by Israeli snipers</a>. </p> <p> And so the world should be. But she was only one person. What about the 1,139 people, mainly Israelis, killed on <a href="https://en.wikipedia.org/wiki/Israel%E2%80%93Hamas_war">3 October 2023</a>? That's several orders of magnitude worse, and the world rightly condemns <a href="https://en.wikipedia.org/wiki/Hamas">Hamas</a> for perpetrating it. But it almost seems that the death of Ayşenur Eygi was worse, a US citizen killed by Israeli snipers. Well, half: she was also Turkish. </p> <p> But then there are the <a href="https://www.aljazeera.com/news/longform/2023/10/9/israel-hamas-war-in-maps-and-charts-live-tracker">currently 41,821 Palestinians</a> that the Israelis have killed since 7 October 2023. The number is rising daily. If the newspapers reported on each of them to the extent that they have reported on Ayşenur Eygi, it would take them centuries to complete. </p> <p> Why this discrepancy? One death is enough. 1,140 are a tragedy. But 42,000! How can <i>any</i> country support such carnage, let alone one claiming to abide by human rights? And that's before you look at the destruction that was once <a href="https://en.wikipedia.org/wiki/Gaza_Strip">Gaza</a>. <a href="https://www.aljazeera.com/news/longform/2023/10/9/israel-hamas-war-in-maps-and-charts-live-tracker">This page</a> summarizes the horrors perpetrated by Israel. Currently it states that in Gaza <i>every hour</i> 42 bombs are dropped, 12 buildings are destroyed, 15 people (6 of them children) are killed and another 35 people are injured. </p> <p> How can a supposedly law-abiding government commit such atrocities? How can the USA support it? By comparison, the Russian aggression in <a href="https://en.wikipedia.org/wiki/Ukraine">Ukraine</a> is almost civilized. The US government is losing international credibility with the support of this carnage. If the death of Ayşenur Eygi can draw people to look at the atrocities, she won't have died in vain. </p> End deselected topic Pho (½ US American? 1139 Israelis? 41,821 Palestinians?) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240917-023903</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240917-023903</link><category>technology</category><category>opinion</category><title>Back to the old slog</title> <description> <![CDATA[ <div align="justify"> <p> So after only 11 days I have finally copied my disk. Time to shut down <i>quartet</i> and revert to the other system configuration issues I have. </p> <p> One relatively minor issue is that the function of the <b>Prev</b> and <b>Next</b> keys (marked <b>PageUp</b> and <b>PageDown</b>) on <i>xterm</i> has changed: it should, as the inscriptions suggest, page up and down, but now it pages through the shell history. </p> <p> Why? It must have something to do with the <a href="https://www.x.org/">X</a> resources. I've been playing with them recently, adding additional key bindings. Initially they were: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>*VT100.Translations: #override \<br /> <Key>Next: scroll-forw(1,page) \n\<br /> <Key>Prior: scroll-back(1,page) </tt></div></blockquote> <p> But in the course of my configuration attempts, I changed them to: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>xterm*VT100.Translations: #override \<br /> <Key>Next: scroll-forw(1,page) \n\<br /> <Key>Prior: scroll-back(1,page) \n\<br /> Shift <Key>Insert: insert-selection(SELECT) \n\<br /> Ctrl Shift <Key>V: insert-selection(SELECT) \n\<br /> Ctrl Shift <Key>C: copy-selection(SELECT)<br /> </tt></div></blockquote> <p> That looks straightforward enough to me. Adding <tt>xterm</tt> to the resource shouldn't make any difference. Tried reverting. No difference. </p> <p> OK, <a href="https://en.wikipedia.org/wiki/RTFM">RTFM</a> time. And <i>once again</i> I couldn't find anything that helps. What does <tt>#override</tt> mean? I haven't found any reference to it, though <i>xrdb</i> has an <tt>-override</tt> option: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><b> -override</b><br /> This option indicates that the input should be added to,<br /> instead of replacing, the current contents of the specified<br /> properties. New entries override previous entries.<br /> </tt></div></blockquote> <p> But that doesn't seem to work here. Much searching, no results. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-17T02:39:03+00:00</dc:date></item> <!-- topic ah not selected <p> It's been quite cool in the few weeks since the beginning of spring, a minimum temperature of 1.8°. And the <a href="https://en.wikipedia.org/wiki/Bureau_of_Meteorology">Bureau of Meteorology</a> has been warning of dead lambs. </p> <p> And indeed, that seems to be the case: </p> <p> Those white spots in the foreground are dead lambs. There were more a couple of days ago, so it seems that they're dying daily. I wouldn't have expected that, especially since the weather wasn't <i>that</i> cold. </p> End deselected topic ah (Oh the springtime it brings on the deadlambs) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240917-023958</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240917-023958</link><category>general</category><category>technology</category><category>opinion</category><title>Banking: the pain</title> <description> <![CDATA[ <div align="justify"> <p> Last Wednesday <a href="http://www.lemis.com/yvonne/">Yvonne</a> had problems with the <a href="https://en.wikipedia.org/wiki/Bank_of_Melbourne_(2011)">Bank of Melbourne</a>: her debit card expired in February, and she had to go to the branch in <a href="https://en.wikipedia.org/wiki/Ballarat">Ballarat</a> to get money. It took her nearly an hour, and the person who dealt with it—clearly an immigrant—spoke English so badly that she could barely understand him. And this at the “we speak your language” Bank of Melbourne! </p> <p> So why did she not get a new card? After over 30 frustrating minutes on the phone, we established: </p> <ul> <li class="fullwidth"> <div class="listitemwidth"> Yvonne never received a replacement card. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> The Bank cancelled Yvonne's card (but not mine) because it hadn't been activated. They didn't inform us of this action. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> To reinstate the card, they needed information from Yvonne like the number and nature of the accounts that we have with the bank. No outside help allowed, so I (standing next to her) couldn't help. But she didn't even know what accounts we had! </div> </li> <li class="fullwidth"> <div class="listitemwidth"> They wanted her to identify herself with things like bank passwords. Again, no help from me allowed. Since she never used online banking, she couldn't know. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> Finally she “identified” herself with date of birth, driver license number and address, along with the last transaction (at the branch in Ballarat). With the exception of the last, which even I didn't know, there's nothing difficult to find here. Any well-prepared criminal could have found them. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> They can't send us the card. It has to be picked up at the branch! At least it seems that there could be a possibility of making an appointment to avoid the long waits. </div> </li> </ul> <p> Why is this all so complicated? Yes, establishing identity is good. But at the very least Yvonne didn't receive her replacement card, and they cancelled it without warning. There's no excuse for that. By contrast, my <a href="http://www.lemis.com/grog/diary-dec2022.php?subtitle=ANZ%20card:%20done?&article=D-20221221-001234#D-20221221-001234">last similar interaction</a> with <a href="https://en.wikipedia.org/wiki/ANZ_(bank)">ANZ</a> went more smoothly, though I had my issues there too. </p> <p> Change bank? Yes, it sounds like a good idea, but only if I can be sure that the new one is better. After my experience with <a href="https://en.wikipedia.org/wiki/Bankwest">Bankwest</a> <a href="http://www.lemis.com/grog/diary-oct2021.php?subtitle=Reporting%20Centrelink%20security%20breach&article=D-20211013-015651#D-20211013-015651">3 years ago</a>, I'm particularly wary. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-17T02:39:58+00:00</dc:date></item> <!-- topic ko not selected <p> I have a surprising number of different kinds of <a href="http://www.lemis.com/grog/recipes/noodle-cooking-times.php">noodles</a>, mainly of Chinese origin. Time to get rid of some of the older ones. Today I had <a href="http://www.lemis.com/grog/recipes/mi-udang-tean.php"><i>mi udang</i></a> for breakfast, requiring noodles like these: </p> <p> What's the difference? Very little in taste. They're intended as single servings: boil in water for about 2 minutes and they're done. But the first (<a href="http://www.lemis.com/grog/recipes/noodle-cooking-times.php#Sau-Tao">Sau Tao</a>) only weigh 45 g per portion, giving about 100 g of cooked noodles—not enough. They're also more difficult to separate than the other one (<a href="http://www.lemis.com/grog/recipes/noodle-cooking-times.php#Chefs-world">“Chef's world”</a>), which give a better 150 g of finished noodle. </p> <p> So: I had 5 baskets of Sau Tao, in this case only 207 g. Cook them for 472 g of finished noodles. That's three portions and the last of the Sau Tao. </p> End deselected topic ko (Sau Tao: too much trouble) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240918-013547</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240918-013547</link><category>technology</category><category>opinion</category><title>X breakthrough</title> <description> <![CDATA[ <div align="justify"> <p> Why don't my changes to <i>.Xdefaults</i> make any difference? On a whim, started a new <i>xterm</i>. It works! So the whole issue was that <i>xterm</i> doesn't pay any attention to the <i>.Xdefaults</i> after it has been started. That makes sense, but it took me a while to get to it. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-18T01:35:47+00:00</dc:date></item> <!-- topic g not selected <p> Paul Donaghy along today to mow the lawn, the first time this season. For once nothing seems to have gone wrong, though he thinks that the drive belt for the lawn mower is slipping. </p> End deselected topic g (First lawn mowing of spring) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240918-014453</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240918-014453</link><category>technology</category><category>opinion</category><title>Back to the makefs bug</title> <description> <![CDATA[ <div align="justify"> <p> It's been <a href="http://www.lemis.com/grog/diary-sep2023.php?subtitle=Debugging%20again&article=D-20230921-023558#D-20230921-023558">nearly a year</a> since I started investigating a <a href="https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=273725">bug</a> in <i>makefs(8)</i>, or at least in the <a href="https://www.freebsd.org/">FreeBSD</a> version. It comes from <a href="https://en.wikipedia.org/wiki/NetBSD">NetBSD</a>, where it works. </p> <p> High time to get back to it. But it seems that <i>kimchi</i>, my NetBSD box, no longer has a debug version of <i>makefs</i>, and to build it I need to build a kernel. And for some reason, the build fails. </p> <p> Dammit, any old (well, newer) version of NetBSD will do. But first I have to download things. And that took the rest of the day. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-18T01:44:53+00:00</dc:date></item> <!-- topic g not selected <p> Our <a href="https://en.wikipedia.org/wiki/Curry_tree">Curry tree</a>, now 14 years old, was not happy about a lapse in watering a few weeks back, combined with the annual attack of mites: </p> <p> But new growth is coming, and it seems that it might even flower: </p> <p> It has only done that <a href="http://www.lemis.com/grog/diary-jan2024.php?subtitle=Some%20garden%20improvement&article=D-20240113-005030#D-20240113-005030">once before</a>, and I had attributed that to the lack of mites: </p> <p> But that was only 8 months ago, and maybe it is just mature enough. </p> End deselected topic g (Curry tree flowering?) --> <!-- topic po not selected <p> Taking photos of the <a href="https://en.wikipedia.org/wiki/Curry_tree">curry tree</a> wasn't easy. Clearly I needed flash, so I chose the mecablitz 15 MS-1, which also needs a trigger flash. This was on the <a href="https://en.wikipedia.org/wiki/Olympus_OM-D_E-M1_Mark_II">Olympus OM-D E-M1 Mark II</a>, so the obvious choice was the toy FL-LM3 flash that came with the camera. But I couldn't get it to fire! It seems that it has died after being used about once a year since I bought it 7 years ago! </p> <p> Not surprisingly, the photos were ridiculously underexposed: </p> <p> More surprisingly, my photo processing software almost recovered it. This is the same photo after processing: </p> <p> But even after replacing it with the <a href="http://www.lemis.com/grog/photography/Mecablitz-58-AF-1.php">mecablitz 58 AF-2</a>, the results were ridiculously underexposed. Once again, as I have been doing for 60 years, I wonder what use on-camera flash is. </p> <p> So what do I do now? More and more I appreciate the studio flash units that I have spread around the house. In this case, though, I think focus stacking with natural light will do the trick. Mañana. </p> End deselected topic po (Bloody flash!) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240919-025101</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240919-025101</link><category>technology</category><title>Exploit?</title> <description> <![CDATA[ <div align="justify"> <p> Into the office this morning to see lots of messages streaming off the log screen. The first that hit me were these: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 18 08:27:20 eureka Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from tiwi: postfix/smtpd[39164]: connect from localhost[127.0.0.1]<br />Sep 18 08:27:20 eureka Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from tiwi: postfix/smtpd[39164]: NOQUEUE: reject: MAIL from localhost[127.0.0.1]: 452 4.3.4 Message size exceeds fixed limit; from=<> proto=ESMTP helo=<tiwi.lemis.com><br />Sep 18 07:45:54 eureka Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: postfix/local[11450]: C04D126359D: to=<grog@lemis.com>, relay=local, delay=0.2, delays=0.19/0/0/0.01, dsn=2.0.0, status=sent (delivered to command: exec /usr/local/bin/procmail -t 2>>/home/grog/Mail/procmailerr || exit 75)<br />Sep 18 09:05:40 eureka Forwarded from 192.109.197.137: postfix/cleanup[25410]: EB30626359D: message-id=<172661433578.1482.3761641752870526039@worldpartners24.com><br />Sep 18 09:05:40 eureka Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: postfix/smtpd[25400]: disconnect from www.lemis.com[45.32.70.18] ehlo=1 mail=1 rcpt=1 data=1 quit=1 commands=5<br />Sep 18 08:55:25 eureka Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: postfix/local[18921]: 9124426359D: to=<grog@lemis.com>, relay=local, delay=1.4, delays=1.4/0/0/0.01, dsn=2.0.0, status=sent (delivered to command: exec /usr/local/bin/procmail -t 2>>/home/grog/Mail/procmailerr || exit 75)<br /> </tt></div></blockquote> <p> “Message size exceeds fixed limit”. From <i>tiwi</i>? What's going on there. Shut down <i>eureka</i>'s mail system while I'm investigating. </p> <p> But then there were other messages: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 18 09:19:11 192.109.197.137 Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: kernel: Limiting icmp unreach response from 203 to 200 packets/sec<br />Sep 18 09:15:06 192.109.197.137 Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: kernel: Limiting icmp unreach response from 215 to 200 packets/sec<br />...<br />Sep 18 09:12:46 192.109.197.137 Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.137: Forwarded from 192.109.197.1<br /> </tt></div></blockquote> <p> What's this nonsense? <i>named</i> was locked at 100% CPU, so I stopped that too. And still they kept coming, so I stopped <i>syslogd</i>. Gradually I was able to investigate things, which included messages like: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Sep 18 06:56:32 eureka Forwarded from homephone: HT802 [c0:74:ad:37:66:d8] [1.0.21.4]CallRecord::writeCDRFile, No space! current file size =51158bytes, need extra 91 bytes.<br />Sep 18 07:54:58 eureka Forwarded from homephone: HT802 [c0:74:ad:37:66:d8] [1.0.21.4]CallRecord::writeCDRFile, No space! current file size =51158bytes, need extra 92 bytes.<br />Sep 18 07:40:08 eureka Forwarded from homephone: HT802 [c0:74:ad:37:66:d8] [1.0.21.4]CallRecord::writeCDRFile, No space! current file size =51158bytes, need extra 91 byte </tt></div></blockquote> <p> They're from the <a href="http://www.grandstream.com/products/gateways-and-atas/analog-telephone-adaptors/product/ht802">Grandstream HT802</a> <a href="https://en.wikipedia.org/wiki/VoIP">VoIP</a> <a href="https://en.wikipedia.org/wiki/Analogue_telephone_adapter">ATA</a>. And I had turned off all logging! Where do they come from? Why are they out of chronological order? About the only thing that isn't a surprise is the content of the message: that's indicative of the terminally buggy firmware of the HT802, as I have <a href="http://www.lemis.com/grog/diary-aug2023.php?subtitle=HT802%20noise&article=D-20230825-024731#D-20230825-024731">already discovered</a>. </p> <p> So what is it? The other messages don't make much sense either. Just these “Forwarded from” prefixes. <i>syslog</i> loop? No, it was only ever the same IP address, <tt>192.109.197.137</tt>, which is <i>eureka</i>. I gradually restarted everything, and things worked normally. No idea when it started: the flood of messages had overwritten all the older copies of <i>/var/log/messages</i> and <i>/var/log/maillog.</i> </p> <p> I'm baffled. I should investigate the <tt>Forwarded from</tt> prefix. <i>syslogd</i> sources? </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-19T02:51:01+00:00</dc:date></item> <!-- topic Po not selected <p> <a href="https://www.reuters.com/world/middle-east/dozens-hezbollah-members-wounded-lebanon-when-pagers-exploded-sources-witnesses-2024-09-17/">News of the day</a>: a large number of <a href="https://en.wikipedia.org/wiki/Pager">pagers</a> used by <a href="https://en.wikipedia.org/wiki/Hezbollah">Hezbollah</a>, a <a href="https://en.wikipedia.org/wiki/Lebanon">Lebanese</a> political party, exploded at 15:30 local time, killing round 10 people and injuring thousands. </p> <p> How did that happen? People are still trying to put the pieces together, not only of the hundreds who lost hands or eyes. But it seems accepted (not a good word) that <a href="https://en.wikipedia.org/wiki/Israel">Israel</a> is behind it. </p> <p> That's terrorism! From the <a href="https://en.wikipedia.org/wiki/Terrorism">Wikipedia page</a>: </p> <blockquote class="fullwidth"> <div class="listitemwidth"> The term is used in this regard primarily to refer to intentional violence during peacetime or in the context of war against non-combatants (mostly civilians and neutral military personnel). Different definitions of terrorism emphasize its randomness, its aim to instill fear, and its broader impact beyond its immediate victims. </div> </blockquote> <p> I've been disgusted by Israel's actions for at least a year now. Somehow there's no room left for an increase in disgust. But when will people do something about the US-backed Israeli terrorists? </p> End deselected topic Po (US-backed Israeli terrorism?) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240919-032427</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240919-032427</link><category>technology</category><category>opinion</category><title>More NetBSD pain</title> <description> <![CDATA[ <div align="justify"> <p> My <a href="https://en.wikipedia.org/wiki/NetBSD">NetBSD</a> sources are in place, so I followed the <a href="https://www.netbsd.org/docs/guide/en/chap-fetch.html">build instructions</a>. Oh. For some reason, they're for cross-builds. And the old <tt>make build</tt> seems no longer to be the way to go. OK, move on to the next chapter, <a href="https://www.netbsd.org/docs/guide/en/chap-kernel.html">compiling the kernel</a>. And it failed! A typical situation is: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>kimchi# <b>config GENERIC</b><br />Build directory is ../compile/GENERIC<br />Don't forget to run "make depend"<br />kimchi# <b>make depend</b><br />make: don't know how to make depend. Stop<br /> </tt></div></blockquote> <p> Clearly it wants that done in a different directory, but which? <i>/usr/src</i>? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>kimchi# <b>make depend</b><br /><i>... much output</i><br /># create lib/pcap-common.d<br />CC=/tooldir.NetBSD-10.0-amd64/bin/x86_64--netbsd-gcc /tooldir.NetBSD-10.0-amd64/bin/nbmkdep -f pcap-common.d.tmp -- -std=gnu99 -Wall -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-sign-compare -Wsystem-headers -Wno-traditional -Wa,--fatal-warnings -Wreturn-type -Wswitch -Wshadow -Wcast-qual -Wwrite-strings -Wextra -Wno-unused-parameter -Wno-sign-compare -Wsign-compare -Wformat=2 -Wno-format-zero-length -Werror -DLBL_ALIGN --sysroot=/ -DPCAP_DONT_INCLUDE_PCAP_BPF_H -DHAVE_CONFIG_H -DENABLE_REMOTE -DPCAP_SUPPORT_RPCAP -I/usr/src/external/bsd/libpcap/lib -I/usr/src/external/bsd/libpcap/lib/../include -I/usr/src/external/bsd/libpcap/lib -I/usr/src/external/bsd/libpcap/lib/../dist -DYYBISON -DHAVE_CONFIG_H -DINET6 /usr/src/external/bsd/libpcap/lib/../dist/pcap-common.c && mv -f pcap-common.d.tmp pcap-common.d<br />/usr/src/external/bsd/libpcap/lib/../dist/pcap-common.c:1131:2: error: #error The LINKTYPE_ matching range does not match the DLT_ matching range<br /> 1131 | #error The LINKTYPE_ matching range does not match the DLT_ matching range<br /> | ^~~~~<br />nbmkdep: compile failed.<br />*** Error code 1<br /> </tt></div></blockquote> <p> Never mind, I didn't want to build a kernel. How about the next chapter, <a href="https://www.netbsd.org/docs/guide/en/chap-updating.html">Updating an existing system from sources</a>? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>kimchi# <b>./build.sh -O ../obj -T ../tools -U distribution</b><br /><i>...much output</i><br /># install /usr/src/../tools/include/compat/nbtool_config.h<br />/usr/src/../obj/tools/binstall/xinstall -c -r nbtool_config.h /usr/src/../tools/include/compat/nbtool_config.h<br />cd: can't cd to include<br /><br />*** Failed target: includes<br />*** Failed command: (cd include && find . -name '*.h' -print | while read f ; do /usr/src/../obj/tools/binstall/xinstall -c -r $f /usr/src/../tools/include/compat/$f ; done)<br />*** Error code 2<br /> </tt></div></blockquote> <p> This is ridiculous. I've never had problems building NetBSD before. Yes, I can try to debug the thing, but this is a standard install following the instructions. What has gone wrong? </p> <p> I have another image, probably the one I used last year. If that's the case, I already have a debug version of <i>makefs</i>, so I don't need to do anything. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-19T03:24:27+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240920-021318</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240920-021318</link><category>technology</category><category>history</category><category>opinion</category><title>IBM 3330 space requirements</title> <description> <![CDATA[ <div align="justify"> <p> For years I've been comparing the capacity and size of my disk drives with those of the <a href="https://en.wikipedia.org/wiki/IBM_3330">IBM 3330</a>, <i>the</i> big disk drive of the 1970s. Here an image from the <a href="https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/MagneticDataStorage/MagneticDisks.php">University of Auckland</a>: </p> <a href="https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/MagneticDataStorage/DataStorageImages/DiskPacks/IBM3330.jpg"> <img alt="" title=". Click to see in original size." src="https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/MagneticDataStorage/DataStorageImages/DiskPacks/IBM3330.jpg" width="400" /></a> <p> It had 8 or 9 drives, though I've only ever seen 8, as in that photo. Each drive initially held 100 MB, but by the time I came on the scene they had doubled that to 200 MB, so an 8 drive configuration held 1.6 GB, coincidentally 0.01% of the capacity of my newest drive. </p> <p> My most recent disk drive has a capacity of 16 TB, roughly 10,000 times the size of the 3330. But how much physical space does it take up? My current drive is a pretty standard 5¼" drive. And the 3330? <a href="https://gemini.google.com/app/12f6ded0cb11427e">Google Gemini</a> fails badly and gives the information: </p> <table summary="Automatically generated table"> <tr> <td valign="top" align="left">Model</td> <td> </td> <td valign="top" align="center">Width</td> <td> </td> <td valign="top" align="center">Depth</td> <td> </td> <td valign="top" align="center">Height</td> <td> </td> <td valign="top" align="center">Floor space</td> </tr> <tr> <td valign="top" align="left">3330-1</td> <td> </td> <td valign="top" align="center">48 inches (122 cm)</td> <td> </td> <td valign="top" align="center">36 inches (91 cm)</td> <td> </td> <td valign="top" align="center">72 inches (183 cm)</td> <td> </td> <td valign="top" align="center">16.5 square feet (1.53 square meters)</td> </tr> <tr> <td valign="top" align="left"></td> </tr> </table> <p> It also includes identical information for the 3330-2, 3330-3 and 3330-4. </p> <p> So what are those measurements? The disk packs were 14" in diameter. Can each bay really be 48" wide? That seems like a lot, more than 3 times the diameter of the pack. But the pack covers on top of the unit are about 15" in diameter, so it seems plausible. Certainly the height appears correct. </p> <p> So: let's guess. Each bay is 1.22 m wide, with the exception of the controller at the right. There seem to have been many controllers with different widths, but for the sake of convenience I'll assume the same width as a disk bay. That makes a total of 6.1 m and an area of 7.65 m². But that's just the footprint. To change packs, you need to open a drive drawer. Assuming that there are several rows, how much space does each unit need? I'd guess an absolute minimum of 60 cm, making for a total footprint of 6.1 × 1.82 m or 11.1 m², or about 144 MB/m². It's been a long time since data took up that much area. </p> <p> So at that density, you'd need 16,000 ÷ 0.144 or 11.11 ha of area. That's considerably more than <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=c&subtitle=Installing%20the%20new%20disk&article=D-20240906-024555#D-20240906-024555">my last estimate</a>, but still smaller than the size of the <a href="https://en.wikipedia.org/wiki/Vatican_City">Vatican City</a>. For that you'd need the equivalent of nearly 5 16 TB disks. </p> <p> And transfer speed? Hard to say. I haven't measured the speed of my new disk, but let's assume round 400 MB/s. The 3330 arrays would have a total of 80,000 disk packs, all capable that would be a transfer rate of round 40 GB/s, more than anything currently available. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-20T02:13:18+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240921-021403</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240921-021403</link><category>technology</category><category>multimedia</category><category>opinion</category><title>Secure confirmation</title> <description> <![CDATA[ <div align="justify"> <p> For some reason I wasn't signed up for the <a href="https://en.wikipedia.org/wiki/ARD_(broadcaster)">ARD</a> mailing list. OK, fight my way through the broken web site and sign up. </p> <p> Obligatory confirmation message: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Date: Fri, 20 Sep 2024 03:03:58 +0200<br />From: ARD Mediathek Newsletter <newsletter@ard.de><br />Subject: Bitte bestätigen Sie Ihre Anmeldung<br /><br />Bitte klicken Sie auf den folgenden Link, um Ihre Anmeldung zu bestätigen:<br /><br /> https://seu2.cleverreach.com/f/136025-139443/wss/1220015-cb7511128984<br /> </tt></div></blockquote> <p> “Please click on this link to confirm your registration”. <a href="https://www.cleverreach.com">https://www.cleverreach.com</a>? What's that? </p> <p> Yes, on some reflection there's little that can go wrong. But it requires reflection, and it encourages sloppiness. <i>Bad</i> ARD. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-21T02:14:03+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240921-021724</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240921-021724</link><category>technology</category><category>opinion</category><title>More makefs debugging?</title> <description> <![CDATA[ <div align="justify"> <p> So why can't I build a <a href="https://en.wikipedia.org/wiki/NetBSD">NetBSD</a> system any more? Do I really need to? No, I still have the disk image that I was using <a href="http://www.lemis.com/grog/diary-sep2023.php?subtitle=More%20makefs%20insights&article=D-20230923-011158#D-20230923-011158">last time</a>. OK, replace the “disk” in <i>kimchi</i> with that image and reboot. </p> <p> No networking! I've seen that <a href="http://www.lemis.com/grog/diary-sep2023.php?subtitle=More%20VirtualBox%20fun&article=D-20230923-011058#D-20230923-011058">before</a> too, though I've never been able to work out why. Create a new VM using that disk, and all is well. </p> <p> Except that I don't have a debug version of <i>makefs</i> there, only some output: <a href="http://www.lemis.com/grog/diary-sep2023.php?subtitle=Looking%20at%20makefs&article=D-20230922-023940#D-20230922-023940">as I discovered later</a>, <i>makefs</i> contains debug options. OK, time for a new <a href="https://www.freebsd.org/">FreeBSD</a> machine. Install 14.1 and upgrade to <tt>-CURRENT</tt>. And that almost worked: it seems that the <i>/usr/obj</i> has blown out to 18 GB! So it filled up the virtual disk, and I had to put it via <a href="https://en.wikipedia.org/wiki/Symbolic_link">symlink</a> on <i>/hydra</i>. But then things ran without trouble, only not fast enough for me to do anything more today. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-21T02:17:24+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240922-014640</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240922-014640</link><category>general</category><category>technology</category><category>opinion</category><title>Spring rains arrive</title> <description> <![CDATA[ <div align="justify"> <p> Heavy rain overnight, causing me to decide to turn off the heating in case we had a power failure. That didn't happen, but we got round 22 mm of rain, or, as the <a href="https://en.wikipedia.org/wiki/Bureau_of_Meteorology">Bureau of Meteorology</a> put it, 2.2 mm: </p> <a id="Photo-5" name="Photo-5" href="diary-sep2024.php?dirdate=20240921&imagesizes=111112&topics=c#Photo-5"> <img alt="This should be Weather-report-2.jpeg. Is it missing?" border="0" id="Photo_5" title="Photo Weather-report-2.jpeg. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240921/small/Weather-report-2.jpeg" width="174" height="387" /></a> <p> It's easy to understand inaccuracies in weather forecasts, especially for areas far from weather stations. But this app shows quite accurate rainfall maps, so it's hard to understand why they should be so completely inaccurate in recording rainfall. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-22T01:46:40+00:00</dc:date></item> <!-- topic go not selected <p> Jesse Walsh along today to do some weeding. Somehow it's not working. After an hour he had removed about 95% of the weeds from an area of about 8 m². In two weeks it will look the same as it did yesterday. </p> <p> How do professional gardeners address the issue? </p> End deselected topic go (More weeding) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240922-020349</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240922-020349</link><category>technology</category><category>general</category><category>opinion</category><title>Measure your IQ</title> <description> <![CDATA[ <div align="justify"> <p> Today I saw an advertisement from a company called <a href="https://www.cerebrumiq.com/">Cerebrum</a> offering to tell me my type of intelligence: spatial, musical, mumble. Over the years I've noticed that different people address problems in different manners, and I've decided for myself that I'm very much spatial. But I've had very intelligent people at work who couldn't find their way round the building. Clearly not a spatially defined intelligence. </p> <p> So this test looked interesting. Disappointing: it was all a matter of pattern matching, and some of the questions were quite difficult. Finally I finished. Show me the results? No, they wanted to send them to me. OK, another throwaway email address. And after a while they sent me a message telling me all about IQ, something that I didn't want to know. No mention of kind of intelligence. </p> <p> OK, click on <b>Get my IQ Score Now!</b>. Error 404! </p> <p> OK, contact. They have a bot. OK, tell it: “Your web link is 404”. </p> <p> Hang. After 20 minutes, I asked “Are you still there”. Yes, “Hi, I'm the Cerebrum bot. What do you need help with?”. OK, maybe an easier-to-parse response: “You have sent me an email with a broken link”. Yes, this time I got an answer: “Hi, I'm the Cerebrum bot. What do you need help with?”. </p> <p> Things continued like that, with alternate hangs, repeats and nonsense. In the end I gave up, but first I noticed the information at the bottom of the page: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>© Cerebrium IQ - 2024. All rights reserved<br />30 N Gould St Ste R Sheridan, WY 82801 USA<br /> </tt></div></blockquote> <p> They can't even spell their own name! Somehow they seem completely unintelligent. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-22T02:03:49+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240922-025435</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240922-025435</link><category>technology</category><category>opinion</category><title>More on software maintenance</title> <description> <![CDATA[ <div align="justify"> <p> <a href="http://www.lemis.com/grog/diary-feb2024.php?subtitle=Making%20software%20maintenance%20easy&article=D-20240209-013125#D-20240209-013125">Earlier this year</a> I discovered a bug in the man page for <i>make(1)</i>. <i>make</i> first looks for a file <i>BSDmakefile</i>, but it's not mentioned anywhere in the man page. In fact, it's not mentioned anywhere in the sources. Where does it come from? After some discussion on IRC, discovered that <i>make</i> reads a file <i>/usr/share/mk/sys.mk</i>, and that contains: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt># Tell bmake the makefile preference<br />.MAKE.MAKEFILE_PREFERENCE= BSDmakefile makefile Makefile<br /> </tt></div></blockquote> <p> OK, that's specific to <a href="https://www.freebsd.org/">FreeBSD</a>. <a href="https://en.wikipedia.org/wiki/NetBSD">NetBSD</a> doesn't know a <i>BSDmakefile</i>. How do we handle it? Started a mail discussion with Simon J. Gerraty, who considered the possibility of adding operating system dependent text. But I wonder if it wouldn't make more sense to remove <i>BSDmakefile</i> altogether. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-22T02:54:35+00:00</dc:date></item> <!-- topic ko not selected <p> <a href="http://www.lemis.com/grog/recipes/roast-beef.php">Roast beef</a> for dinner today. That's nothing new, but somehow the details evade me. Cook at 180° fan-forced to 53°, taking about 42 minutes per kilogram. That's the easy part. But this roast had a layer of fat on top. That calls for heat from above, also suggesting that the beef should be higher in the oven: </p> <p> Started like that. After 20 minutes, the inside temperature was still showing the initial 18°. Something wrong with the (wireless) thermometer? </p> <p> OK, try a second thermometer, inserted not quite as deeply as the first. And sure enough, it rose more quickly. After the expected 42 minutes the temperature was 53°, just what I had expected. But the first thermometer was only showing round 40°. After 20 minutes' rest they were closer, round 50° and 60°. </p> <p> That's understandable, but it brings back to me how difficult it is to specify a meat temperature. I don't know how to address the problem. </p> <p> And then <a href="http://www.lemis.com/grog/recipes/yorkshire-pudding.php">Yorkshire pudding</a>. Not a success: it takes much longer to cook than I expected, and after 20 minutes at 210° it was still undercooked. More investigation required. </p> End deselected topic ko (Roast beef again) --> <!-- topic hko not selected <p> Vanessa Milamber along for breakfast today. She's travelling Australia (and the world) and is currently staying with Chris Bahlo. </p> End deselected topic hko (Breakfast with Vanessa) --> <!-- topic go not selected <p> It's the <a href="https://en.wikipedia.org/wiki/September_equinox">September equinox</a>, time for the monthly <a href="http://www.lemis.com/grog/Albums/garden/Flowers-20240922.php">garden flower photos</a>. </p> <p> It has been one of the mildest winters since I started keeping records: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>mysql> <b>select year(date),<br /> min(outside_temp),<br /> avg(outside_temp),<br /> max(outside_temp)<br /> from observations<br /> where month(date) > 5<br /> and month(date) < 9<br /> group by year(date);</b><br />+------------+-------------------+-------------------+-------------------+<br />| year(date) | min(outside_temp) | avg(outside_temp) | max(outside_temp) |<br />+------------+-------------------+-------------------+-------------------+<br />| 2017 | -2.2 | 8.560618937140722 | 19.7 |<br />| 2018 | -3 | 8.791814473911545 | 19.7 |<br />| 2019 | -0.9 | 9.059936482458161 | 20.4 |<br />| 2020 | -0.8 | 8.716560253881685 | 20.7 |<br />| 2021 | -0.6 | 9.301671782613512 | 23 |<br />| 2022 | -2.6 | 9.085142510706916 | 20.4 |<br />| 2023 | -1.4 | 9.892861121292666 | 23.6 |<br />| 2024 | -2.8 | 9.608775174628834 | 24.4 |<br />+------------+-------------------+-------------------+-------------------+<br /> </tt></div></blockquote> <p> But somehow the garden looks less developed than in previous years. There's almost nothing worth highlighting. About the only thing of interest is this <a href="https://en.wikipedia.org/wiki/Abutilon">Abutilon</a> outside my bedroom window, which has had its ups and downs: </p> <p> The parent tree of our blue-flowering wildflower is flowering nicely: </p> <p> But the best we can say about the one that we transplanted is that it's not dead yet: </p> <p> It looked much worse a few months back, so there's still (a little) hope. </p> <p> The rest aren't dead yet, but they're taking their time coming. Here they are: </p> End deselected topic go (Garden flowers in early spring) --> <!-- topic ko not selected <p> Grilled chicken legs for dinner tonight, something that we haven't done very much recently. In particular, I didn't have much for the “hair dryer” <a href="http://www.lemis.com/grog/recipes/air-fryer.php">“air fryer”</a>. Tried them for 30 minutes at 200°, but they were ready after 25 minutes: </p> <p> That was a relatively painless experiment. </p> End deselected topic ko (“Air fried” chicken legs) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240924-014113</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240924-014113</link><category>technology</category><category>opinion</category><title>DVD disaster</title> <description> <![CDATA[ <div align="justify"> <p> When I finalized the specifications for <i>hydra</i> <a href="http://www.lemis.com/grog/diary-sep2023.php?topics=c&subtitle=New%20computer&article=D-20230922-014301#D-20230922-014301">almost exactly a year ago</a>, one of the requirements was a <a href="https://en.wikipedia.org/wiki/DVD">DVD</a> drive. That seemed like a no-brainer, but it later occurred to me that DVDs and <a href="https://en.wikipedia.org/wiki/Compact_disk">CDs</a> are gradually becoming a thing of the past, and that for the few occasions that I might need one, I could use a <a href="https://en.wikipedia.org/wiki/USB">USB</a> drive. As it was, the choice of an inbuilt drive limited my choice of case. </p> <p> Today I had some data CDs that I had been asked to copy. OK, put them in the drive. It's behind a door. How do you open the thing? While trying to do so, saw a red light go on on top of the cabinet. The machine shut down! </p> <p> This <i>stupid</i> idea of putting the controls on the top of the case! </p> <a id="Photo-6" name="Photo-6" href="diary-sep2024.php?dirdate=20231018&imagesizes=1111112&topics=c#Photo-6"> <img alt="This should be hydra-18.jpeg. Is it missing?" border="0" id="Photo_6" title="Photo hydra-18.jpeg. Click to redisplay larger version of image." src="https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20231018/small/hydra-18.jpeg" width="369" height="183" /></a> <p> The one on the left is the power button, and it's only that far from the front of the case (extreme left). I must have pressed it while trying to open the door, which (I discovered) opens in the opposite direction. Why do people do this? It must be (relatively) modern, since Bruce Evans' <i>zetaplex</i> also has it. I need to put a plastic guard on top of it. </p> <p> The machine came up relatively smoothly, as was to be expected with a clean shutdown. But: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>mount: /dev/da0p1: R/W mount on /Video denied. Filessystem is not clean - run fsck.: Operation not permitted </tt></div></blockquote> <p> Huh? What went wrong there? OK, mark it <tt>noauto</tt> in <i>/etc/fstab</i> and continue. Run <i>fsck</i>, which went for hours. Didn't I have journalling enabled? When <i>fsck</i> finished, confirmed: no, it wasn't enabled. OK, enable, then <i>umount</i> and remount 3 different file systems on <i>teevee</i>. But: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/home/grog</font> <font color="blue">10</font> -> <b><tt>mount /spool/Videos</tt></b></code><br />[tcp] hydra:/Video/spool/Videos: No such file or directory<br /> </tt></div></blockquote> <p> <i>/spool/Videos</i> is mounted on <i>hydra:/Video/spool/Videos</i>. <i>/Video</i> is already mounted. What's the problem? </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt><code><font color="blue">=== root@teevee (/dev/pts/12)</font> <font color="red">/home/grog</font> <font color="blue">11</font> -> <b><tt>ls -l /Video</tt></b></code><br />total 2019944<br />drwxrwxr-x 2 root operator 512 15 May 13:35 .snap<br />-r-------- 1 root wheel 655,589,376 23 Sep 13:59 .sujournal<br />drwxr-xr-x 11 root wheel 512 15 May 13:54 home<br />-rw-r--r-- 1 root wheel 2,116,891,893,624 16 May 04:28 homeimage.gz<br /> </tt></div></blockquote> <p> <b><i>OH</i></b>. What's that? Where have all my videos gone? I don't have a backup for this file system. </p> <p> Took a moment to recover. Did <i>fsck</i> find an old version of the disk? Potentially that could happen in very small areas, but we're talking about 2 TB of data here. The old random detection of USB drives after boot? But I only have one drive! </p> <p> More searching. Oh, how about that, there <i>is</i> more than one drive. But it's not connected. </p> <p> Still more searching. Oh, how about that, the drive <i>is</i> is connected, not just where I thought. And sure enough, not only was the drive undamaged, it was also clean. Panic over. And in the process discovered that the other disk has the same size as <i>/Video</i>, so I can use it as a backup disk. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-24T01:41:13+00:00</dc:date></item> <!-- topic ho not selected <p> <a href="http://www.lemis.com/grog/diary-sep2024.php?subtitle=Banking:%20the%20pain&article=D-20240917-023958#D-20240917-023958">Last week</a> we had one of many annoyances with the <a href="https://en.wikipedia.org/wiki/Bank_of_Melbourne">Bank of Melbourne</a>. I didn't note it at the time, but I asked them to take a complaint, for which they didn't give a reference number. </p> <p> Today we received mail from the Bank of Melbourne. The promised new debit card? Information on how to pick it up? No, no mention of a new card, which had been promised by today. </p> <p> It addressed the complaint. At least I now know the reference number, CS95140069. But I don't know what use it is: </p> <blockquote class="fullwidth"> <div class="listitemwidth"> Thank you for your time sharing your complaint. We are please we have been able to resolve the matter with you. </div> </blockquote> <p> “With you”? There was no interaction, and if they have resolved the matter, then only to their satisfaction. You'd think they were deliberately trying to insult her. </p> <p> If anything, the only effect of this message is to reinforce our intention to change banks. </p> End deselected topic ho (Insults from Bank of Melbourne) --> <!-- topic ko not selected <p> It's been <a href="http://www.lemis.com/grog/diary-aug2021.php?subtitle=KL%20Hokkien%20Mee&article=D-20210803-013826#D-20210803-013826">over 3 years</a> since I first started to try to cook <a href="http://www.lemis.com/grog/recipes/KL-Hokkien-Mee.php">KL Hokkien Mee</a>. Gradually it doesn't seem to be as complicated any more, but that's at least because of some of the simplifications made: use fried pork skin and onion flakes instead of greaves (chu yau cha, 猪油渣), don't brown the squid, making it possible to cook without smoking the place out, not like <a href="http://www.lemis.com/grog/diary-mar2022.php?subtitle=Smoke%20in%20the%20kitchen&article=D-20220318-000621#D-20220318-000621">on this occasion</a>: </p> <p> I also add chili and much more garlic than the recipes I have found. The result looks like this, before and after the onion flakes: </p> <p> </p> End deselected topic ko (KL Hokkien Mee: my version) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240925-014345</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240925-014345</link><category>technology</category><category>opinion</category><title>New ports</title> <description> <![CDATA[ <div align="justify"> <p> Somebody has broken <a href="http://www.gregroelofs.com/greg_xv.html"><i>xv</i></a>, a program that I have been using for ever. It was one of the programs on my <a href="http://www.lemis.com/grog/Software/Applications-for-UnixWare.php">Applications for UnixWare</a> CD, released in 1993. But the version I have on <i>hydra</i> doesn't recognize <a href="https://en.wikipedia.org/wiki/PNG">PNG</a> images. I've <a href="http://www.lemis.com/grog/diary-jul2024.php?subtitle=Fixing%20xv&article=D-20240730-225557#D-20240730-225557">established</a> that the problem is with a library, not <i>xv</i> itself, but does that help? </p> <p> There's a <a href="https://lists.freebsd.org/archives/freebsd-questions/2024-September/005709.html">thread</a> on the <a href="https://lists.freebsd.org/archives/freebsd-questions/">FreeBSD-questions</a> list about mail clients, but there was <a href="https://lists.freebsd.org/archives/freebsd-questions/2024-September/005751.html">one reply</a> suggesting <a href="https://feh.finalrewind.org/"><i>graphics/feh</i></a>. Installed that. Yes, it works, displays PNG. But there's nothing that I can do to influence the display. As far as I can see, no rotate, no resize. Not really a replacement for <i>xv</i>, not even in this limited context. </p> <p> Alastair Boyanich also recommended another tool for comparing source trees: <a href="https://meld.app/"><i>meld</i></a>. It's a <a href="https://en.wikipedia.org/wiki/GNOME">GNOME</a> application, but surprisingly installed easily and didn't pull in the normal hundreds of GNOME helper ports. I still don't dare try it out. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-25T01:43:45+00:00</dc:date></item> <!-- topic ko not selected <p> <a href="http://www.lemis.com/grog/diary-sep2024.php?subtitle=Roast%20beef%20again&article=D-20240922-030313#D-20240922-030313">Saturday's</a> <a href="https://en.wikipedia.org/wiki/Yorkshire_pudding">Yorkshire pudding</a> was not a success. Today we ate more of the roast, and I put the pudding in the “bathtub” <del><a href="http://www.lemis.com/grog/recipes/air-fryer.php">air fryer</a></del> convection oven for another 20 minutes at 210°. Still not a real recovery: </p> <p> But at least it tasted firmer. Clearly Yorkshire pudding is not a dish that you can recover like this, but it would be interesting to consider whether I could make it in the bathtub. </p> End deselected topic ko (Recovering the Yorkshire pudding) --> <!-- topic ah not selected <p> The cats have been interested in the garage for quite a while now, and we have suspected that they have found rodents. But I was still surprised when <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> came out of the garage this evening with a live mouse in his mouth. </p> <p> He went off to the front door with it, where he started playing with it. I put <a href="http://www.lemis.com/grog/Albums/Animals/Mona.php">Mona</a> there too, and they investigated in some detail: </p> <p> But they didn't attempt to kill it, confirming my suspicion <a href="http://www.lemis.com/grog/diary-sep2024.php?topics=a&subtitle=A%20mouse?%20%20Rats!&article=D-20240907-025631#D-20240907-025631">a few weeks ago</a> that Mona killed the rat by accident rather than by design. </p> <p> So: what do I do with the mouse? <a href="http://www.lemis.com/yvonne/">Yvonne</a> wanted to put it outside, but then it would just have found its way back into the garage. So: put it in a mousetrap: </p> <p> It's amazing all the things you can do with chopsticks. After that I put it outside to wait until the morning. </p> <p> In passing, it's interesting that when <a href="http://www.lemis.com/grog/Albums/Animals/Larissa.php">Larissa</a> came to join in the fun, Mona let her come really close without getting upset. Clearly there were more interesting things to do. </p> End deselected topic ah (Cats and mouse) --> <!-- topic a not selected <p> The mouse that we caught <a href="http://www.lemis.com/grog/diary-sep2024.php?subtitle=Cats%20and%20mouse&article=D-20240926-031346#D-20240926-031346">yesterday evening</a> was dead when I came out this morning. Probably it was on its last legs when <a href="http://www.lemis.com/grog/Albums/Animals/Bruno.php">Bruno</a> caught it. What should I do with it? The cats were clearly not interested. Offered it to <a href="http://www.lemis.com/grog/Albums/Animals/Lena.php">Elena</a>, who sniffed at it and walked off. Offered it to <a href="http://www.lemis.com/grog/Albums/Animals/Larissa.php">Larissa</a>, who took it off somewhere and presumably disposed of it. </p> End deselected topic a (What to do with a dead mouse) --> <!-- topic G not selected <p> This morning at breakfast, <a href="http://www.lemis.com/yvonne/">Yvonne</a> was looking decidedly unwell, and she told me that she was feeling unwell in a manner that suggested that her heart medication wasn't sufficient. She took another <a href="https://en.wikipedia.org/wiki/Flecainide">Flecainide</a> tablet, as instructed for such occasions, and gradually things weren't as bad. But it's happening more often. Time to talk to <a href="https://ballaratcardiology.com.au/staff/dr-rod-reddy">Rodney Reddy</a> again. Called up and—as always—was put on hold waiting for an answer. But none came: after less than 4 minutes, the phone system decided it had had enough, told me to call back later, and hung up. Moral: don't call in the middle of the day. </p> <p> Do we even have a referral for Dr. Reddy? Called up <a href="https://www.healthfirstballarat.com.au/">Health First</a> and asked for one. More hold while the receptionist spoke with Paul Smith, who suggested that she go to the Emergency Department—that way madness lies—but reported that she had a permanent referral. </p> <p> Called up <a href="https://ballaratcardiology.com.au/">Ballarat Cardiology</a> later and was given the information that Dr. Reddy is on leave until the end of the month, but that she should contact Paul Smith, whom she would inform. That looks like a waste of time, but we should check. </p> <p> Also asked her about mobile phone ECG devices. <a href="https://en.wikipedia.org/wiki/Apple_Watch">Apple Watch</a>! I don't see that happening; apart from the presumed fact that there are cheaper and better devices, that Yvonne wouldn't want to wear something that obstructive, and that she would almost certainly need an <a href="https://en.wikipedia.org/wiki/iPhone">iPhone</a> to work with it. She didn't have any other suggestions, but at least I didn't get the line “these ECG devices are all useless”. So I'll do some further investigation. </p> End deselected topic G (Yvonne unwell) --> <!-- topic go not selected <p> Nick Macdonald of <a href="https://instagram.com/macgardens">macGARDENS</a> along today, rather unexpectedly: he had been due to come yesterday, but didn't inform us, so we had written him off. He was here to take a look at the garden with a view to tidying it up. He recognized me instantly: “Wasn't I here a couple of years ago, and we didn't see eye to eye? You were quite antagonistic”. </p> <p> Huh? Was he the bloke who wanted to come only twice a year? I do recall him, and that our intentions didn't match up. But no, he wasn't that one. Spent some time talking, and though he's <i>really</i> expensive—$1,430 per day—he doesn't have to come that often, and we should be able to get by with Jesse doing most of the work for a fraction of the price. </p> <p> And was he here before? Yes, <a href="http://www.lemis.com/grog/diary-sep2021.php?subtitle=A%20new%20gardener?&article=D-20210922-012117#D-20210922-012117">three years ago</a>. He was the bloke who wanted to come only twice a year—or at least, that's what I wrote. Clearly some kind of misunderstanding. </p> End deselected topic go (Another gardener?) --> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240927-030749</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240927-030749</link><category>technology</category><category>history</category><category>opinion</category><title>Linux on Intel 4004</title> <description> <![CDATA[ <div align="justify"> <p> Interesting reference in email today: Dmitri Grinberg has run <a href="https://en.wikipedia.org/wiki/Linux">Linux</a> on an <a href="https://en.wikipedia.org/wiki/Intel_4004">Intel 4004</a>: </p> <iframe width="400" height="300" src="http://www.youtube.com/embed/NQZZ21WZZr0" allowfullscreen frameborder="0" > </iframe> <p> This video is at least partially in fast forward. It's worth looking at the clock in the bottom right corner of the laptop display. </p> <p> But how can it work? The 4004 only had maximum of 4 kB of memory. It seems that <a href="https://dmitry.gr/?r=05.Projects&proj=35.%20Linux4004#_TOC_6cd03fe22e73587a2147bd353d8c3391">it really just emulates a MIPS instruction set</a>. Still, an impressive achievement. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-27T03:07:49+00:00</dc:date></item> <item><guid isPermaLink='true'>http://www.lemis.com/grog/diary-sep2024.php#D-20240928-021431</guid><link>http://www.lemis.com/grog/diary-sep2024.php?topics=c#D-20240928-021431</link><category>technology</category><category>general</category><category>opinion</category><title>Instagram pain</title> <description> <![CDATA[ <div align="justify"> <p> I've had an <a href="https://en.wikipedia.org/wiki/Instagram">Instagram</a> account for I don't know how long. It's not the kind of site that I find useful, but Nick Macdonald seems to have <a href="https://instagram.com/macgardens">his only web presence</a> on it. OK, take a look. </p> <p> “Please login to view the remainder of this page”. OK, Instagram, if you wish. “<span style="color:red">Sorry, your password was incorrect. Please double-check your password</span>”. </p> <p> Damn you, Instagram. You have changed my password without telling me. OK, send me a “Login Link”. But nothing came. Repeat. Nothing came. The mail logs showed that it didn't even try. </p> <p> OK, new email address, new account. Select nick? Sorry, they're all taken or “not available”: </p> <table summary="Automatically generated table"> <tr> <td valign="top" align="left">stupidinstagrambpas</td> <td> </td> <td valign="top" align="left">user already exists</td> </tr> <tr> <td valign="top" align="left">biglier</td> <td> </td> <td valign="top" align="left">not available</td> </tr> <tr> <td valign="top" align="left">endlessfrustration</td> <td> </td> <td valign="top" align="left">not available</td> </tr> <tr> <td valign="top" align="left"></td> </tr> </table> <p> And yes, the first nick was really “stupidinstagrambpas”; I must have accidentally hit some extra keys. But “user already exists”? Come on, Instagram. You know that's not true, though “endlessfrustration” sounds likely. How about something in German? Yes, it accepted “verdammteidioten”. I suppose I should change it if I really want to do something with Instagram, but that's becoming ever less likely. </p> <p> But my success was short-lived. A message from Instagram that at least looked amusing in the inbox: </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt> 10 26-09-2024 To broken-insta ( 125) Instagram Action needed on your account, verdammteidioten </tt></div></blockquote> <p> The content, in <a href="https://en.wikipedia.org/wiki/HTML">HTML</a> </p> <blockquote class="fullwidth"><div style="text-align:left"> <tt>Take action or lose access to verdammteidioten<br />Hi verdammteidioten,<br /><br />Your Instagram account has been suspended. This is because your account, or activity on it, doesn't follow our Community Guidelines on account integrity and authentic identity.<br /><br />If you think we made a mistake, you have until March 26, 2025 to appeal. </tt></div></blockquote> <p> It gave me a login link that didn't work. Somehow my link name seems appropriate. My guess is that they had a background checker (maybe human) that decided that my nick was not appropriate. But then they should have said so and not immediately disabled my login. Should I care? None of this would have happened if they hadn't broken my original login, which remains broken. </p> </div> ]]> </description> <dc:creator>Greg Lehey</dc:creator> <dc:date>2024-09-28T02:14:31+00:00</dc:date></item> <!-- topic Go not selected <p> So it seems that it's a good idea to buy an <a href="https://en.wikipedia.org/wiki/ECG_monitor">ECG monitor</a> for <a href="http://www.lemis.com/yvonne/">Yvonne</a>. But which? Clearly the <a href="https://en.wikipedia.org/wiki/Apple_Watch">Apple Watch</a> is out of the question. But they have one at <a href="https://en.wikipedia.org/wiki/Chemist_Warehouse">Chemist Warehouse</a> that looks reasonable: the <a href="https://www.biosigdx.com/ecgme">Biosignals ecgme</a>. I could go into town and get one right now, assuming they're in stock. </p> <p> But it requires a mobile phone app to operate. How closely are the data tied to the app? Can I back them up? Off to look for the <a href="https://www.biosigdx.com/_files/ugd/a8ca81_8221aa60a8ca4e0db7fee967751e9c78.pdf">user manual</a>, presented as a single strip of 30 pages. Lots of warnings and things. How do I back up the data? No mention. Clearly this is a modern device. </p> <p> Much more searching brought me to the <a href="https://store.kardia.com/products/kardiamobile">Alivecor KardiaMobile</a>, and also the better <a href="https://store.kardia.com/products/kardiamobile6l">KardiaMobile 6L</a>, both of which have <a href="https://kardia.com/ifus/6l-row/19LB75.2.pdf">much better instructions</a>, in particular describing the fact that the ECGs can be downloaded in <a href="https://en.wikipedia.org/wiki/PDF">PDF</a> format, which I suppose is the best we can hope for at the moment, though the text suggests that something gets lost in the translation: </p> <blockquote class="fullwidth"> <div class="listitemwidth"> Health care professionals (HCPs) may review and analyze the recorded EKG. The Kardia app provides users with the ability to forward the recorded EKG to their HCP in one of two ways, one by using a referral code with HCPs who use Kardia Pro, and the other, by emailing the EKG PDF to the HCP. When connected to Kardia Pro, the user’s EKG recordings are automatically transferred and made available to the HCP. HCPs can review the EKG PDF to perform rhythm assessments as well as measure the QT interval. </div> </blockquote> <p> Also found a report that suggested that the KardiaMobile 6L performed much more accurately than the Apple Watch. </p> <p> In passing, a few recognitions: </p> <ul> <li class="fullwidth"> <div class="listitemwidth"> The SmartWatch approach has the obvious advantage that it can monitor continuously, while the phone-attached devices require manual intervention. </div> </li> <li class="fullwidth"> <div class="listitemwidth"> The approach in each case is via electrodes. Presumably the smart watches only have one, most phone-attached devices have two (one for each hand), and the best have three (the third for the left knee). </div> </li> <li class="fullwidth"> <div class="listitemwidth"> A dual electrode device such as the KardiaMobile is documented as supplying a single lead ECG. The triple electrode device such as the KardiaMobile 6L is documented as supplying a 6 lead ECG, thus the name. Others also repeat this claim, including the better of the two Biosignals ecgme. By comparison, a clinical ECG monitor has 12 leads. Nobody explains how you can make 1 out of 2 and 6 out of 3. </div> </li> </ul> End deselected topic Go (ECG monitor: walk before you can run) --> </channel></rss> If you would like to create a banner that links to this page (i.e. this validation result), do the following:
Download the "valid RSS" banner.
Upload the image to your own server. (This step is important. Please do not link directly to the image on this server.)
Add this HTML to your page (change the image src attribute if necessary):
If you would like to create a text link instead, here is the URL you can use:
http://www.feedvalidator.org/check.cgi?url=http%3A//www.lemis.com/grog/diary.xml.php%3Ftopics%3Dc
