|
| Greg's diary |
| March 2024 |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Friday, 1 March 2024 | Dereel | Images for 1 March 2024 |
| Top of page | ||
| next day | ||
| last day |
|
More VicEmergency insights
|
Topic: general, technology, opinion | Link here |
Following up on the bushfire today didn't bring much in the way of new insights. Yvonne drove down Kleins Road further than I had been able to do yesterday, but around the corner the road was closed, and there still wasn't very much to see on the 47 Kleins Road property. We'll have to wait a few days more.
But Petra Gietz has found another trick with the VicEmergency app. I had discoved the map settings (or was that “Filter”?) page and noted that I could only select one setting. But she had another: one that showed the wind direction. Currently:
That's useful, but there seem to be a number of issues with the function. First, as I had already mentioned, There Can Only Be One. Select this “filter” and the others are deselected. And the arrows don't scale. If I enlarge the display, the arrows stay as they are, and maybe disappear outside the display area:
|
|||||||||||||||||||||||
And at some point, the setting just disappears. I haven't yet established why or when.
And then there's a big difference between the mobile display (first image) and the web browser display:
|
|
|||||||||
|
|
|||||||||
Apart from the poor contrast, the burnt area (if that's what it is) is missing in the mobile phone display. Why?
|
Finally cataract surgery?
|
Topic: health, opinion | Link here |
I've been dragging my heels on my cataract surgery. I'm not afraid of the procedure, but I don't like the idea of having only partially functional vision for a week, and that for each eye. But it does have to happen. David Fabinyi wants to do my primary (left) eye first, but that doesn't seem a good idea. What if the vitrectomy shows up problems that we couldn't expect? I can get by with just cataract surgery on the left eye if it proves to be an issue. And should I really go home in the evening, or stay overnight in hospital? In principle I want to go home, but if there's a question of monitoring, it might be better to stay overnight.
Called up his surgery on 5221 3098 and spoke to Estelle, who told me that he wasn't available today, took notes and said that she would call back. Potentially the first eye could be done later this month.
|
Phone smart, again
|
Topic: technology, general, opinion | Link here |
One of the things that really annoys me about my Android mobile phone is that it's so difficult to keep apps running. In particular, basic functionality like an FTP server keeps stopping, though I have set it in some obscure place (not the app configuration) to stay running at all times.
It was all the more surprising that I recently found that it had been running for 4 or 5 days. But then it stopped, and now it keeps stopping every few hours. That can't be intended. How I hate these things!
|
Real uptime
|
Topic: technology, opinion | Link here |
Seen recently:
Current time: 2024-02-29 16:11:21 UTC
System booted: 2013-10-22 02:22:24 UTC (540w2d 13:48 ago)
Protocols started: 2020-10-26 16:08:37 UTC (174w3d 00:02 ago)
Last configured: 2024-01-07 01:40:54 UTC (7w4d 14:30 ago) by root
4:11PM up 3782 days, 13:49, 2 users, load averages: 0.00, 0.01, 0.00
That's a router somewhere in the FreeBSD complex. Yes, not a real computer, but still worthwhile. I'm told, though, that while the software is running fine, the hardware is worn out, so it's not likely to stay there much longer.
| Saturday, 2 March 2024 | Dereel | Images for 2 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Out, foul bot!
|
Topic: technology, opinion | Link here |
My web server continues to suffer from extreme overload:
2:09AM up 338 days, 17:57, 8 users, load averages: 177.41, 158.34, 121.08
And since it's the beginning of the month, I got my web server bill, round double what it has been in the past. And the server continues to hang. Clearly I need to do something.
Most of the traffic seems to come from web crawlers. OK, how about excluding them for a while? Put this in robots.txt:
User-agent: *
Disallow: /
That effectively tells new crawlers to stay away. What about ones that already have their tendrils in the system? Watching the server load and access times to robots.txt showed a correlation; after some hours, the load average gradually dropped to round 15, still not exactly idle (the system has 2 CPUs). But it looks as if things are under control for the moment. I'll keep an eye on it for a couple of days before allowing the bots back in a more limited manner.
|
Chasing down the Hugin stitch problem
|
Topic: technology, photography, opinion | Link here |
House photo day again today, and once again I had fun with Hugin, particularly with this panorama, stitched with hydra (Hugin version 2023.0.0.d88dc56ded0e) and with eureka (Hugin version 2018.0.0.5abfb4de7961) (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What's causing that? There's nothing obvious, and they're both stitched from exactly the same project file using exactly the same command:
hugin_executor --stitching e-from-house.pto
What does hugin_executor do behind the scenes? Ah, that would be telling. The man page says:
DESCRIPTION
hugin_executor is a command line tool, useful for stitching projects in
shell scripts
And that's really everything except the options, which aren't very interesting. But clearly it invokes nona and enfuse. Ah! There's a non-standard -d option for just printing the commands that it executes, along with an undescribed requirement to use either the -s (stitch) or -a (execute assistant, whatever that means in this context). Try that and get:
=== grog@eureka (/dev/pts/39) ~/Photos/20240302 147 -> hugin_executor -s -d C/e-from-house.pto
/usr/local/bin/nona -v -z LZW -r ldr -m TIFF_m -o e-from-house /Photos/grog/20240302/C/e-from-house.pto
/Photos/Tools/Enblend -w -f15158x4900+0+931 --compression=LZW -o e-from-house.tif -- e-from-house0000.tif e-from-house0001.tif e-from-house0002.tif e-from-house0003.tif e-from-house0004.tif e-from-house0005.tif
exiftool -overwrite_original -TagsFromFile /Photos/grog/20240302/C/e-from-house-0.tiff -WhitePoint -ColorSpace -@ /usr/local/share/hugin/data/hugin_exiftool_copy.arg -@ /var/tmp/he1Zxy10 e-from-house.tif
rm /var/tmp/he1Zxy10 e-from-house0000.tif e-from-house0001.tif e-from-house0002.tif e-from-house0003.tif e-from-house0004.tif e-from-house0005.tif
=== grog@hydra (/dev/pts/28) ~/Photos/20240302 89 -> hugin_executor -s -d C/e-from-house.pto
/usr/local/bin/nona -v -z LZW -r ldr -m TIFF_m -o e-from-house /Photos/grog/20240302/C/e-from-house.pto
/Photos/Tools/Enblend -w -f15158x4900+0+931 --compression=LZW -o e-from-house.tif -- e-from-house0000.tif e-from-house0001.tif e-from-house0002.tif e-from-house0003.tif e-from-house0004.tif e-from-house0005.tif
exiftool -overwrite_original -TagsFromFile /Photos/grog/20240302/C/e-from-house-0.tiff -WhitePoint -ColorSpace -@ /usr/local/share/hugin/data/hugin_exiftool_copy.arg -@ /var/tmp/heQauSlR e-from-house.tif
rm /var/tmp/heQauSlR e-from-house0000.tif e-from-house0001.tif e-from-house0002.tif e-from-house0003.tif e-from-house0004.tif e-from-house0005.tif
Yes, there is a difference between the two command outputs, but it's not important: the name of the temporary output file from exiftool is different. Apart from that, it's identical. So what is the problem? And why does it almost always happen in the same place in this particular panorama? One clue is in the invocation of enblend: it's /Photos/Tools/Enblend, a script that I wrote. In itself it's uninteresting:
#!/bin/sh
# $%39d: Enblend,v 1.1 2018/07/01 01:32:35 grog Exp grog $
# Wrap enblend to be able to report finish.
echo enblend $* > myenblend
enblend $*
# espeak "enblended"
But it makes clear that something is going on in the background, presumably to read some configuration information. And not surprisingly the problem occurs when running enblend. Changing horses in mid-stream shows that I can use enblend on eureka to stitch the output from nona and get good results.
| Sunday, 3 March 2024 | Dereel | Images for 3 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Web server calmer
|
Topic: technology, opinion | Link here |
Watched the web server load all day today. And yes, the load average stayed relatively constant in the 8 to 20 range.
Fixed? Hardly. I do like to appear on Google search results, especially for things for which there are few hits, for example (surprisingly) Swine Bismarck, which we ate last night. And my robots.txt has only excluded the well-behaved crawlers, as my log file shows:
47.76.35.19 - - [03/Mar/2024:01:04:46 +0000] "GET /grog/diary-mar2020.php?dirdate=20190323&imagesizes=1223021022213223222222223221021222221221322220222222320232401232222212202311232012222221323212323222222222222231022223222332221222340222221122212222312202222202232232222203221222122222220212222222222220221&size=2 HTTP/1.1" 200 1340527 "-" "Mozilla/5.0 (Windows NT 6.2; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/77.0.2827.33 Safari/537.36"
47.76.35.19 - - [03/Mar/2024:01:04:47 +0000] "GET /grog/diary-mar2020.php?dirdate=19960415&imagesizes=02022210022102222202022222223212200110223113101223213223221412320222204022211322120202212032233232032202012022021221201002022021203122212022222322223122222322032013332421022211221023222211121221222222102221&size=2 HTTP/1.1" 200 1225067 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3234.7 Safari/537.36"
47.76.35.19 - - [03/Mar/2024:01:04:47 +0000] "GET /grog/diary-mar2020.php?dirdate=20191206&imagesizes=20222220022222221222222223222112222222224222222221223203332111322232212222210320202222212223222032022222221222222210221222022222202222222221012222222222232212022002220202022212221022112231102222221112202223&size=2 HTTP/1.1" 200 1308264 "-" "Mozilla/5.0 (Windows NT 6.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.2023.7 Safari/537.36"
At the very least I still need to look at replacing GET with POST.
|
CAPTCHAs on the rise!
|
Topic: general, technology, opinion | Link here |
I had hoped that CAPTCHAs were finally going away, but today I was confronted with two particularly virulent ones, from Vultr and Wise, the vaguely named money transfer service. Not only multiple CAPTCHAs, but ones that change when you click things, and some which are so vague that I can't recognize them, apart from the use of US American terminology like “crosswalk”.
Will they never go away? Can't somebody use Artificial Intelligence to make a CAPTCHA-solving browser plugin? Please?
|
UnWise
|
Topic: general, technology, opinion | Link here |
Yvonne wants to transfer money to Austria again. OK, that's a case for this stupidly renamed Wise funds transfer service. They used to be called Transferwise, but clearly that name was too descriptive.
OK, the transfer is for round € 1000. I don't have that much in my account, so I needed to transfer it from a real bank account. But first I had to solve these really horrible CAPTCHAs. And though normal bank transfers in Australia are performed in real time, Wise told me that it would take a day. Wise told me that it would take two days. Clearly they couldn't decide. And my bank also said that it would take two days.
Why? Are they trying to annoy me? No, I don't think so. It's more likely that they're trying to profit on the infinitesimal interest on the money for one or two days. Is it worth annoying your customers for that? It's time to look for alternatives.
| Monday, 4 March 2024 | Dereel | Images for 4 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Will the genocide never end?
|
Topic: history, politics, opinion | Link here |
It's been nearly 5 months since the 2023 Hamas-led attack on Israel, in which 2,752 people were killed, 1,143 of them Israeli civilians. A further 242 civilians were taken hostage. If I read the Wikipedia page correctly, most of the casualties were Hamas fighters, something that people seldom mention.
That's horrible, of course. If the Palestinians want justice, there are other ways. Or are they? They've been under siege for 16 years, and all attempts to get out from under the Israeli yoke seem to have failed. The Israelis, who were maintaining the siege, should have expected that they wouldn't just sit back and take it. Still, murdering innocent civilians is Just Plain Wrong.
So what do the Israelis so? They fight back with their US-backed weaponry. So far they have killed over 30,000 people (“Palestinians”), most of them innocent civilians, mainly women and children. Sorry, Israel, if it's wrong to kill Israeli civilians (and of course it is), it's wrong to kill other civilians. But they haven't stopped, and they are systematically destroying Gaza and starving the remaining population.
Words fail me. How can they continue doing this? How could they start doing this? Why is the USA supporting the genocide? Both countries have completely lost any legitimacy. The USA, claiming to hold the moral high ground, is complicit in some of the worst atrocities I have ever experienced. By comparison, the Russian invasion of Ukraine seems like nothing.
Somehow many Israelis have a mental block. I've been gradually watching Asher Susser's lectures about the Al Nakba. But that's not a word that he likes to use: it suggests a natural catastrophe, one that absolves the Palestinians of the blame. And the information that he presents, though undoubtedly correct, seems to be biased towards the Israeli viewpoint. It's hard to be sure if there are any inaccuracies, but round 1948 the story strongly differs in emphasis from the Al Jazeera videos.
And that seems to pervade Israeli society. For some years I have been reading the Times of Israel and considered it relatively neutral. But here, too, the emphasis is off-centre. Return the remaining hostages, many of whom were killed by Israeli fire. Who cares about the Gazans? Was anybody shot while trying to get food? No, they stampeded and accidentally got bullets in their skulls.
When will this nightmare end? What would have happened in the Second World War if the news media had had the coverage that they do now? Would that have changed the behaviour of the Nazis?
But then, is news free? Look any day on Al Jazeera News to see the conditions. Or look on any US news site to see an expurgated version, or on an Israeli site to see nothing at all.
|
More gardening and bushfire insights
|
Topic: gardening, general | Link here |
Paul Donaghy along to trim the grass today. He brought the news that his property suffered only minor damage, just the back fence (west) damaged. Not by the fire, but by the fire crews, who tore it down to create a fire break.
And it seems that the fire was caused by a defective electric fence device (“actuator”), though the reports I have seen seem a little confusing. But clearly that's a thing to think of on days of total fire bans.
|
Another migraine
|
Topic: health | Link here |
I don't know if the Israeli situation has something to do with it, but this afternoon I had yet another minor migraine.
|
Evil passwords again
|
Topic: technology, opinion | Link here |
The FreeBSD is using Matrix for internal communications. And that requires a Kerberos password. How do you do that? Ah, yes, from this diary. OK, generate a password. This silly “evil random passwords” again!
===================================================================
Generating strong, evil random passwords...
===================================================================
Your new, ready to forget, password: ung...
Your new, ready to forget, mail password: ahXoog...
But it worked. Only mail doesn't work any more. It always creates a new password. More pain. Maybe that's why they're evil.
| Tuesday, 5 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
CJ's pain again
|
Topic: technology, opinion | Link here |
CJ Ellis round this morning with his computer and most of the stuff I asked to bring with him. He has been locked out of Facebook! Oh horror! Hard to say what caused that, and I don't really want to know, but they sent him instructions on how to reset it, requiring email with a code being sent to his mail server (Gmail). But the code was only valid for 30 seconds, and he didn't know how to open two tabs on his web browser. By the time he had written down the code and re-entered firefox, the code had expired.
OK, I have access to his Gmail system, so for this exceptional case I'll connect to the mail system. CJ solicits the code, I read it out to him, and all is well. Except that I can't log in. The password must have changed.
So for all that he brought his machine over here. Off looking for the missing cables, and got the thing up and running. Yes, no problem. Fire up Chrome. No go, wrong password.
Ah, says CJ, I don't know that. Fires up firefox and logs in. No problem. Same password.
So why did I have difficulty? Why did his machine have difficulty with Chrome? That's the Google browser, after all. My only guess is that it must be some misguided security measure posing as a different problem.
On with Facebook recovery. Yes, got the code. Now please upload your “ID”. Take a photo of it and send it to us.
Problem: CJ doesn't have a mobile phone, doesn't have a computer with a camera, doesn't have a camera. So one way or the other he had to come to me so that I could scan his driver license and send it to him.
But how? On a real computer I would just fight my way past firefox breakage, load the image and send it. But this is Microsoft. How do I do it? Send it round the world to Gmail and back? Did that. Now use this famous Microsoft “drag” to drag the image to firefox. No, it didn't want to know, politely overlaying it with an “end of limit” symbol to allow me to guess that this was a limitation, that dragging didn't work. Why not? Again, too polite to say, but presumably because the image is really on the Gmail server at the other end of the world. OK, save the image to disk. Sorry, can only save the entire message.
Did that anyway, and of course firefox didn't want to know.
Dammit, how do these things work? Do these things even work? But deep down in “Windows”, struggling to get out, is a program called FTP.EXE. Fire that up, ftp to eureka, and pull it across. Just be careful to get it in BIN format, which for some reason requires two attempts. And then go tree-climbing with firefox and pull it up, send it off.
We will examine what you sent us and respond within 48 hours.
Well, that's CJ's problem, not mine. The whole thing took about an hour, and at the end it was all I could do not to have a fit of screaming. I wonder whom I can get to solve CJ's next problem.
|
Gmail annoyances
|
Topic: technology, opinion | Link here |
Part of the ordeal with CJ's computer was complicated by another issue with Gmail:
<cjellis@gmail.com>: host gmail-smtp-in.l.google.com[142.251.2.26] said:
550-5.7.26 This mail has been blocked because the sender is
unauthenticated. 550-5.7.26 Gmail requires all senders to authenticate with
either SPF or DKIM. 550-5.7.26 550-5.7.26 Authentication results:
550-5.7.26 DKIM = did not pass 550-5.7.26 SPF [hydra.lemis.com] with ip:
[45.32.70.18] = did not pass 550-5.7.26 550-5.7.26 For instructions on
setting up authentication, go to 550 5.7.26
https://support.google.com/mail/answer/81126#authentication
x2-20020a1709027c0200b001dca8522501si8881415pll.276 - gsmtp (in reply to
end of DATA command)
I've seen this before. On that occasion, like Schrödinger's cat, it both got delivered and got rejected. Since then I had discovered that if I sent a message from hydra, it got rejected, but bouncing exactly the same message from eureka worked.
Today I tried it differently: bounce from hydra. Works! So what's wrong with this system? Breakage of a new kind? Everything seems to be breaking lately.
|
Updating NBN
|
Topic: technology, general, opinion | Link here |
It seems that the NBN truck that I saw last week was ahead of its time. Today I received mail with more verbiage. But it seems that it only applies to fibre: upgrade data rates from 100 Mb/s to 500 Mb/s, or up to 1 Gb/s for faster connections. And somehow this will end up costing more money, not because the links are any more expensive, but because people will use more data:
“The average household now consumes 443 gigabytes per month across 22 internet-connected devices," she said. "We predict that average will grow to 33 connected devices by 2026 and 40 by the end of the decade."
That sounds remarkably high. I don't consider my household to be “average”, but I think that I have more “devices” than most people. Counting them, including the ones that I don't really use, I have about 7 computers, 3 mobile phones, a TV and a PV inverter, a total of 12. What do other people have that I don't? And though I've been downloading lots of videos this month, my usage will only reach about 870 GB. And at 100 Mb/s, I could transfer that in about 20 hours. But even if it's correct, it seems unlikely that the use will increase significantly due to higher link speeds, especially for people with a traffic cap. It may, however, serve as an excuse to increase the pricing.
And what does this say about the NBN truck in Dereel? I had noted at the time that the person there didn't seem to be very well-informed. Will there in fact be an upgrade of the Fixed Wireless service? I wouldn't be surprised if there were not.
| Wednesday, 6 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
10 more years of Internet
|
Topic: history, technology, opinion | Link here |
Ten years ago I was doing a Coursera course on the history of the Internet, and one of the assignments was: “Write an essay that imagines how the Internet will be different 20 years from now”. And that in 1000 words.
I did make the 1000 word maximum, but I also wrote a slightly longer version that said what I wanted to say.
So: now we have the halfway point. I've been keeping my eye on proof or disproof of my claims, and I'll continue to do so. Here the biggest takeaways so far:
Most purchases will occur on-line, and the few remaining shops will mainly exist to order and supply goods available on the Internet.
This seems to be well on its way.
I think this has largely happened already.
Radio and TV broadcasts will gradually cease.
Not yet, for sure. I can't see any evidence that there has even been a drop in terrestrial broadcasts, though it could be happening. Some programmes, though, are only available by networks.
Traditional publication will mainly cease.
There's a lot of evidence of that. I read news“papers” mainly online, and it seems that that's a common practice.
The public switched telephone network will cease to exist.
Yes, there's a lot of evidence of this, but not the way I thought. Most people now use mobile phones, and while they can use the Internet, there's a dichotomy between the traditional mobile phone networks and the Internet. It'll be interesting to see whether they can get their act together in the next 10 years.
Internet privacy will gradually become less powerful.
Maybe less powerful, but more obnoxious. I can't decide on this one.
Pornography will not go away.
This is clearly the case. Porn seems to account for a large portion of Internet traffic, and it may provide the impetus for its expansion (why does the National Broadband Network really think that people require a 1 Gb/s link?).
Software and media licensing will continue to be a contentious issue.
Surprisingly, this seems to be becoming less of a problem, at least from my viewpoint.
And then there were two things that I didn't want to predict:
How important will social networks be? Will they reach a plateau or even decline?
I still don't know the answer to this one. I don't follow social networks (well, hardly), but I have seen some evidence of decline. But maybe that will just be to a plateau. Certainly they seem to have taken over messaging such as email, breaking interoperability into incompatible, ugly proprietary formats.
Will the ideal of telecommuting become a reality?
I haven't seen much evidence. The COVID-19 pandemic gave a good reason to do so, but after it has quietened down, things seem to be reverting to on-site work.
So: what do the next 10 years hold? More of the same, I would think.
|
Keeping up to date
|
Topic: technology, opinion | Link here |
Various sites are telling me that my Chrome is out of date. And that only 5 months after installing the then latest and greatest!
OK, pkg upgrade. “I will remove all the unrelated packages that are dear to your heart, this time with the exception of Emacs”. No thanks. Build from scratch, trying not to get too annoyed at the continual resizing of the screen. Fail:
===> node-20.10.0_1 has known vulnerabilities:
node-20.10.0_1 is vulnerable:
WWW: https://vuxml.FreeBSD.org/freebsd/77a6f1c9-d7d2-11ee-bb12-001b217b3468.html
1 problem(s) in 1 installed package(s) found.
=> Please update your ports tree and try again.
=> Note: Vulnerable ports are marked as such even if there is no update available.
=> If you wish to ignore this vulnerability rebuild with 'make DISABLE_VULNERABILITIES=yes'
What kind of nonsense is that? This is a ports tree updated this morning, and it seems that node is required for one of the most important ports. How do they build it? OK, try again, with DISABLE_VULNERABILITIES set to an offensive comment.
That worked, then
===> Installing for gmake-4.4.1
===> Checking if gmake is already installed
===> An older version of gmake is already installed (gmake-4.3_2)
You may wish to ``make deinstall'' and install this port again
by ``make reinstall'' to upgrade it properly.
If you really wish to overwrite the old port of gmake
without deleting it first, set the variable "FORCE_PKG_REGISTER"
in your environment or the "make install" command line.
So Chrome is so finicky that it really requires the latest and greatest GNU make? I don't believe a word of it. But once again I had to restart the build.
After that, Chrome built and installed. But why is this almost always such a pain? Looking back through this diary, I have been having this problem every few weeks for the last 25 years.
I wasn't done yet. Stop the Chrome on hydra:0.3 and restart it. It displayed on hydra:0.1! Google breakage? Doubtless, but not the way it seems. The Chrome on hydra:0.3 was running from eureka, something that I thought no longer worked. Stop the instance on hydra:0.1 and I was able to start it as I wanted. How I hate broken software!
And, of course, was my flood of CAPTCHAs recently related to this down-rev Chrome? Potentially in part, but it also happened with a recent version of firefox.
|
Security above all
|
Topic: technology, opinion | Link here |
CJ's and my pain yesterday had a justification, at least from Facebook's point of view. Somebody had compromised CJ's account, and they wanted to confirm that it was really he. So they sent a 6 digit code to his email system, which ultimately I read and typed into the Facebook screen. Then they wanted a photo of his driver licence, which I scanned in for him. CJ wasn't involved at all in verifying his identity! And of course, they would by far have preferred to send him a code to his mobile phone, if he had one.
When will people realize that this is a completely useless way to confirm identity?
A related issue is my pain with Wise. These CAPTCHAs drive me mad. OK, try the workaround: Microsoft. And it worked. Yes, I had to fill out a CAPTCHA, but at the end it was happy enough to tell me that I wouldn't have to do so again. And finally I was able to send my money, with only two messages sent to my mobile phone. People, what's secure about a mobile phone? What if I had been stupid enough to enter the transfer details on the phone? Then a message to the same phone would be of no use at all.
In passing, the login on Microsoft also sent me an email message, asking me to confirm that the following information was correct:
Was this you?
Device Firefox, Windows 10
Location Germany
Date and Time 5 March 2024, 04:54 (GMT)
It's been over a year since my network was registered as being in Australia. What kind of out-of-date data is telling them that I'm in Germany? They could at least have used last year's exchange rates, which would have given me an advantage of between 5% and 10%.
| Thursday, 7 March 2024 | Dereel | Images for 7 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Crash!
|
Topic: technology | Link here |
One of the first things I do in the morning is check the load on www.lemis.com, canonically named lax.lemis.com. But the top program had stopped. And so had all the other windows to lax.
Oh:
=== grog@lax (/dev/pts/4) ~ 1 -> uptime
11:16PM up 3:38, 8 users, load averages: 2.78, 3.14, 2.91
The system rebooted! And yes, /var/log/messages confirmed:
Mar 6 19:35:23 lax qpopper[19872]: Stats: ...
Mar 6 19:38:41 lax syslogd: kernel boot file is /boot/kernel/kernel
Mar 6 19:38:41 lax kernel: ---<<BOOT>>---
What caused that? The only other clues were:
Mar 6 19:38:41 lax kernel: Trying to mount root from ufs:/dev/ufs/rootfs [rw]...
Mar 6 19:38:41 lax kernel: WARNING: / was not properly dismounted
And that was all. The second clue was that there was no dump or panic message, though arguably that's because I don't have a dumpdev on lax. So it looks like a Vultr problem. The good news was that everything seems to have come up normally.
|
Aussie: support me!
|
Topic: technology, opinion | Link here |
In the afternoon, my IRC proxy failed again. It's on lax, so off to check. Can't reach system! Has it gone down again? It didn't come back. Can I access it via ffm.lemis.com (in Frankfurt/Main, obviously)? Yes! So it's a routing problem.
A quick traceroute showed:
=== grog@hydra (/dev/pts/45) ~/FreeBSD 57 -> traceroute lax
...
5 be400.lsr1.nextdc-s2.syd.aussiebb.net (180.150.0.172) 174.822 ms 174.776 ms *
6 be30.bdr1.coresite-la1.lax.aussiebb.net (202.142.143.199) 182.125 ms 179.557 ms 180.318 ms
7 * * *
Close, but no cigar? It's reasonable to assume that lax.aussibb.net, like lax.lemis.com, is located in Los Angeles. Does Aussie have routing problems? Braved the expected response from “support”, called Aussie support on 1300 880 905 and explained my problem. Or at least, I thought I did. I explained where the traceroute stopped, and got the astounding question: “does it come up or go in circles?”. “Was the page loading?”, she meant.
My best guess is that she thought I was using a web browser to access the site, and that “going in circles” is what some browser or another does while it's waiting. After 4½ minutes on the phone it was clear that she thought that a router was a box on your desk that connected to the wall. I explained the very basics of Internet routing and that the buck stopped in lax. She said that routing was nothing to do with Aussie Broadband. I asked for somebody more knowledgeable, but that didn't help much.
In the meantime, the system came back online, but I still wanted to follow up. How do I report a routing problem? I asked to be connected to a supervisor, and after about 25 minutes I was connected to Mark, who at least sounded as he understood the concepts. No, there's no way to contact the networking people. His suggestion: send email and then call support 30 minutes later. I pointed out that they claim to answer email in a time frame of days, but he says that if I follow up with a call, it will get done in a timely fashion (if 30 minutes can fall into that category). I suppose it's better than nothing.
And when I was done, I found that I had received an email:
Subject: Fault #27960640 created
Now what good is that going to do? Will it get looked at? Their "my aussie" web site doesn't want to know about it.
Later Daniel O'Connor told me that I could get a business connection for the same price, with a free static IP address, and with better service personnel. Now that LEMIS has closed down, can I still do that?
|
Working round Gmail breakage
|
Topic: technology, opinion | Link here |
Once again I sent a Schrödinger's cat email today. The recipient was rejected, the Cc: address worked.
From MAILER-DAEMON Fri Mar 8 09:44:52 2024
<jeffrab@gmail.com>: host gmail-smtp-in.l.google.com[142.251.2.26]
said: 550-5.7.26 This mail has been blocked because the sender is
unauthenticated. 550-5.7.26 Gmail requires all senders to authenticate with
either SPF or DKIM. 550-5.7.26 550-5.7.26 Authentication results:
550-5.7.26 DKIM = did not pass 550-5.7.26 SPF [hydra.lemis.com] with ip:
[45.32.70.18] = did not pass 550-5.7.26 550-5.7.26 For instructions on
setting up authentication, go to 550 5.7.26
https://support.google.com/mail/answer/81126#authentication
s31-20020a63451f000000b005dc493c9496si14163290pga.507 - gsmtp (in reply to
end of DATA command)
OK, we know that. Just bounce it to him:
From MAILER-DAEMON Fri Mar 8 09:46:23 2024
<jeffrab@gmail.com>: delivery via
gmail-smtp-in.l.google.com[142.251.2.27]:25: 250 2.0.0 OK 1709851582
m12-20020a633f0c000000b005e438e96ae6si14850286pga.601 - gsmtp
On this occasion the IP address of the receiving MTA was different, but that doesn't seem to be relevant.
|
Cataract surgery date
|
Topic: health | Link here |
Call from Estelle from David Fabinyi: yes, we can do my right eye first, I don't need to stay overnight in Geelong, and the two operations can be done on 21 March and (pending success of the first) 11 April. It should all be over by ANZAC Day.
|
Visit for next Anke clinic
|
Topic: general, animals | Link here |
Anke Hawke is coming for the weekend for another clinic. Not a good time, as it turns out: the daytime temperatures will be round 38° for the whole time, and already 3 of 8 participants have canceled. But one, Ainslie Kimber, is already here and staying with us. She and Yvonne have plenty of things planned that don't rely on the clinic.
|
Google: so nice, so nice, we do it twics
|
Topic: technology, opinion | Link here |
Yvonne and Ainslie over to Chris Bahlo's place today, despite relatively tight timing. I kept an eye on their location, and indeed they made it back in time, apparently one by one:
|
|||||||||||||||||||||||
Now how did that happen?
|
IP address location: still broken
|
Topic: technology, opinion | Link here |
A few years ago I noted how inaccurate IP location services are. Now I have officially moved my address block to Australia. Have they improved? Tried again today and got:
| Site | Groggy | |
| iplocation.io | Schellnhausen | |
| db-ip | Frankfurt, Braubachstrasse | |
| criminalip | NSW, Garnpung lake | |
| ipinfo ipinfo | Melbourne GPO | |
| (ip2location) | Schellnhausen | |
| nordvpn | fail | |
| keycdn | NSW, Garnpung lake | |
iplocation.net offers multiple lookups, some conflicting. It claims that ipinfo has me in Niederdorla, but ongoing there I'm suddenly in Melbourne again.
So what's wrong with this picture? We have 5 different locations, all wrong:
The bottom line, though: although I have registered the net block in Australia a year ago, the majority of the lookups put me in Germany, a couple even in a location that disappeared over a quarter of a century ago, or three-quarters of the Internet history. There's really no excuse for such completely inaccurate reporting.
|
Ladyhawke? Werewolf?
|
Topic: language, multimedia, opinion | Link here |
Watching Ladyhawke on TV tonight. Somehow I didn't like it as much as on previous occasions. But a question arose: the heroes were the man who turned into a wolf by night, and the woman who turned into a hawk by day. For the latter they coined the term “Ladyhawke” (no explanation for the terminal e). What do you call a half man, half wolf? There's a word for that: “were” (an old word for “man”) and “wolf". But that was probably not the impression they wanted to create.
| Friday, 8 March 2024 | Dereel | Images for 8 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Is it a total fire ban today?
|
Topic: rant, general, technology, opinion | Link here |
Seventeen years ago, even before we moved to Dereel, I checked the Victorian emergency services and discovered, as I put it at the time, that they “... think of the web sites more as a geek sport than a way to disseminate information”. You had to search to find that kind of information.
It's looking like being a really hot weekend, again potential fire danger. Times have changed, and now I get informed when there's a total fire ban—by the Ballarat Courier, who make the content available without a subscription. Today I received such a notification at 15:00. The emergency services web sites have learnt something in the last 17 years, right?
Wrong. At 16:47:42 they issued a weather warning, but no mention of a fire ban. And of course a notification was sent to my mobile phone:
|
|||||||||||||||||||||||
Looking further at my phone, I saw:
|
|||||||||||||||||||||||
That was after waiting many seconds for the page to update. Before, which I unfortunately missed capturing, it showed data for “2/28/24”, presumably a US American format. What are they thinking by displaying that information at all, and in that format?
No mention of a fire ban. But with a bit of coaxing it was prepared to give a clue:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Where's the information? In the colour of the map background. Any fool (well, those at VicEmergency, anyway) knows that orange means “total fire ban”. If you have to, select “Incidents” (which it isn't), and get:
|
|||||||||||||||||||||||
People, this is literally a matter of life and death! Why can't you get your act together?
|
What is “feels like” temperature?
|
Topic: technology, opinion | Link here |
The other half of the fire danger is the weather, of course. I've already established that the Australian Bureau of Meteorology has two competing weather forecasts, which can differ by a degree or two. So when Ainslie mentioned her favourite mobile phone app, AccuWeather, I tried it out.
It didn't start well. I needed the “AccuWeather Australia” app, which made sense. And it came up telling me that the outside temperature was 70°.
|
|
|||||||||
No way! This silly thing doesn't understand modern units! Well, it does if you insist, but why should you need to? It should display in the units of the country for which it was prepared. Following “See 4 Hours” suggested that its biggest concern was rainfall, not temperature. Could it be that it comes from the United Kingdom? That would at least explain the offer of Imperial units, where they don't even mention US Customary.
But the most interesting thing, and one that Ainslie particularly likes, is the “feels like” temperature. I'm obviously not clever enough to find it in the app, but in the image above it claims that “RealFeel®” is 75° F (24° C)”, 3° higher than the actual temperature. And in general “RealFeel®” is higher. But the BoM also has a “feels like” temperature:
|
|||||||||||||||||||||||
And here it's lower than the real temperature. So what does it mean? Ainslie considers the “RealFeel®” temperature more important than the real temperature. Should she?
|
CJ exploited?
|
Topic: technology, opinion | Link here |
Mail from Eddie Purcell, who suggested that Facebook would never ask for the kind of information that CJ Ellis had to provide on Tuesday. Is he right? Certainly it looked strange, but as I have been commenting lately, lots of things look strange.
But then today Yvonne got a message from CJ via Facebook, asking her to confirm who he was; she should have received a code on her phone. He needed her confirmation to be able to set up his mobile phone.
What, CJ and mobile phone? Called him, no reply. Sent him email, which Gmail refused until I bounced it, and he called back in the evening. No, still no Facebook access. So somebody has stolen his identity, and I helped them! That's what comes of getting mixed up in things you don't understand.
What do we do now? Cancel what we can, maybe including his driver license, and report it. But how? These social networks should make it easier to freeze accounts.
|
Bruno stuck
|
Topic: animals | Link here |
During dinner, we heard a bang, followed by a plaintive “meow”. Bruno had fallen behind the washing machine:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's still better than last time, where Yvonne found him in the machine with the dirty linen just before she started the wash cycle.
|
Werwolf or Werewolf?
|
Topic: language, opinion | Link here |
After writing yesterday's article on Ladyhawke and werewolves, I checked my dictionaries. The OED disagrees with me about the etymology of the word “werewolf”:
Old English werewulf (once), = Middle Dutch and Dutch weerwolf, Middle High German werwolf (German wer-, wehrwolf), Low German werwulf; also West Frisian waerûl, warûle (and waerwolf after Dutch), ...The first element has usually been identified with Old English wer man were n.1, but the form were- in place of wer- (compare however were- and wergild wergild n.), and the variants in war-, var-, makes this somewhat doubtful.
So we have the rather strange claim that the derivation is doubtful, though it exists in all West Germanic languages. What do the Brothers Grimm say?
werwolf, m. , mythologische bezeichnung für einen menschen, der auch in wolfsgestalt auftreten kann, mannwolf (wer 'mann, mensch', s. 1DWb wergeld). das wort ist im ahd. lediglich als eigenname bezeugt (Weriuuolf, s. zs. f. dt. altert. 12, 252) und in späterer zeit dann bei Burchard v. Worms und Berthold v. Regensburg (werwolf, s. u.) als gattungsbezeichnung nachweisbar. mhd. dichter haben werwolf offenbar nicht verwendet; erst seit dem 15. jh. erscheint das wort häufiger, teilweise in volksetymologischer umdeutung: bärwolf (so z. b. bei Luther; vgl. teil 1, 1146, beerwolf teil 1, 1244). german. parallelen finden sich im ae. wer(e)wulf, im mnd. warwulf, im mndl.-ndl. weerwolf;...
Despite the archaic presentation, it seems that that was written in 1959. And the OED article?
First published 1926; not yet revised
Presumably it could do with revision. My guess is that they would find another explanation for the medial e.
| Saturday, 9 March 2024 | Dereel | Images for 9 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Weather forecasts
|
Topic: technology, general, opinion | Link here |
Very hot day today, with a top of 40°, only 0.2° less than the maximum all summer:
How did the various weather apps predict the temperatures?
| Time | AccuWeather | BoM | Measured | |||
| 12:00 | 32° | 33.1° | ||||
| 13:00 | 34° | 34° | 36.0° | |||
| 14:00 | 36° | 36° | 36.5° | |||
| 15:00 | 38° | 37° | 38.2° | |||
| 16:00 | 40° | 37° | 38.0° | |||
| 17:00 | 39° | 37° | 38.3° | |||
| 18:00 | 37° | 36° | 39.5° | |||
| 19:00 | 36° | 33° | 37.6° | |||
| 20:00 | 33° | 30° | 32.3° | |||
| 21:00 | 31° | 26.6° |
What should I make of that? Both are incorrect, but that was to be expected. In general AccuWeather shows higher temperatures than the Bureau of Meteorology, but the relationship to reality is no better.
|
Clinic dinner fail
|
Topic: food and drink, animals, opinion | Link here |
First day of the Anke Hawke clinic today. In the past the majority of the participants came to dinner with us, but it seems that those days are over. Apart from Yvonne, Ainslie and Anke herself (whom I saw briefly in the morning), the only other participants were Chris Bahlo and Julie Donaghy, both of whom were too busy. In the end, it was only Ainslie who joined us, and she would have anyway.
A good thing too. The dinner was a catastrophe. Ainslie's “vegetarian”, though she also eats fish, so we had ikan goreng, alu masala and dal.
But there were problems. We usually make the ikan goreng with tenggiri (Spanish mackerel) steaks, but we couldn't find any, so Yvonne bought swordfish instead. The alu masala needs black mustard seeds, but we were out of them, so I had to use yellow mustard seeds instead. And our new blender wasn't able to puree the onions.
On serving, the fish was partially tough. I thought that that was probably my fault, but it was only in one place, so maybe it was just a problem with the fish itself. And the alu masala wasn't properly cooked! It's potato cubes, and it had been cooking slowly (clearly too slowly) for nearly an hour! In addition, both of them weren't flavoursome enough. It seems that the tamarind paste that we used for both dishes was much weaker than what we used to use. I've seldom had such a poor result from considerable effort.
| Sunday, 10 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Relax!
|
Topic: general, technology, photography | Link here |
Somehow I didn't do very much today. The continuing hot weather didn't help, but it seems that I was just catching up with my mail; the Unix Heritage Society seems to be particularly active lately. And I've been given for review the draft of a book on running email servers, something very close to my heart. And then Hugin came up with some problems that took me a while to fix.
But is that enough to keep me going all day? It seems so. Somehow I'm slowing down.
| Monday, 11 March 2024 | Dereel | Images for 11 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
RIP Baron Louis Binos de Pombarat
|
Topic: general, history | Link here |
Yvonne received sad news by Facebook today: Baron Louis Binos de Pombarat died yesterday at the age of 82.
Who? I never met him. But Yvonne did: she was married to him for about 10 years. And that was over half a lifetime ago. How time flies! Here's the first we have seen of him since he was 40, with his nephew Jules:
|
|||||||||||||||||||||||
|
Alu masala wraps
|
Topic: food and drink, opinion | Link here |
What do I eat for breakfast today? I had decided on something different, alu lap chong. But looking at the recipe, it
was almost the same as the alu masala that failed so spectacularly on Saturday. How about using some of that? And while I was at it, how about the couple
of roti “wraps” that were also left over?
|
|||||||||||||||||||||||
The rest was: cut the cubes more finely, add more tamarind (25 g!), lap chong and peas. And it wasn't that bad. The 25 g of tamarind were a slip of the hand, but they proved not to be too much.
|
Learning about email
|
Topic: technology, history, opinion | Link here |
More reading of “Roll your own mail server”, the book I have to review. Normally a review takes me a couple of hours, but in this case I'm particularly interested, so I'm going through it with a fine-tooth comb, and it looks like taking days. I've been running my own mail server for well over 30 years, but I'm amazed how much I have learnt.
| Tuesday, 12 March 2024 | Dereel | Images for 12 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Film from Ukraine?
|
Topic: photography, opinion | Link here |
Watching a forgettable YouTube clip “20 CAMERAS I'll NEVER Sell” (his capitalization) today:
None were exactly like mine, of course, but he had a Kiev 4, fully 20 years newer than my Kiev 3a, and also a Nikon FM3A, coincidentally 20 years newer than my Nikon FM2. And strangely no Asahi camera at all. And for some reason, he fired the shutter of each to hear the sound.
Firing the shutter on a film camera is an issue, of course. What if there's a film in the camera? So he first turned the rewind knob to check that there wasn't: if there's no film, it will turn freely. If there it, there will be resistance. But with his favourite, number 1 camera, a pre-rangefinder Leica, he didn't. And how about that, there was a film in the camera: when he wound the advance knob, the rewind knob moved too. I couldn't decide whether that was deliberate or not.
How similar is the old Leica to my FED 1? Brought it out and compared. Yes, definitely similarities. My camera has a 50 mm f/3.5 FED lens, a copy of the Elmar, while his has some cheap aftermarket lens. Cocked the shutter. The rewind knob turned! There's a film in the camera! And I had never noticed. I have had the camera for nearly 2 years, but because of this horrible early Leica loading method (from below), I hadn't even checked.
OK, I should take some photos with it, though I'd be much more interested to know if there are already some photos on the film. What kind of film? In particular, what sensitivity? I'd assume somewhere between 50/18° and 400/27°. I'll have to bracket them to be on the safe side.
But there's a problem: the aperture selector is stuck at f/4:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I could have sworn that I had turned it when I got it, and that it was very (too) light in operation, but looking at the photos I took when I got it, it was always set at f/4. What do I do? Clearly I shouldn't force it, and the new photos show that there's quite a bit of dirt in there, so possibly it just needs cleaning. Or I could try to borrow a lens, just for the one film.
And the old Leica as his favourite film camera? No, just joking. His favourite is the Nikon FM3A.
|
More email review
|
Topic: technology, opinion | Link here |
Spent a lot more time today reviewing “Roll your own mail server”. It's getting a lot more nitty-gritty now, but I still don't have a good feel about exactly what I should do. It's too early to criticize; for that I need to find my way through the book. Maybe I should write a cookbook.
| Wednesday, 13 March 2024 | Dereel | Images for 13 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
FED 50/3.5: Bug compatible
|
Topic: photography, opinion | Link here |
Am I the only person who has had problems with the aperture setting on his FED 50/3.5 lens? It's a copy of a Leitz Elmar, so how about looking for issues with that lens?
Yes! Many reports, though generally they were just stiff, not jammed, like this one. Suggestions that oiling would to the trick. Tried that, with no immediate effect, and then found a YouTube clip:
It's not very good. No sound at all, and the explanations are marginal. Also, I probably need tools that I don't have. Still, it's worth thinking about.
|
More ryoms
|
Topic: technology | Link here |
Once again spent a lot of time reviewing “Run Your Own Mail Server” (which I keep feeling should be called “Roll Your Own Mail Server”). I'm making good progress, less so with my accompanying cookbook.
|
More bloody CAPTCHAs!
|
Topic: technology, health, opinion | Link here |
I've been bombarded with a lot of things to do for next week's cataract operation, and even more SMSs reminding me to fill things out. Some of them make no sense at all:
Please take a RAT test on the previous day, take a photo of the negative result along with your driver license, and then a screen shot to confirm the time.
What's wrong with this?
Particularly the time is a problem. The target audience would presumably take a photo with their mobile phone, which includes the time in the Exif data. Of course, many people can manipulate the time. But screen shots on my computer don't have any relationship to the time they were taken. And even if you take a screen shot of a photo on your mobile phone (why?), you can change the name. This whole thing makes no sense at all to me.
Of course, it also assumes that you have access to a RAT test kit, but that's reasonable.
And then there's yet another online form to fill out, 45 minutes they say. Sign up with their usual password restrictions. Are you human? Please fill out this CAPTCHA!
SCREAM! What earthly use would it make for a bot to fill out the form? And why do they submit people with visual problems to this insult? Sent off an email to the reception saying that I don't do CAPTCHAs, and asking for an alternative. Yes, call this help line...
Why do people do this?
|
autocutsel
|
Topic: technology | Link here |
Callum Gibson came up with a new (to me, anyway) program today: autocutsel. I haven't investigated in detail, but it claims to keep the cut buffer and clipboard in sync. It's not quite clear whether this is a good idea, but until I understand the alternatives, it could be a good approach.
| Thursday, 14 March 2024 | Dereel | Images for 14 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Registering for operation: the pain
|
Topic: health, technology, opinion | Link here |
Yesterday I vented my spleen (which on the face of it was not involved) about the appalling requirements to register for next week's cataract operation. But there was nothing for it: it had to be done.
Are they less obnoxious towards Microsoft users? Tried it on distress. Yes, they still have a CAPTCHA, but it's just a matter of ticking the “I am not a robot” box—twice on different occasions. On the first occasion it sent me a 6 digit PIN, not to prove that I had entered the mobile phone number correctly, but to “authenticate” me! And to give me time to think, it waited up to 20 seconds before responding to every input.
And what a pain these forms are! They ask all sorts of questions to which they already know the answers, and I have to answer them in the format that they decide. It took me 4 attempts to enter my address. The first time it simply cleared the field again. The second time I ticked the “can't enter address” box. Ha ha, only joking, you have to enter it. Third time it came up with clever alternatives to “Stones Road”, all starting with “St Albans”. But on the fourth attempt it got bored and allowed me to enter the correct address.
After about 40 minutes, I was finally done! Ha, ha, no, that's just the first half of the questionnaire. Now your medical history, please. When did you stop smoking? August 1975? Wrong!. You need to give the exact date!
Finally I was done, and I had to digitally sign the document. OK, I'll bite, how? It sent another PIN to my mobile phone (number provided earlier in the ordeal), which I had to enter. What kind of children are they to think that that proves anything at all?
Why do people still do this? I had a very positive impression of the facility up until now, but these forms have raised bureaucracy to a new level. An obvious example would be my COVID-19 vaccination status, which should be available at Services Victoria. But the pandemic is over, so they're making it really difficult to access that information.
Still, finally done. Oh.
Date: Thu, 14 Mar 2024 05:21:10 +0000
From: myadmission@sjog.org.au
Subject: Online admission - next step is ready
That's enough pain for one day. Mañana. As it is, it took all afternoon, and I had no time to look at “Roll Your Own Mail Server”.
|
Gaza Genocide: the other view
|
Topic: politics, multimedia, opinion | Link here |
After watching the Al Jazeera video series about Al Nakba in January, I went looking for something that described the situation from the other side. I was only partially successful with some videos by Asher Susser late last month.
But now the ABC has produced a programme titled The Forever War, an interview with a number of Israeli politicians and activists, most prominently Ehud Barak. The usual Western gloss over the problems? By no means. The fascinating thing is that many of the people agree with me: Yehuda Shaul, a former IDF commander, states “Bibi is a terrorist, and that it's not possible to destroy Hamas. Nathan Thrall, an “author and essayist”, states that if the USA were to stop delivering arms, the “war” would be over. But most interesting was what Barak said: “Bibi deliberately strengthened Hamas to keep them in conflict with the Palestinian Authority”.
The programme has a few rough edges, but I'm amazed about how the content matches my own assessment. Now when will the rest of the world sit up and pay attention?
|
CAPTCHAs: Not the site's fault?
|
Topic: technology, opinion | Link here |
I've had a remarkable number of particularly irritating CAPTCHAs lately, and I've expressed my displeasure in no uncertain terms. But am I maybe pointing the blame in the wrong direction? In every case, it happened with a modern browser running on hydra, usually firefox. I'm not sure if it happened with chromium as well. But with firefox on Microsoft, things worked with minimal annoyance. Can it be that there's something about the behaviour under FreeBSD that rings alarm bells somewhere in the innards of something?
| Friday, 15 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Fire recovery by Facebook
|
Topic: general, technology, opinion | Link here |
Yvonne told me today that the CFA was holding a “fire recovery day” today, 4 hours at the Dereel Hall. I can do without that, but why wasn't I informed? Ah, only on Facebook, assuming you're subscribed to the right group.
|
Configuring daily cron jobs
|
Topic: technology | Link here |
FreeBSD runs a number of cron jobs every night to check for things that could be problematic. The jobs themselves are in the /etc/periodic/ hierarchy. One of them (/etc/periodic/security/110.neggrpperm) produces this output:
Checking negative group permissions:
find: /VB3/oldbackups/eureka-FreeBSD/Downloads: No such file or directory
141107070 -rw----rwx 1 grog lemis 313303 Oct 11 09:03:38 2017 /home/grog-eureka/DxO/Modules/C14933a_000.caf
141107071 -rw----rwx 1 grog lemis 4713293 Aug 23 14:48:40 2017 /home/grog-eureka/DxO/Modules/C15101a_000.caf
141107087 -rw----rwx 1 grog lemis 4745060 Aug 23 14:48:48 2017 /home/grog-eureka/DxO/Modules/C19627a_000.caf
...
What's that? The issue, if there is one, is that the “other” permissions, the last three letters in the mode output and here “rwx”, are more permissive than the previous 3 “(---)”.
That doesn't make any sense, but is it a security threat? Not that I can see. In this case, it's DxO PhotoLab lens correction modules, firmly entrenched in the Microsoft space, and for some reason Samba maps the permissions like that. The worst thing about them is that they get in the way of potential real error messages. Yes, I can just remove the file, but it turns out that there's a file /etc/periodic.conf that allows you to configure the jobs:
--- periodic.conf 2022/05/12 08:31:23 1.1
+++ periodic.conf 2024/03/14 23:13:09
@@ -268,7 +268,7 @@
security_status_chksetuid_period="daily"
# 110.neggrpperm
-security_status_neggrpperm_enable="YES"
+security_status_neggrpperm_enable="NO"
Aaaah, that's better.
|
Surgery: the other shoe
|
Topic: health, technology, opinion | Link here |
I'm still not done with my webwork (virtual paperwork) for my cataract operation next week. When I thought it was done, I got a mail asking me to perform the next step.
OK, dammit, get it over and done with. “Please enter a credit card number for the gap payment”. That's unusual, but OK. And it went relatively quickly.
And they deducted the money immediately!
Your payment was successful, a payment confirmation will be sent via email.
Please complete your credit card details. If you selected 'Allow Automatic Charging of Incidentals’ your Credit Card details will be stored securely and only accessed if you have further incidentals payments after your hospital stay.
Please do not refresh this page, it may result in a duplicate transaction.
Savour that “please do not refresh”. Their software is so bad that they can't recognize duplicates! That was a basic tenet of the Tandem Operating System 50 years ago, even before they changed its name to “Guardian”.
|
Larissa in pain again?
|
Topic: animals | Link here |
This evening, we called Larissa for her daily meal. She didn't get up. She moved a little, and then apparently gave up. It wasn't until we lifted her into position that she finally walked off and ate her food.
Why? It's not the first time that something like this has happened. Is she having difficulty coordinating her legs? Yvonne thinks that it might be related to her being on the tiles, where she slips when she tries to get up. We need to keep an eye on that, maybe a video of it next time it happens.
| Saturday, 16 March 2024 | Dereel | Images for 16 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More garden work
|
Topic: gardening | Link here |
Jesse Walsh along today again, this time with different tasks, notably pruning the bushes, notably the Buddleja × weyeriana that has grown out of all proportion.
|
Putting ryoms to the test
|
Topic: technology, opinion | Link here |
I'm about half way through my review of “Roll your own mail server”, and I'm running into trouble. So far I've been reviewing with my own knowledge of the material, but how he's moving into areas that I don't know. That's good, of course: that's why I'm interested in the book. But how can I review this content? I had thought of moving through more quickly, but I really need to try this stuff out for myself.
OK, what do I need? As the author recommends, two systems that initially talk only to each other. Fine, revive tweedledum.lemis.com and tweedledee.lemis.com. Set up virtual machines.
OK, where? hydra is the new machine, of course, but I haven't configured VirtualBox for it so far. It proves that I had already installed the software. How hard can it be?
Oh. The GUI has changed, of course. After working my way through that, I discovered that I needed at least to tell VBox where the files are. Did that, but it didn't find anything.
Oh, of course, I have to add the machines. Why? No idea. And not much help. Go to Tools, click on the green + symbol, climb down the directory tree to find a file with a name like eureso/eureso.vbox.
|
|||||||||||||||||||||||
Click Start. Oh:
|
|
|||||||||
Clearly it's too polite to complain, but if I ask for Details, I get:
Result Code: NS_ERROR_FAILURE (0x80004005)
Component: SessionMachine
Interface: ISession {c0447716-ff5a-4795-b57a-ecd5fffa18a4}
What does that mean? The web knows, lots of conflicting reasons. The one that matches is that I wasn't in the vboxusers group. OK, add that and try again. Same message, but with different details:
Could not launch the VM process for the machine 'eureso' (VERR_ACCESS_DENIED).
Result Code: VBOX_E_IPRT_ERROR (0x80BB0005)
Component: MachineWrap
Interface: IMachine {85632c68-b5bb-4316-a900-5eb28d3413df}
What does that mean? Maybe the web knows that too, but what I found suggested that I had a VM that was too old, and that's possible.
OK, why am I using VirtualBox anyway? That's an old, worn-out virtual environment, I'm told. The new one is Bhyve. Read up on how to install that at FreeBSD as a Host with bhyve (just below FreeBSD as a Host with VirtualBox), both well hidden in the Virtualization chapter, and not helped by the web, which didn't include that link, just the Wiki page with mainly FAQs, the first of which is Is it "bhyve", "Bhyve", "BHyVe" or BHyve?.
Oh. It's non-trivial, and it leaves me wondering whether I shouldn't stick with VirtualBox on eureka, at least for current purposes. The trouble is that I have migrated all the VM files from eureka to hydra. Once again logistics are getting in the way of doing anything.
|
Microsoft does it again
|
Topic: technology, opinion | Link here |
It's middle of the month, past the time where Microsoft crashes and restarts distress and somehow manages to stop my rwhod from running. They're getting more cunning all the time. In the past I have had to reinstall the program, but recently I haven't even been able to do that: the file was marked immutable or whatever Microsoft calls it, and I couldn't delete the old one. I had to start as “Administrator” and manually remove it before I could reinstall it.
But even that doesn't work any more. First, I can no longer start
a shell COMMAND.EXE as Administrator. Have they changed something, or
have I forgotten something? Something about “right click”, but not the way I remember it.
A quick web search came up with 13 ways to use "Run as administrator" in Windows 10. Thirteen! Is that an unlucky
number?
But even as “Administrator”, I can't delete it! But I can rename it to an appropriate expletive. What a pain this Microsoft is!
|
Kangaroo visit
|
Topic: animals | Link here |
It's been a while since we've had a kangaroo come visit:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
| Sunday, 17 March 2024 | Dereel | Images for 17 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
New “Laksa” noodles
|
Topic: food and drink | Link here |
Penang laksa for breakfast today, but no noodles in the freezer. Where are the fresh ones? I didn't find them, only this kind:
|
|||||||||||||||||||||||
OK, the eternal question with a new kind of noodle: how long do you cook them? These instructions were interesting: soak for 15 minutes, then boil them for 5 minutes. That makes more sense in light of other noodles that just want to be soaked, like the Twenty-twenty noodles that I had actually been looking for. OK, try it out. Yes, for once it works. But somehow they don't seem to be laksa noodles, just ordinary rice noodles. That's probably because they're from Singapore, which already has a very different kind of laksa.
|
Reprocessing old photos
|
Topic: technology, photography, history, general | Link here |
Ten years ago today I drove along the Great Ocean Road with Jörg Micheel and his son Richard. Lots of photos, of course, which I processed with the means at my disposal in those days.
Can I improve on things? The biggest difference is the “optimizer”. In those days I used the Ashampoo optimizer, which frequently improved things, but left something to be desired. In November 2021 I switched to “Perfectly Clear”, which, though discontinued, seems to be better.
OK, I already have a script for reprocessing old photos, and it works most of the time (very old directories have different structures). Part of the script generates a web page that compares the images before and after optimization. But what about the difference between Ashampoo and Perfectly Clear? Today spent a surprisingly short period of time modifying the script for a three-way comparison. And it seems that this reprocessing was a particularly good example. I had noted at the time that the extreme lighting had caused problems, particularly with the koalas. Here the original, the results with Perfectly Clear, and the results with Ashampoo (run the cursor over the first or third image to compare it with Perfectly Clear):
That's a particularly clear example for the improvement.
|
More VirtualBox setup
|
Topic: technology | Link here |
Yesterday I established that I couldn't run my old VirtualBox VMs on hydra, at least not without changes that I didn't want to make. What about tweedledum? No, it didn't want to know about that either.
OK, run on eureka. Install a standard FreeBSD 14.0-RELEASE system, even with ZFS, and as close to default as made sense. The only changes were to enable NTP. Everything went well, but of course the network interface didn't work: the default network is NAT, and what I want is a bridged network. I've seen this before, but I now wonder why these defaults just don't work.
OK, shut down the VM, set the network parameters. Oh. It only offered me the “choice” of interface re0. But eureka doesn't have an re0, just xl0 (which I don't want, and which in the past has been its preference) and em0 (which I do). Where did re0 come from? That's on hydra! But the VirtualBox is running on eureka! And I couldn't start it again, with a meaningless error message that looked like the one I had seen before.
Dammit, does nothing work? Yes, I had started VirtualBox to display on hydra, but that's no reason for it to look at anything else on hydra. So I need to go back to eureka and see if I can get it to work there.
| Monday, 18 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Virtual machines for ryoms
|
Topic: technology | Link here |
More playing around with virtual machines for “Roll Your Own Mail Server” today. Strange things happened when I ran VirtualBox on eureka but displayed on hydra. What about displaying on eureka?
Yes, almost. Still this issue with the kernel modules. There are three of them, but vboxnetadp was missing. In the past I've had issues with the sequence of loading, but today I found the solution: load only vboxnetadp, and it will load the other two, and, it seems, also ng_ether.
And, to my surprise, things mainly worked, modulo a couple of surprises. The network wasn't set up correctly, presumably because of the issues getting it to work when I installed it, so I had to set up /etc/resolv.conf and the host IP address and default router in /etc/rc.conf. But after that I was able to install bash, Emacs and xterm, allowing me to access the machine relatively easily.
Next, install the packages that the author wants:
[root@tweedledum /home/grog/ryoms/Logs]# Log pkg install postfix mutt dovecot rspamd mariadb apache
pkg: No packages available to install matching 'mariadb' have been found in the repositories
pkg: No packages available to install matching 'apache' have been found in the repositories
Ah, of course, it's apache24 and mariadb106-server and mariadb106-client. Another of my favourite gripes about the Ports Collection.
But then it installed, and quickly. What a surprise! Now I need to find out how to set up X.509 certificates and I can set up my test servers.
|
If it's broke, don't fix it
|
Topic: technology, opinion | Link here |
tweedledum is my first “machine” with ZFS. It installed a default set of file systems:
[root@tweedledum /home/grog/ryoms/Logs]# <b>df -c</b>
Filesystem 1K-blocks Used Avail Capacity Mounted on
zroot/ROOT/default 61665232 1852404 59812828 3% /
zroot/tmp 59812932 104 59812828 0% /tmp
zroot/home 59812964 136 59812828 0% /home
zroot/var/audit 59812924 96 59812828 0% /var/audit
zroot 59812924 96 59812828 0% /zroot
zroot/var/log 59813004 176 59812828 0% /var/log
zroot/var/crash 59812924 96 59812828 0% /var/crash
zroot/usr/ports 60633592 820764 59812828 1% /usr/ports
zroot/var/mail 59812952 124 59812828 0% /var/mail
zroot/usr/src 59812924 96 59812828 0% /usr/src
zroot/var/tmp 59812924 96 59812828 0% /var/tmp
total 660615165 2674188 657940977 0%
OK, I'll bite. I assigned 64 GB of disk (tweedledum.vdi), which correspond to 2 GB on disk. But df is telling me that I have 660 GB!
Clearly it's lying. The Used column is probably correct; the total size and Avail are clearly wrong. So I mentioned it on IRC, and once again I got responses that blow my mind.
Naysayer: Unix/POSIX doesn't have the concept of file system(s) sharing a pool of storage
Naysayer: it would also fail if you use (say) PEFS
Naysayer: or probably any other layered file system
Groggy: If it fails, then it's broke.
Naysayer: what about other overlay file systems?
Naysayer: seems like you just want ZFS to not exist with the weak excuse they didn't magically fix POSIX brain damage
Naysayer: one I use: PEFS
Groggy: Where's the brain damage?
Naysayer: /local0/crypt on /local0/crypt (pefs, local, nosuid, mounted by radar)
Naysayer: [patest 4:48] ~ >df -c /local0 /local0/crypt
Naysayer: Filesystem 1K-blocks Used Avail Capacity Mounted on
Naysayer: /dev/gptid/523c6d2b-ca8f-11ec-b6c5-94de8081be56 395648196 1290052 362706292 0% /local0
Naysayer: /local0/crypt 395648196 1290052 362706292 0% /local0/crypt
Naysayer: total 791296392 2580104 725412584 0%
Groggy: Right, that looks broke too.
Naysayer: but I'm not taking on responsibility for fixing kernel interfaces so du can learn what file systems are backed by others
Groggy: So: looking at a ZFS collection of file systems.
Groggy: In any case, presumably these pools, or whatever they're called, belong to a single group or a small number of groups.
Groggy: So if I allocate something out of one of them, there's less left for the rest.
Naysayer: "hey I wrote an encrypted overlay file system, isn't it neat?" "Go away because it breaks df -c"
Groggy: No, "df -c can't handle it. It needs to be fixed before we can put it in the tree".
Naysayer: but in saying that you are pushing extra work on the submitter
Naysayer: and in this case, a _lot_ of work
Naysayer: then you have some non standard API
Groggy: PEFS and ZFS are the non-standad API.
Naysayer: it is a question the POSIX committee should solve
Groggy: POSIX isn't the issue.
Groggy: We don't stick to it now, and never have.
Naysayer: it _IS_
Somehow I'm not getting across that to import a component into a system, it needs to be compatible. And here it clearly isn't. The silly thing, of course, is that this really has nothing to do with POSIX, which doesn't define the -c option. But nobody seems to think that it's necessary to fix anything. And that's sad.
What's needed? Clearly df is getting the wrong information, which might be ZFS' fault, or (as Naysayer implies) it might be a kernel interface restriction. The Avail seems to show the real size of the pool available to the file system, including the space already used by any member of the pool. The Used seems to be correct. And the first column, the “1K-blocks" below, seems to be the sum of the two, so Used gets counted twice.
That's ridiculous, of course, and it needs to be fixed. It also has nothing to do with the -c option. But it's not half as ridiculous as “it's too hard, let's forget it”. Or, as Callum Gibson said,
Groggy: would have us remove all offending filesystems
Is that really so hard to explain? He did, though, point to another issue: NFS-exported ZFS file systems have the same issue, so it's not as simple as having df look at the file system type and acting accordingly.
And, of course, this isn't the only place where it's possible to count storage more than once. One of my favourites is this:
$ du -sc D/ DL/
66306 D
38473 DL
104779 total
# du -sc DL/ D/
104779 DL
1 D
104779 total
I asked for an explanation, but Naysayer carefully avoided the question.
| Tuesday, 19 March 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Dreary day
|
Topic: general, health, opinion | Link here |
Somehow I didn't do anything all day. I wasn't feeling the best. Was that the concern about the upcoming cataract operation, a change in the weather, or just the fact that I'm gradually getting (very) old? I don't think it's the operation itself. That doesn't concern me, and I'm sure that the results will be worthwhile. But the recovery phases, two weeks of candidacy for king of the land of the blind, gives me some concern. Still, to go by what bad dreams I have, that's not as much of a concern as the plight of the population of Gaza. One way or another, as a result I did little of interest today.
|
Emacs crash
|
Topic: technology | Link here |
While doing something non-spectacular today, Emacs instance on eureka stopped for no apparent reason. That's unusual. Normally my Emacs processes run for months.
But only a little later, another one stopped, this time with a But even more unusual was that another one crashed, this time with a signal 6 (SIGABRT):
Mar 19 13:32:11 eureka kernel: pid 4734 (emacs-24.5), uid 1004: exited on signal 6 (core dumped)
If it had been hydra, I would wonder about hardware issues, but eureka has been running reliably on the current hardware for over 10 years. Coincidence? Should I try to debug the core dump? I fear that that way madness lies.
| Wednesday, 20 March 2024 | Dereel | Images for 20 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Monthly garden photos
|
Topic: gardening | Link here |
Today was the March equinox, time for my monthly garden flower photos.
Parts of the front garden are looking less than happy. The newly planted oak has decided that it's autumn, and the Japanese cherry doesn't look much better::
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The other oak closer to the arena looks happy enough, so I assume that there's some more irrigation work to be done, as the lawn also indicates:
|
|||||||||||||||||||||||
Other trees in the area also look less than spectacular:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
On the other hand, the Corymbia ficifolia has simply finished flowering:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And the Hibiscus rosa-sinensis “Uncle Max” seems to have recovered from its freezing nearly 2 years ago:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The transplanted Camellia japonica still looks anything but happy, but there's hope:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Those completely green leaves are something new. Hopefully it will continue to recover.
The Abutilons are both looking much better than in the past:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Clearly adequate watering makes a big difference.
The Banksia serrata is now flowering again:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Presumably that's just a question of the season.
We've had our lime tree for so long that I don't know when I bought it. It could have been in August 2008, in which case it's a Tahitian lime. And it has seldom had much in the way of fruit. Until now:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
It's still not overly abundant, but more or less enough for us in the season.
I haven't seen these bulbs before:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I think that they must be survivors of the plants that Jane Ashhurst sent us. For reasons I don't understand, just about all of them died, but I recall that some were supposed to flower like that.
This sad-looking lawn is the result of a glyphosate attack by Yvonne:
|
|||||||||||||||||||||||
To make up for it, the compost heap is growing happily:
|
|||||||||||||||||||||||
|
Preparing for surgery
|
Topic: health, technology, opinion | Link here |
Tomorrow's the date of my cataract surgery. When do I need to be there? Hopefully not in the early morning, but they haven't told me so far. But there's one more thing to do:
All patients having elective surgery will have to undertake a Rapid Antigen Test the day prior to scheduled surgery. Patients will need to show proof of a negative test result when arriving at the surgery centre. It is suggested to take a photo of the test result next to a drivers license or medicare card and then take a screen shot of the photo which will capture the date and time of the result.
I've already grumbled about the stupidity of this requirement, but OK, if it makes them happy. How can they tell that I have really taken the test? Anybody could have done it. And when? My screen shots have time stamps that say when the shot was taken, not when the photo was taken. OK, take the photos, process them. And the screen shot? No problem, but of course it's on eureka. And they want me to show it to them.
How? Two possibilities: print it out, or download it to hirse, my mobile phone. Both are possible, since I have an FTP client for the phone. Did that. Where are the screen shots? “Camera” shows photos, but clearly they're something completely different. What does Google say? Dozens of different answers, many requiring installing an app!
Finally I found an answer: use the “File Manager”, a bare-bones ls substitute. It doesn't understand locations, and it groups files by category, in this case Images, which includes both screen shots and photos:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And yes, when you select them, it displays them with—as far as I can tell—no details whatsoever beyond the modification timestamp. Once again I'm amazed at the uselessness of these apps. But I have them, and I'll take a printout and the dead RAT to be on the safe side.
Finally, at 14:25, I received my information, per SMS!
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
OK, I'll bite. When? Called up Preadmissions, who told me that the time should be at the top of the SMS. No. OK, I'll connect.
Long wait; I thought it would time out. Connected to Steph, who apologized for the delay in sending the SMSs, but it should be there soon. After I made it clear that I had received the SMS, and that was the problem, she told me that she has a time of 12:30, but she would connect me to the surgery. This time it did time out, and I came back to Steph. OK, try the surgeon's rooms. Another time out. OK, I'll connect you to Preadmissions. But that's where I came from!
Back to Preadmissions, where they apologized that the SMSs hadn't gone out yet. After I made it clear that I had received the SMS, and that was the problem, she told me that she has a time of 12:30.
OK, that seems final. Only 13 minutes on the phone! Back to check things, and discovered that the message displayed had been truncated! The real SMS was 3 times as long, and it included all the data, but for that I had to go elsewhere in the menu maze. It would be nice to be able to save it, but I still haven't been able to find anything that will do that for me. I expected that Microsoft Phone Link would do it. That displays the image on a Microsoft Window, and it allows me to select the text (it would be too simple just to store it, right?), but then I can't do anything with it, not even store it in another window: C-v merrily stores the last text from the X cut buffer instead.
Installed a nameless app that wanted to copy my SMSs to Google Drive, but not store them on any local machine. Tried again and got something that at least sent one test SMS to my email address:
Incoming - +1234567890 (Contact undefined) <br/>Message: Test (20/3/24 16:23)
You received this letter because you set up SMS forwarding from your phone via the "SMS to phone / mail - auto redirect application" (https://play.google.com/store/apps/details?id=com.gawk.smsforwarder)
If you don't want to receive such emails anymore - Unsubscribe: https://sms.forwarder.api/email/unsubscribe?email=groogled@gmail.com&token=X0eXG71IHkiqSPjFK2GUquQlDCY2a38UVVdrqf5GlghvizDax9YgZ
This corresponds to an SMS with the single word “Test”. I haven't found out how to forward existing messages. Is there nothing better?
|
Bird's nest
|
Topic: animals | Link here |
Yvonne found this on the ground a couple of days ago:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What is it? A bird's nest, of course, but what kind? The red feather suggests some kind of parrot, but it's too small for any kind that I know. And it's amazingly light, only 1.42 g.
| Thursday, 21 March 2024 | Dereel → Geelong → Dereel | Images for 21 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Eye rebuild
|
Topic: health | Link here |
To Geelong to have my cataract surgery and vitrectomy. Everything went surprisingly smoothly. The appointment was at 12:30, but I arrived a few minutes early and was whisked away for preparations almost immediately. Various relevant questions, forms filled out, a second nurse to check all the details, confirm the correct eye and mark it:
|
|||||||||||||||||||||
Somehow I find it irritating that the arrow is off-centre.
Then she subjected me to two sets of eye drops, the first of each of which stung, after which David Fabinyi came in and confirmed yet again that we were doing cataract surgery and vitrectomy on the right eye, though in January we had talked of the left eye first.
Then I was wheeled in to the anaesthetist, who put in a particularly painful catheter into my hand, covered me over, filled me with anaesthetics, and off they went. I was conscious the whole time, and I expected something interesting, but all I saw were a few strange geometrical patterns, and I felt a couple of pricks. And almost before it began, it was over, at 14:00. Out for a set of sandwiches and a coffee, advice on what to do next—keep the covers on the eye until tomorrow morning, when I should come back to have my eye looked at again. And by 14:45 we were out and on our way, after only 2 hours, 20 minutes:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
As Yvonne said later, at that time after being admitted to Epworth Hospital two months ago, she was still waiting to be attended to.
And the results? No pain, not even itching. As advertised, there's a bubble in the eye, and as I move I can see the liquid moving backwards and forwards. Somehow it's less colourful than I expected. Probably the most irritating thing was the pad over my eye (an iPad, if I misunderstood the nurse correctly), which I needed to keep on until tomorrow morning, and which made it difficult to put my glasses on the good eye.
|
Understanding Android displays
|
Topic: rant, technology, health, opinion | Link here |
Yesterday I spent an inordinate amount of time complying with the hospital's silly RAT requirements, notably a screen shot of a photo of the device.. Yes, I did that, and I also brought the device itself.
When I got there, one of the first questions from reception was "Did you do a RAT?”. Yes. OK, then, go on through!
Ha ha, only joking. The next nurse wanted to know, but she was happy with the device, so all the messing around with Android was unnecessary. But independently of the operation, I did come to one discovery: yes, Android's file manager does distinguish between “photos” and “screenshots”. I just didn't see it because of the appalling display layout: first “photos”, then most of the screen full of advertising, and squashed at the bottom a couple of scree shots trying to make themselves seen:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And what about the dates of the photos? They're all JPEG with Exif data. But it takes a lot
of smashing swiping to find any
details—even the file name—and then I get something like this:
|
|||||||||||||||||||||||
What it should really be saying is something like this:
File Dead-RAT-3.jpeg
Date taken: Wednesday, 20 March 2024, 13:18:10
Exposure: 1/250 sec, f/5.6 (EV 12.9), 200/24 ISO
Camera: Olympus OM-D E-M1 Mark II, serial BHUA20626
Lens: Olympus M.Zuiko Digital ED 12-100mm f/4 IS PRO
Focal length: 41.0 mm (full frame equivalent: 82 mm)
Focus: S-AF+MF 0.535 m (0.52 - 0.55 m)
Crop factor: 0.51 (0.63 horizontal, 0.80 vertical)
Effective FL: 58 mm (FF: 117 mm)
Field of view: 15.1 horizontal, 14.5 vertical, 20.8 diagonal
Meter mode: ESP Manual
Stabilization: Body
Size: 3259 x 3126 pixels (10.19 megapixels)
Copyright: Greg Lehey
Not only is it ridiculously superficial, it's wrong. It claims that the photo was taken at 16:34, politely omitting the seconds, when the Exif data clearly states that it was taken at 13:18:10. And no, it's not a case of stripping the Exif data: I retrieved the image from the phone and ran Exiftool against it.
In addition, I have a number of such images. How do I tell them apart? By name, of course, but for that I need to select each image and then select the data. That would take for ever. It's easier, once you get a sensible file copy tool, to download them to a Real Computer and look at them there.
| Friday, 22 March 2024 | Dereel → Geelong → Dereel | Images for 22 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
A new eye
|
Topic: health, opinion | Link here |
Off to Geelong this morning into a bright rising sun that made the drive a misery. To add insult to injury, the windscreen was dirty and the windscreen washer empty. Stop at a service station on the way? On the entire journey, 79 km, there was exactly one service station, in Inverleigh. But it was on the wrong side of the road, and Yvonne didn't want to go there.
At the surgery, I was immediately assigned to a nurse who told that I should have taken the covering off my eyes last night. She took off the bandages. Brilliance! I was expecting things to be fuzzy for a while, but no, everything looked sharp.
Well, sort of. I knew about the air bubble at the bottom of my field of vision, but there were various other artefacts: a flickering in the middle of my field of view, small, mobile black floaters that are apparently droplets of blood which will go away. And she gave me an eye test chart that I could barely read. So she gave me a little device with small holes to look through, after which I could read them better, but still not well.
Then in to see David Fabinyi, who told me that everything went well from his point of view, and we made an appointment to do the other eye on 11 April, which I thought had already been made. An appointment with his locum a week before to confirm that everything had gone according to plan. I forgot to mention the (to me) disappointing results of the eye test chart. I must write these things down for next time.
Still, a couple of strangenesses. I was surprised that I could see my outstretched hand so clearly. Ah, that's because the lens focus is set to slightly shorter than infinity, and he has chosen a lens with particularly good depth of field. When I first saw him, I talked about depth of field, and at the time I got the impression that he hadn't heard the term before. Clearly there's no difference in depth of field from one lens of the same focal length and aperture to the next. And it's not clear to me whether “slightly shorter than infinity” relates to a hyperfocal distance (probably a term he still hasn't heard) or not. And for some reason he thought that I might no longer need multifocal lenses in my glasses. That puzzles me.
On the way home I discovered that yes, I could see the dashboard very clearly, but not the road signs. Wrong lens? When I got home I heard from Daniel O'Connor that his father had had a cataract operation, and that he needed to have it “redone”—no further details. Hopefully they haven't put in a too powerful lens.
At home, the impression continued. I couldn't clearly read text on the TV, 5 m away, but I could read things that were closer. Is this a problem, or will it go away with the bubble? By the evening things looked better, so maybe it will. In any case it confirms the validity of my decision to have the right eye done first.
And the bubble in itself is interesting. I discovered that I can see clearly through it when looking straight down. But the focus is different: it seems to be about 10 dioptres stronger than the aqueous part, effectively giving me a built-in magnifier. It's almost a pity that it will go away.
So what's it like? I was expecting more sharpness, of course, but it really hit me in the face. The colours were also more vivid, and the old left eye has a yellowish tinge that I hadn't expected. I need to find a way to take some photos showing “before” and “after”. Certainly I can read LCD displays easily now, something that had been becoming increasingly difficult.
And the eye drops? Four times a day, and they're apparently the same kind that I got before the operation yesterday, the ones that sting considerably. Still, it doesn't last long.
|
Google Maps from the worst side
|
Topic: technology, opinion | Link here |
I've been using Google Maps on almost every trip in the last couple of years, and they often work well. But yesterday they identified congestion on the way into Geelong, and took me straight through it anyway, taking 5 minutes longer than their own estimate.
|
|
|||||||||
OK, there are things we can do about that. Go via a diversion:
|
|
|||||||||
That all went fine up to a few hundred metres before the waypoint, when it became clear that Google Maps had no intention of taking me any further. So I had to ask it to take me to the correct address—fortunately I remembered it. “Exit Navigation? Yes or No?”. No, of course, I wanted to continue. Wrong answer. To navigate the new route, I first needed to “exit” the old route. By the time that was sorted out, we were way past the non-obvious turnoff:
|
|
|||||||||
The whole effort took 5 minutes longer than going through the congestion.
| Saturday, 23 March 2024 | Dereel → Ballarat → Dereel | Images for 23 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Seeing more clearly
|
Topic: health | Link here |
My concerns about cataract surgery at the beginning of the month proved to be accurate: the surgery itself was not an issue—even less so than I expected—but I have the problem that I can only see clearly with one eye at at time. With glasses it's the left eye. Without glasses it's the right eye, and then only at a distance.
The solution is obvious: remove the right-hand lens from my glasses. But how? The web has lots of solutions that appear to apply to other kinds of glasses. For plastic, heat it and remove the lens. For metal frames, which is what I have, unscrew the frame and remove the lens. But my glasses don't have screws. I know that cats and dogs are quite good at removing lenses, especially if you subsequently sit on the frame, but though I'll never need the lenses again, I found that a little destructive.
OK, call up Specsavers. Sure, come on in, we'll do it immediately. So off before breakfast, where they took the glasses off and disappeared for a surprising amount of time—about 5 minutes—before returning with half glasses:
A good thing that I didn't try to do it myself. In that time, Yvonne, who arrived later after parking the car, had her glasses adjusted and plenty of time to wait. I wonder what the trick is.
Back home, restrictions became clear. The view of the two eyes was offset vertically. I had to put the glasses on at an angle to compensate for it. I assume that that's normal: glasses change the vertical angle of view slightly. As long as there are lenses in both sides, that doesn't make a difference, but with only one lens it's irritating.
And, of course, things didn't help when I was at the monitor: my new eye can't focus that close. So I needed a fake Moshe Dayan after all:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But things are definitely looking up.
|
Simulating cataract surgery improvement
|
Topic: health, technology | Link here |
How do I describe the change in vision after cataract surgery? Obvious: tap into my brain and download the images.
But I don't know how to do that, so the next best thing is to take some photos and process them as “before” and “after”. Here a few attempts (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The important thing here is that I was previously unable to see whether the door was locked (lock vertical) or open (diagonal, as here).
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And here the car dashboard. Previously I could barely recognize the needles on the dashboard, and I had extreme difficulty reading the LCD display. This photo probably comes closest to showing the difference.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I'm not sure that this doesn't understate the difference. But I could read the LCD display before, so maybe it's relatively accurate.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And this seems reasonable. In bright sunlight I could see things relatively well, though things like the trees in the background were hard to distinguish. And the colour cast is about right.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
This one was more a coincidence. I went into the bedroom with my good eye covered over, looked for Bruno, and didn't see him.
|
Seeing triple
|
Topic: health | Link here |
Some of the more stupid things I have seen (no pun intended) were cartoons and similar showing the effects of drink: people start seeing double. Drink far too much and they see quadruple or worse. That makes no anatomic sense at all, of course.
But now I'm seeing triple! Two different views through my two eyes, and then a third view through the bubble in my right eye, effectively a close-up lens. It's marginally useful, but it also means that I currently can't look downward with that eye.
|
Lens improvements
|
Topic: photography, opinion | Link here |
It's been almost exactly 60 years since I started getting interested in photography. At the time I had a Voß Diaxette, with a 45 mm f/2.8 Steinheil Cassar lens, which looked pretty much the same as this one, which I bought on a couple of years ago:
|
|||||||||||||||||||||||
I only used it until July 1964, but I took a lot of photos that are now of historical interest. But it was a pretty bare-bones camera, and the lens (an f/2.8 Cooke triplet!) was really not good. And yet I bought another one, this time 50 mm, 6 years ago, with an Edixa Reflex. OK, that could be a good lens to include in my cataract photos. Poor image quality? Flare? Here's a comparison with the Olympus Zuiko Digital ED 50 mm f/2.0 Macro, itself over 20 years old (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's really a surprising difference. And no, the lens isn't dirty, and it is coated (at least to the standards of 70 years ago). And though all images were taken af f/2.8 (the maximum aperture of the Cassar), the camera selected a shutter speed of 1/80 s for the Cassar and 1/160 s, fully a stop more, for the Zuiko. Is the transmission of the Cassar really that bad?
| Sunday, 24 March 2024 | Dereel | Images for 24 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More eye recovery
|
Topic: health | Link here |
My eye is gradually improving. The bubble is getting smaller, and a good thing too: it's particularly irritating because I can't look down.
|
Another beef and broccoli recipe
|
Topic: food and drink, opinion | Link here |
Yvonne bought some broccoli when she went shopping on Wednesday. What do we do with it? The only thing that I know that goes with it is beef, and that wasn't on the menu.
OK, reschedule. Beef and broccoli it is. But can we improve on the recipe I have? Off looking on the web. The quality of the recipes is gradually improving, and today I found Made with Lau, which looks interesting. They have a recipe that explains some of the things that have puzzled me in the past. Why baking soda (Sodium bicarbonate)? To tenderize the meat. But then the recipe includes:
1 tsp baking soda (you can also tenderize meat physically by hammering the meat with a meat mallet)
I use the mallet for different reasons, to thin the “sizzling steaks”, but it's good to understand the reasoning.
And 1 tsp baking soda? Yes, these horrible US measurements. But no worry, you can select metric. 4.9 mL baking soda! 454 g broccoli! 1.2 mL white pepper! Unfortunately it doesn't run to using appropriate ways to measure things. measu
So: the recipe calls for carrots, which I chose to ignore. The result was:
| quantity | ingredient | step | ||
| 200 g | beef (“sizzling steak”) | 1 | ||
| 10 g | oyster sauce | 1 | ||
| 10 ml | Chinese rice wine | 1 | ||
| 10 g | sesame oil | 1 | ||
| 10 g | soya sauce | 1 | ||
| 6 g | garlic paste | 1 | ||
| 100 g | broccoli florets | 2 | ||
| oil for frying | 3 | |||
| 6 g | garlic | 4 | ||
| 22 g | spring onions | 4 | ||
| 10 g | oyster sauce | 5 | ||
| 6 g | light soya sauce | 5 | ||
| 6 g | dark soya sauce | 5 | ||
| 10 g | water | 5 | ||
| 10 g | cooking wine | 5 | ||
| 6 g | cornflour | 5 | ||
| chopped coriander leaf | 5 |
Cut the beef into thin strips about 1×3 cm in size. Hammer to get it about 3 mm thick. Mix with oyster sauce, wine, sesame oil, soya sauce and garlic and marinate for at least 30 minutes.
Cut the broccoli into small florets. Blanche in boiling water for 30 seconds, drain and let cool
Fry the beef until cooked, place on a serving dish.
Fry the garlic and spring onions in the oil until about half-cooked. Add the broccoli and stir fry briefly.
Add sauce ingredietns and bring to the boil.
Warm the beef in the serving dish, then place the vegetables on top. Mix together, garnish with coriander leaf and serve.
And how was it? Surprisingly close to the recipe I already had. The proportions of this recipe were off, of course, because of the units chosen. In particular, there was too much cornflour, which I've tried to modify in the recipe above. In many ways I think my old recipe was better. So rather than update that recipe, I'll wait until we want to do it again, and then I'll make a decision.
| Monday, 25 March 2024 | Dereel | Images for 25 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Eye recovery
|
Topic: health | Link here |
It has now been four days since I had my cataract surgery. Things are going completely according to plan, but it's still an interesting experience. The bubble in my eye is gradually becoming smaller, and in the process the magnification effect is diminishing. On Friday it was about 10 dioptres, and now it can't be more than 1.5 dioptres. And the real vision seems to be shifting in focus too. Hopefully by the time the bubble is gone (tomorrow? Wednesday?), the focus will be as intended.
And apart from that, the floaters are interesting too. Initially they looked like flies flying around, but now they're longer and joined. Today I saw one, very thin and in L shape that covered about half the field of view and gradually moved into the bubble. Presumably they'll all go away with the bubble. But still it's irritating, and it's keeping me from doing anything useful.
|
Depth of field? Hyperfocal distance?
|
Topic: photography, health, opinion | Link here |
When I visited David Fabinyi in January, I mentioned depth of field, and I got the impression that he hadn't heard the term before. But it's important for my decision on what kind of vision aids I will need. Of to search the web. Yes, strangely the term doesn't seem to be in use for human optics. What about hyperfocal distance? Also barely a mention. It's not because the concepts don't apply—on the contrary, they're just as relevant. So why?
It seems that there's only one paper on the subject, “The Depth of Field of the Human Eye”, written by Fergus William Campbell in 1959, and only available for significant quantities of money. Then there are discussions like this one, which refer to dead links and come out with claims like “Hyperfocal distance of human eye is 6 or 7 meters in indoor light”. That would correspond to a depth of field of about 12 cm at 60 cm distance. Could that be right? Elsewhere I saw a similar claim, except that it was in feet. So somehow I'm no closer to an answer.
| Tuesday, 26 March 2024 | Dereel | Images for 26 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Where am I?
|
Topic: technology, photography, opinion | Link here |
One of the recurring themes in film series that we watch is that somebody gets lost (in the Alps, for example). They have mobile phone connection, but they don't know where they are.
That's simple: all modern phones have a GPS module that can identify where they are to within a couple of metres. And there are silly things like Plus Codes that convert the output into incomprehensible codes.
That's nonsense, of course. The phone should be able to send its location coordinates with every call, or on a call-by-call basis for the paranoid. But its not available anywhere that I can see. I've ranted about this numerous times, most recently in August 2022.
But the matter of the RAT last week got me thinking. Mobile phones always (as far as I can tell) embed the current GPS coordinates in every photo. And with only a moderate amount of coaxing you can not only get the information, but pass it on to Google Maps!
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Isn't that nice? Finally something that works.
Well, almost. I suppose it's modern that Google Maps gets the address wrong (it relates to something at the top of the screen, under the coordinates). But more to the point, it only works when you don't need it. This is a photo taken on my own phone. What about a photo sent with the messaging app? Tried that with Yvonne's phone and got:
|
|||||||||||||||||||||||
No Exif data! And look at the image sizes. Yvonne's phone is effectively the same as mine, but the image has been reduced from 12 MP to 1.9 MP (a variable which it's too polite to mention). About the only useful piece of information there is the path name, for which I had seaarched in vain with inadequate tools through the Android directory hierarchy.
So: once again Android fail. I can understand the paranoia of so many people who don't want their location to be known, but it should be easier than this, at least as an option.
One day I'll find something really good about Android. I'll be amazed.
| Wednesday, 27 March 2024 | Dereel | Images for 27 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Blacklist removal: success
|
Topic: technology, opinion | Link here |
Yesterday I received mail from Wendy Bartlett, whom I knew at Tandem Computers. Replied, or at least I tried to:
<wendy_bartlett@randomisp.net>: host ff-ip4-mx-vip2.prodigy.net[144.160.159.22]
said: 553 5.3.0 flpd594 DNSBL:RBL 521< 45.32.70.18 >_is_blocked.For
assistance forward this error to abuse_rbl@abuse-att.net (in reply to MAIL
FROM command)
Damn, will this never end? For the sake of completeness, sent a message to abuse_rbl@abuse-att.net to say that I was just an innocent bystander.
Surprise, surprise:
Thank you for contacting the AT&T Postmaster.
The mail-server IP address(es) associated with your request will be removed from the block list within 24-48 hours from the date of this letter. AT&T and its affiliates do NOT intentionally block legitimate mail in the course of our anti-spam initiatives and regret any inconvenience this may have caused.
And yes, now it works! I think this is the first time that I have been successful in such an appeal.
|
Eye progress
|
Topic: health, opinion | Link here |
My eye continues to improve. The bubble is still there, but barely. Hopefully it will soon be completely gone, and the new floaters with it.
Today Yvonne went shopping and picked up some reading glasses for me. We had carefully measured the distance from my head to monitor and established that it was about 60 to 62 cm. That falls clearly between 67 cm (1.5 dioptres) and 57 cm (1.75 dioptres). I suggested getting one set of each power if we could return one.
That didn't happen: nobody had 1.75 dioptre glasses on offer. So she came back with a 1.5 dioptre pair. But that's not an issue: I couldn't focus to 67 cm with them, more like 40 cm. That corresponds to 2.5 dioptres! And no, it's not the reading glasses. They focus images from outside at exactly 67 cm.
So it seems that currently, at any rate, my right eye is short-sighted by about 1 dioptre. Is that intentional? Is that what David Fabinyi meant when he said that “the lens focus is set to slightly shorter than infinity”? 1 dioptre isn't “slightly shorter”, it's considerably short-sighted. Hopefully things will improve.
| Thursday, 28 March 2024 | Dereel | Images for 28 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Bursting the bubble
|
Topic: health, opinion | Link here |
The bubble in my eye has been reducing in size, as advertised, but it was still there at 8:00 this morning. I had hoped that it would have been gone by yesterday at the latest. But then I went back to bed for an hour or so, and at 9:00 it was gone.
All OK now? Not really. The floaters are still there, probably as many as before, just longer strands now. And I still can't see clearly when reading the monitor (with the reading glasses) or watching TV (without glasses in the right eye). I can't make up my mind whether the focal length is increasing or not. I can no longer see my outstretched hand clearly, so it could be, but it seems still to be less than the 3.5 m or so distance to the TV, indicating 0.3 dioptres or more.
And then there's an unexpected additional issue. Walking the dogs today, with only non-prescription sunglasses, I had difficulty aligning the view from each eye. The image for the left eye was higher than for the right eye. I had noticed this on Saturday, but attributed it to the glasses. On the other hand, how could the operation cause this kind of issue? Off-centre lens?
Is this going to go away? If not, I have two choices: have the lens replaced (and hope that it'll be better next time) or wear glasses all the time, one of the things that I had hoped to avoid. But there's not much point in coming to any decisions now. I certainly wouldn't have anything done until after the other eye has been done, and even that looks less likely to happen on the appointed date 11 April: David Fabinyi won't be available before then, and his locum is unlikely to be able to help. Hopefully things will clear up before then. I don't want the present situation to last too long.
Received a card (credit card size) in the mail today, with no further comment. It seems that I have a TECNIS™ Eyehance with TECNIS Simplicity™ Delivery System DI800. What a lot of meainingless buzzwords! About the only thing of interest is that the lens has a power of +21.5D, which doesn't really say very much about where the focus is set.
|
I clean everything
|
Topic: general, opinion | Link here |
Petra Gietz helps us with the cleaning in the house. She takes her job seriously:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
| Friday, 29 March 2024 | Dereel | Images for 29 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
A cage for Bruno
|
Topic: animals, general, opinion | Link here |
Bruno is the first cat we've had who isn't allowed outside. It's clear that he doesn't like the restriction.
For some time Yvonne has been suggesting that we cover over the dog pen behind her bedroom. I've been reluctant because I couldn't see a way of really ensuring that Bruno can't get out of it. But of course there are people who make cages for exactly that. Yvonne came up with these:
Jesse Walsh is planning to build one for his own cats, and he's prepared to do the work for us too, so it sounds like a good idea to let him do his first. Then we can take a look at it and get a better feeling for what needs to be done.
|
Sweet and sour fish, yet again
|
Topic: food and drink | Link here |
I've had a somewhat strained relationship with Sweet and sour fish, but Yvonne had asked for it, so off to try things out again.
Last time I had come to the conclusion that deep frying is the way to go, so that's what I did today. I also added garlic for the first time. I had always noted that there was no garlic, and I considered it a feature, but maybe it's a bug after all. Certainly it tasted good with the garlic, to the point that Yvonne commented on how good it was this time. As usual, I'm still not completely happy. It wasn't until I updated the recipe afterwards that I discovered I had forgotten one detail: “reduce the quantity of sauce”.
| Saturday, 30 March 2024 | Dereel | Images for 30 March 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Recipe treasure trove
|
Topic: food and drink, opinion | Link here |
While looking for something else, probably with locate, discovered a directory /grog/recipes/alt.gourmand/recipes that seems to have been on my system for ever. It contains over 500 recipes and appears to be a snapshot of the USENET Cookbook. Files had troff markup:
=== grog@hydra (/dev/pts/28) ~/public_html/recipes/alt.gourmand/recipes 56 -> grog gravlax
groff -mm gravlax
Are they of any use? None of the ones that I saw seemed very interesting. And though the Wikipedia article considers it a pioneering attempt at Electronic Publishing, the dates of the recipes (1985 to 1989) overlap considerably with my own COSIG recipe collection.
|
Open letter to the New York Times
|
Topic: general, technology, opinion | Link here |
To any human at the New York Times:
TL;DR: I'm not receiving email, and I can't find any way to report it. I don't consider a “Customer Care” only available in the middle of the night to be acceptable. Please fix.
In more detail: I've had a digital subscription to the New York Times for over 6 years, and I've found it a useful resource. But today it occurred to me that I hadn't received any email from them for “a while”, How long? Well, you keep bombarding my old email address with subscription offers that aren't as good as the one I have, but in the last week the only message I had to the real email address was an excruciatingly long survey.
Looking back further, it seems that I haven't received anything for 9 days. The last log entry (with modified recipient address) was:
Mar 21 16:44:19 lax postfix/qmgr[863]: 5C95728099: from=<nytimes@e.newyorktimes.com>, size=36475, nrcpt=1 (queue active)
Mar 21 16:44:20 lax postfix/smtp[47810]: 5C95728099: to=<news@lemis.com>, relay=mx0.lemis.com[121.200.11.253]:25, delay=1.3, delays=0.03/0.01/0.55/0.75, dsn=2.0.0, status=sent (250 2.0.0 Ok: queued as 1B41E2635C7)
What's going on? I still have access to the site, and I receive messages on my mobile phone. Check my settings on the web site. All OK, though I'm still subscribed to a discontinued newsletter. OK, delete that and wait for the confirmation email. It didn't even try.
OK, try the help pages. Nothing useful, no way to contact a human being. With some difficulty I found Contact information for international subscribers, which gives me a free call number in Australia, 1800-170-145, available today from 22:00 to 6:00, exactly the time when I'm asleep.
OK, the times make perfect sense in New York (where they're 9:00 to 17:00), but these are also the times for Australia. Clearly that doesn't work.
So, New York Times, please:
You know my real email address. Let me know when you have fixed it.
|
Cleaning the windows
|
Topic: gardening, animals | Link here |
Jesse Walsh along today to clean the windows I spoke about his planned cat enclosure, but it looks as if he won't do it in a hurry. On the other hand, he has a client for whom he has already built an enclosure, so maybe we can take a look at that.
|
File system strangenesses
|
Topic: technology, opinion | Link here |
I've gradually been tidying up my old TV series archives. One of the bigger jobs was Um Himmels Willen, where I reloaded some because the originals were truncated. Finally got round to moving things around, but found something unexpected:
=== grog@tiwi (/dev/pts/21) /spool/Already/Series/Um-Himmels-Willen/D 413 -> ls -l 07-01-Am_seidenen_Faden.deu.vtt
-rw-r--r-- 4 grog home 58478 5 Mar 13:29 07-01-Am_seidenen_Faden.deu.vtt=== grog@tiwi (/dev/pts/21) /spool/Already/Series/Um-Himmels-Willen/D 414 -> mv 07-01-Am_seidenen_Faden.deu.vtt ../07=== grog@tiwi (/dev/pts/21) /spool/Already/Series/Um-Himmels-Willen/D 415 -> echo $?
0=== grog@tiwi (/dev/pts/21) /spool/Already/Series/Um-Himmels-Willen/D 416 -> ls -l 07-01-Am_seidenen_Faden.deu.vtt
-rw-r--r-- 4 grog home 58478 5 Mar 13:29 07-01-Am_seidenen_Faden.deu.vtt
So I move a file to a different directory. Move is successful. And the file is still there! How could that be?
=== grog@tiwi (/dev/pts/21) /spool/Already/Series/Um-Himmels-Willen/D 437 -> ls -il ../07/07-01-Am_seidenen_Faden.mp4 07-01-Am_seidenen_Faden.mp4
3196646 -rw-r--r-- 4 grog home 1638402168 4 Mar 03:30 ../07/07-01-Am_seidenen_Faden.mp4
3196646 -rw-r--r-- 4 grog home 1638402168 4 Mar 03:30 07-01-Am_seidenen_Faden.mp4
The very same file (same inode number) was already there, so the “move” was a success. I still can't see whether it's a bug not to report it, or not to report the original.
Another issue was more serious:
=== grog@tiwi (/dev/pts/3) /spool/Series/Watzmann-ermittelt 1766 -> rmdir 03/
ufs_rmdir: Bad link count 2 on parent inode 4228608 in file system /spool
rmdir: 03/: Invalid argument=== grog@tiwi (/dev/pts/3) /spool/Series/Watzmann-ermittelt 1767 ->
What's that?
=== grog@tiwi (/dev/pts/3) /spool/Series/Watzmann-ermittelt 1768 -> l -a
total 1
drwxr-xr-x 2 grog home 512 14 Oct 21:14 .
drwxr-xr-x 86 grog home 2,560 30 Mar 18:43 ..
drwxr-xr-x 2 grog home 3,584 30 Mar 17:59 03=== grog@tiwi (/dev/pts/3) /spool/Series/Watzmann-ermittelt 1769 ->
The directory 03 is there, but the parent directory only has two links. A directory should have two links plus the number of subdirectories, here 1, so the link count should be 3. What happened there? Some kind of file system corruption,clearly, but how? The modification time stamp (14 October 2023, 21:14) could give some indication, but I don't see anything in my diary.
|
More hydra completion
|
Topic: technology, opinion | Link here |
Apart from the various issues that I still need to think about, there's one thing about hydra that I've put on the back burner: the “beep” doesn't work. But then I found this diary entry from ten years ago:
... looked for the speaker connector, which I hadn't set up when I built the machine. With good reason: this high-quality enclosure doesn't have a speaker! Not a big issue, since they almost never fail, and I had dozens of old machines from which I could cannibalize a speaker. But why didn't they include one?
I need to take a look, which could require shutting down the machine. But if speakers are going out of fashion, which would make sense, isn't it time to find an alternative way to beep? One that could be customized to say who beeped would be good.
|
Another Postfix tweak
|
Topic: technology, opinion | Link here |
Why does my mail from hydra come from, well, groggyhimself@hydra.lemis.com, and not groggyhimself@lemis.com, like from eureka? I still don't know why, but Tom Ivar Helbekkmo, told me how to fix it. Put this in eureka's /usr/local/etc/postfix/main.cf:
masquerade_domains = $mydomain
I've been using Postfix for well over 20 years, and I'm still learning.
|
Das linke Osterschwein
|
Topic: food and drink, opinion | Link here |
Roast ham for dinner tonight, which I chose to call „Das linke Osterschwein“ (“the crooked Easter swine”) as contrast to BWV 158 „Das rechte Osterlamm“ (“the right Easter lamb”). We've done this before, but there was room for improvement.
First, as I had planned last time, cook it in the “bath tub” air fryer, which was at least big enough to take both the ham and the temperature probe. But how high? The stands made the choice for me:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Based on prior experience, I baked at a setting of 190°, in contrast to the 170° in the hair dryer. It took 54 minutes to bring the 560 g piece to 74° internal temperature:
|
|||||||||||||||||||||||
But it was still just as difficult to remove the net:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Nothing for it: we'll have to remove the net before roasting next time.
|
Eye: not there yet
|
Topic: health, photography, opinion | Link here |
My eye continues to improve, and it's really conceivable that it will end up with normal infinity vision. But this evening, looking at the garage lights, I saw streaks reminiscent of problems that I have had with camera lenses. It's amazingly difficult to find an example of the effect, though in the past I've seen them seeing something like the streaks in this image (from SIGGRAPH 2011 paper on Physically-Based Real-Time Lens Flare Rendering via Wikipedia):
The part I'm seeing are the spikes from the central source, not the reflections, and in my case they're irregular and mainly in a horizontal plane. Will they go away? Possibly not, but I could live with it.
In passing, the Wikipedia page barely distinguishes between flare and streaks. For me they're two very different things.
| Sunday, 31 March 2024 | Dereel | Images for 31 March 2024 |
| Top of page | ||
| previous day |
|
The wrong Easter lamb
|
Topic: food and drink, language, opinion | Link here |
Today was Easter Day. After yesterday's „linkes Osterschwein“ (“the crooked Easter swine”), it only seemed reasonable to follow Bach's BWV 158 „Das rechte Osterlamm“ (“the right Easter lamb”). But Yvonne doesn't like lamb any more, like so many other things. How about seekh kebab?
One reason is that I've always had trouble keeping the kebabs together. But not-so-recently I bought a mixture that might make life easier:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The package is interesting for a number of reasons, notably language. It can't decide whether it's for Seekh Kabab or Shish Kebob. The instructions are in English and German, where I can read „keine Preservative“ (“no condoms”), but some details are in French, and there are vestiges of Italian. It promises recipes in French, Urdu, Hindi and Bangla inside the package, but I didn't find them.
And of course it gives ingredients in archaic measures: teaspoons and tablespoons, whatever they may be for solids. I find this the most frustrating part of trying a new recipe, trying to guess what they meant.
And there more language hits me. The ingredient list calls for butter, but the instructions mention oil.
|
|
|||||||||
|
|
|||||||||
What is it? My guess is that the butter is ghee, and in the original language (Urdu?) it can also be called oil. That's certainly the case in Malay, where ghee is called minyak sapi, and the word minyak means oil. But by the time the translation goes into German, things have changed completely: „Fügen Sie ½ - 1 Tasse Speiseöl hinzu“: “Add ½ to 1 cup cooking oil”. Not only definitely oil, but also almost certainly much more. Assuming US American cups, that would be between 120 and 240 ml of oil, compared to presumed 30 to 45 ml of butter. I wonder how many people have messed up their food like that.
Things weren't much better for me. My main concern was that the kebabs might fall apart, like they like to so much. I didn't even entertain the idea of putting them on skewers. But make sausages out of them and fry them?
|
|||||||||||||||||||||||
The ones at the back didn't look too bad. But they fell apart when I tried to put them into a serving dish:
|
|||||||||||||||||||||||
And while doing that, also made some roti paratha. But the issues with the kebabs took me longer than I expected, and it showed:
|
|||||||||||||||||||||||
I decided to deep fry some frozen pakoras and samosas from ALDI. How long? 2 minutes! For both of them! The pakoras were far larger, and it was fairly clear that they would need more than 2 minutes. But after 4 they looked OK. Only: they were still frozen in the middle! Grr!
All in all, definitely the wrong Easter lamb.
Do you have a comment about something I have written? This is a diary, not a “blog”, and there is deliberately no provision for directly adding comments. It's also not a vehicle for third-party content. But I welcome feedback and try to reply to all messages I receive. See the diary overview for more details. If you do send me a message relating to something I have written, please indicate whether you'd prefer me not to mention your name. Otherwise I'll assume that it's OK to do so.
| Top of page | Previous month | Greg's home page | Today's diary entry | Next month | Greg's photos | Copyright information |
