|
| Greg's diary |
| April 2022 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Friday, 1 April 2022 | Dereel | Images for 1 April 2022 |
| Top of page | ||
| next day | ||
| last day |
|
Finally! A solution to the foil package problem!
|
Topic: food and drink, opinion | Link here |
For some time I've been grumbling about the lack of tools for emptying foil paste packages like the ones I use for my East Asian breakfasts. It should be as simple as a couple of rollers held together, through which you put your foil package. But it seems that that's too obvious: I can't find one.
And then Yvonne found a solution, on Facebook, of course: a low-resolution version of this YouTube video:
That's actually a whole hour's worth of cooking tips, some of them completely uninteresting, but the idea of using the lid of a pot is so obvious that I don't know why it didn't occur to me before. It's not ideal, of course, but it's available. And some of the other tips are interesting enough that I should spend some time looking at them all.
|
More DMARC investigations
|
Topic: technology, opinion | Link here |
Why did the GLOCK APPS web site refuse to continue yesterday? Clearly it's a badly written web page, possibly designed for Microsoft. So I tried it on distress, my faster Microsoft box. And how about that, still using firefox, it positioned the warnings outside the field that they had previously obscured. And it reacted when I pressed SUBMIT. It didn't like “N0 future!” as a password (too weak, which it let me deduce to be the reason for not continuing). OK, try “Is this str0ng enough, Idiots?”? No, too long! So why did they let me enter it?
Finally got something set up. A free trial for (I think) 30 days, and then pay—through the nose. The pricing page shows at least $708 per annum (which, despite the annual payment schedule, they present as $59 per month), though with a little patience I was able to find prices as low as $9 per month, still out of the question. About the only positive thing I found was that they have multiple names. Apart from GLOCK APPS they also call themselves the marginally less obnoxious GlockApps.
And EASY DMARC? There, too, I found prices. Yes, really free! As long as you have only 1 User, whatever that may mean. What do I get for it? Presumably less than I'm getting now, and that's little enough. Analysis of 3 messages, two passed, one failed. Details? We don't need no steenking details. Or maybe I just couldn't find them.
I suppose DMARC makes sense for people who want to keep statistics, but in their case I'd expect something much more powerful.
|
More paddock work
|
Topic: gardening, animals, opinion | Link here |
Jesse Walsh along this afternoon, ostensibly to do garden work, but Yvonne got him to put up a secondary fence in the middle paddock:
Why that? She has a new riding arena in the north paddock, for which we have paid much money, and for which we will pay even more when Troy Addicoat erects the cover. She's really impatient for the cover, but she's not using the arena. It can't be the lack of cover: there's none in the west paddock either. I suppose things will soon change when the weather gets wetter.
| Saturday, 2 April 2022 | Dereel | Images for 2 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Clearing the house forest
|
Topic: Stones Road house, general | Link here |
The westernmost 20% or so of our property (about 0.8 ha including the part loaned by the Marriotts) are relatively dense trees with bracken underneath:
|
||||||||||||||||||||||||
There are also a number of wildflowers, mainly orchids such as Pterostylis, but they struggle to get through the bracken.
Now Yvonne has found Nathan Delaney, who has had the same problem on his property and is now prepared to remove ours for us. He thinks that 3 attempts at 6 month intervals should be enough to get rid of them for good. He'll be around in 10 days' time for the first, most radical mowing.
Nathan is interesting from another perspective: he bought the house belonging to the perpetrator of the bushfire 9 years ago. It seems that the perpetrator (arsonist) was identified, tried and sentenced, but he fled to Germany, of all places, where apparently he can't be extradited for that kind of crime. As a result, Nathan got the property very cheaply. But I'm surprised that Germany wouldn't extradite the arsonist (whose name I haven't been able to establish).
|
Making friends with dogs
|
Topic: animals | Link here |
Nathan brought his son with him, who, after a little hesitancy, made friends with Larissa:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
And Lena? She didn't want to know. We know that she's timid, but today it was particularly obvious.
|
The power of mulch
|
Topic: gardening, opinion | Link here |
The Abutilons that we propagated from cuttings last year are growing well. But we had noted that the one in the pot was flowering better than the one that we planted in the ground. Here photos from mid-February:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But since then Jesse has mulched the area where the one Abutilon was planted. Now we have:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
Goodbye Marriotts
|
Topic: general, opinion | Link here |
Garry Marriott and Diane, our neighbours, are going walkabout again. They invited us along to afternoon tea this afternoon as part of a going-away party:
|
||||||||||||||||||||||||
They had hoped for good weather and pleasant view from the verandah, but the weather didn't oblige, so they relocated to the garage:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
They'll be gone for at least 6 months, maybe a year, especially if they circumnavigate the continent:
| Sunday, 3 April 2022 | Dereel | Images for 3 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
DST end, more OI.Share pain
|
Topic: general, photography, technology, opinion | Link here |
Daylight Saving Time ended this morning, with the usual issues of resetting clocks. The biggest was with the
cameras. The older ones are simple: go to the menu and set it. But the three newest
(Olympus OM-D E-M1 Mark
I, OM-D E-M1
Mark II and OM-D
E-M5 Mark III) have a remote control program that also sets the time from
the computer device mobile phone to which they can connect. And
that's more accurate.
But this horrible OI.Share app makes it such a pain! In the spring I gave up, but today I tried again. One more surprise: the Toy Shop offers:
|
|
|||||||||
after installing it on Yvonne's phone, it came up with “Ha, ha, I'm the wrong OI.Share. I'm Olympus, you want OM Systems OM.Share”:
|
|
|||||||||
But that's wrong. It reports itself as OI.Share. And why is the old one still in the toyshop? It all gives the impression of complete disorganization.
The new one is no better than the old. The people still haven't understood networking: it makes the camera an access point and excludes everything except the phone. It even takes the phone off the network. And the setup instructions have never worked for me. Here's a typical sequence. After starting, select “Remote Control”. A popup appears at the bottom:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The greyed-out Connect means that it doesn't find anything! But going into the phone's “Settings” shows the device:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The alternative is “Easy Setup” (hah!):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Now there's something new. OM Systems OI.Share asks you for permission:
|
|||||||||||||||||||||||
And then the phone wants you to confirm:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Then I have to scan the QR code. It confirms:
|
|||||||||||||||||||||||
Oh. Now it discovers that it knows about a camera. Go to it? No, remove the entry and start again:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Only one camera can be registered? It seems that it's a different one. That's new. I wonder if it's even correct. OK, remove, register:
|
|||||||||||||||||||||||
It still doesn't see it!
Oh yes, it does:
|
|||||||||||||||||||||||
It only took 30 seconds. And then it connected, only another 15 seconds, bringing another spurious error message with it:
|
|||||||||||||||||||||||
I thought that that might be due to the empty second card slot on the E-M1 Mark II, but it occurs on the E-M1 Mark I (only one slot) too. But it's harmless.
But is the excessively slow reaction the reason that I couldn't connect before? Back to try it out. Yes! For reasons that make no sense whatsoever, it takes 45 seconds to connect to the camera, 30 of which are spent in navel contemplation. My best bet is that this happens when it's expecting a different camera (the one that is registered with it) and always waits for the timeout before continuing; the phone settings menu shows the camera long before the timeout.
And is this “there can only be one” correct? Yes, that seems to be the case, at least for the OM Systems version. And I can't test the old one because it refuses to work. What stupidity!
But wait, there's more. Part of the setup requires turning on remote access on the camera:
|
|
|||||||||
But that doesn't work on the E-M5 Mark III. Ah, the app is prepared for that:
|
|
|||||||||
Which menu, idiot? Off to grep the manual. “WiFi” doesn't appear at all. Of course not, it's “Connecting the camera to a smartphone”. And where in the menu is that? In Playback!
|
|||||||||||||||||||||||
You can also connect by tapping blob in the monitor.
No you can't. There's no such display. But what a crazy idea to hide the connection setup in the playback menu!
All in all it took me 30 minutes to connect the three cameras. I didn't need to do anything more: it sets the time automatically. But what's wrong with this situation? The camera has no business simulating an access point. It should be just another device in an existing network. Connecting should be similar to the way a mobile phone connects to a network.
Another issue is that the app insists on taking JPEG images, even when “raw only” is set. It also ignores raw images on the camera. Another indication that it's a toy.
So how can you live with OI.Share? My best guess is to have one mobile phone per camera and only use that pair. Then it seems to recognize the cameras relatively quickly.
If anybody from Olympus OM Systems reads this and has useful input, please
contact me.
|
Dead battery
|
Topic: photography | Link here |
Messing around with OI.Share meant leaving the cameras powered on while I fought the app. Result: the battery on the Olympus OM-D E-M1 Mark I was empty. OK, replace with the next in line (number 2). Dead! I couldn't measure any voltage at all across the terminals.
OK, this battery is was over 8 years old: I got it shortly after the E-M1, and the other two are still functioning. So it's a far cry
from the issues I had with the E-30.
|
Upgrading MediathekView
|
Topic: technology, multimedia, opinion | Link here |
MediathekView is a program for downloading videos from European TV broadcasters. I've been using it for nearly 5 years, but I never really got round to making it into a FreeBSD port.
Now I'm trying, and I've discovered that Lars Engels beat me to it—and then gave up again because of problems with Java. Now it's my turn.
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 155 -> mediathekview
Error: A JNI error has occurred, please check your installation and try again
Exception in thread "main" java.lang.UnsupportedClassVersionError: mediathek/Main has been compiled by a more recent version of the Java Runtime (class file version 61.0), this version of the Java Runtime only recognizes class file versions up to 52.0
at java.lang.ClassLoader.defineClass1(Native Method)
... another 12 irrelevant stack frames
OK, MediathekView, what's the correct Java version? That proved to be surprisingly difficult to find out, but ultimately I found a correlation between class file version major version and here. There should be a way to find them out in a more official manner.
OK, remove OpenJDK 8 and install OpenJDK 17. Bang! MediathekView gets deinstalled.
Installed packages to be REMOVED:
bootstrap-openjdk8: r450802_2
mediathekview: 13.8.1_1
openjdk8: 8.0
And at the end we had:
This OpenJDK implementation requires fdescfs(5) mounted on /dev/fd and
procfs(5) mounted on /proc.
If you have not done it yet, please do the following:
mount -t fdescfs fdesc /dev/fd
mount -t procfs proc /proc
I don't know why that irritates me, but it does. What business does a programming language have with system configuration?
Still, on. Reinstall MediathekView:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 180 -> make install=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 181 ->
Huh? Why didn't it install? Did it not get deinstalled? Oh yes, it did:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 181 -> make deinstall
===> Deinstalling for mediathekview
===> mediathekview not installed, skipping=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 182 -> make install
===> Installing for mediathekview-13.8.1_1
Somehow there's a disconnect here.
Then run it:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview 172 -> mediathekview
. Configuring for non-portable mode
Exception in thread "main" java.lang.ExceptionInInitializerError
at java.desktop/java.awt.GraphicsEnvironment.getLocalGraphicsEnvironment(GraphicsEnvironment.java:106)
at java.desktop/sun.awt.X11.XToolkit.<clinit>(XToolkit.java:224)
...
Caused by: java.lang.NullPointerException: Cannot read the array length because "this.screens" is null
at java.desktop/sun.awt.X11GraphicsEnvironment.initDevices(X11GraphicsEnvironment.java:235)
... 12 more
this.screens? That looks like a typo, but no, it's a Java array. So why is it NULL? One obvious reason might be that I have my DISPLAY set to eureka:0.2, in other words on a different system. That works with the old version, but maybe that was not intentional.
So what do I do now? There are a number of options: run X on tiwi, probably a good idea, but I have little hope of success. Use the version of Java supplied with the MediathekView distribution? If I can work out how to do that, and if it even works on FreeBSD, that could potentially solve the problem. For the moment, head-scratching seems like a good idea.
|
Cooking rendang
|
Topic: food and drink, opinion | Link here |
I still have a whole lot of spice mixes for East Asian food, including no less than four packages of this powder:
|
|||||||||||||||||||||||
It wants nice, round quantities: 1¼ “cups” (300 ml) of coconut milk, 1 “cup” (presumably 240 ml) of beef broth, 600 g of beef. OK, I can fix that. Beef comes in 500 g quantities, in this household, anyway, and coconut milk comes in 400 ml cans, so I'll use that and maybe 50 g of water to rinse it out. Instead of 600 g beef and 540 g of liquid (liquid 90% of the weight), I had 500 g of beef and 450 g of liquid, coincidentally also 90%. But that was far too much:
|
|||||||||||||||||||||||
Cook for 1 hour, reducing slightly? No, cook on a relatively high power for 2½ hours and reduce considerably:
|
|||||||||||||||||||||||
That looks better, but there's too much oil. I'll have to think about what to do there.
To my surprise Yvonne liked it. But then, rendang should be fiery hot, and this wasn't hot at all (that was presumably the purpose of the chili they asked for in the recipe). It also wasn't salty enough. But that can be fixed, and we have another 3 packages of the stuff.
| Monday, 4 April 2022 | Dereel | Images for 4 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Emptying foil packages, in practice
|
Topic: food and drink, opinion | Link here |
I was really excited about the trick for emptying foil packages that I mentioned on Friday:
I tried it under somewhat uncontrolled circumstances a couple of days ago, but today I was more careful. With a package of laksa paste with net weight 250 g, I had:
| State | Weight | Yield | % | Comment | ||||
| Full | 257.7 g | |||||||
| Normal empty | not measured | |||||||
| Wiped | 26.6 g | 231.1 g | 92.4% | |||||
| Rinsed | 7.7 g | 250.0 g | 100.0% |
I should have first pressed it out before trying the pan lid trick, but I didn't think of it. The important thing, though, is that after squeezing it all out, I was able to recover a further 18.9 g by rinsing the rest of the paste out of the package. That's over 7.5%. In passing, it's also surprising that I recovered exactly 250 g of paste.
Next time I'll try it the old way and weigh how much I get out. But it's not important. The old way is simpler than with the pan lid, and either way I need to rinse out the package. So goodbye, good idea: you're not good enough. It also suggests that I can give up my search for the rollers that I thought would solve the problem more elegantly.
|
Understanding O[IM].Share
|
Topic: technology, photography, opinion | Link here |
Spent much of the morning writing up yesterday's rant about OI.Share. That was useful for a number of reasons, most importantly because I now
understand the bugs features better.
First, what is the name of the app? What would you like it to be? OM Image Share? OI.Share? It seems that both are correct. On the one hand that means that the name of the app hasn't changed to add further confusion, but it also allows mistakes to take place, and there seem to be many of them. Most importantly, the only OI.Share that I can find in the toyshop is the old Olympus one, the one that installs and offers to install the new version.
Or is it? I made the mistake of installing via my phone. Looking at the toy shop on a real computer, I see:
|
|
|||||||||
The first is the old one, which should have been removed long ago, and the second is the new one. But that's not what I saw on my phone.
The real issue, though, is the extreme delay, which I attributed to a timeout. And as I guessed, yes, this is because I dared to try to connect to more than one phone. Got out enzian, Yvonne's old (well, not-so-old) Xiaomi Redmi 9A and installed the app on it. No problems setting up “easy setup”, and connected it to my Olympus OM-D E-M1 Mark II with no difficulties. Turned on the OM-D E-M1 Mark I and it connected immediately with hirse, my Redmi 9T. So as I guessed yesterday, it really is a case of “there can only be one”. What are these people thinking?
Put in a review on the toy shop. Before Google mangled it, it looked like:
This looks what you get when you let novices try to write network code. I'm astounded at how bad it is:
1. It requires a mobile phone. It can't connect to a real computer.
2. It wants to be an AP, requiring the phone to disconnect from any other WiFi network.
3. It only recognizes one camera!
4. If it can't find the camera it knows, it takes 45 seconds to connect to a different camera, by which time most people would assume that it wasn't going to make it.
5. It doesn't recognize raw images.
6. It takes JPEG images even when the camera is set to “raw images only”.
As a result, it's completely useless to me. The camera has all the smarts needed for a correct
implementation, but this app is a waste of everybody's time. See http://www.lemis.com/grog/diary-jun2021.php#D-20210624-043102 for more details
That was really intended as a criticism of either version, but while I was there I looked at the other reviews, nearly all of which were very negative. But it seems that the app was initially also very unreliable: people thought that it was worse than the previous one, as if that were possible. Still, if I have managed to work around Google's broken selection criteria, there were 549 reviews: 4 of them 5 star, 2 of them 4 star, 3 of them 3 star, 2 of them 2 star, the rest (538) 1 star. Somehow Google averages this as 2.5 stars.
So: now I have 3 cameras that OI.Share understands, and 3 mobile phones, each of which knows about one of them. And they connect quickly when turned on, even when all 3 cameras are turned on.
Still, that's only part of the problem. Of the 6 objections in my review, 4 remain:
But at least I can now use them to set the time relatively painlessly.
| Tuesday, 5 April 2022 | Dereel | Images for 5 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Understanding postfix
|
Topic: technology, opinion | Link here |
I've been running my own email server for over 30 years. It seems that I'm one of the few people who still does. That's not overly surprising: you can get free email services on every street corner. But I want to do it My Way, and things aren't getting any easier.
30 years ago I installed sendmail, the only game in town. Less than 9 years later, some time between 16 November 2000 and 29 January 2001, I migrated to postfix, mainly because sendmail had changed its configuration file format to be even more difficult (sorry, Eric).
But postfix is also anything but easy. It doesn't help that I now have 6 mail servers running, and they need to talk to each other. And postfix tries to be so much like sendmail that it's difficult to tell which one I'm using, not made even more difficult by some GNU program that installs a binary called /usr/local/bin/mail with different semantics.
Now, however, mail works. Sort of. dereel caused problems which I initially attributed to not running lockd. While searching for the problem, I discovered that it was running sendmail, though I had installed postfix. And once again it wasn't delivering mail.
OK, enough is enough. Let's finally do it right. First, select postfix, not sendmail. That's in /etc/rc.conf:
postfix_enable="YES"
Then stop sendmail and start postfix:
service sendmail stop
service postfix start
eureka is the main mail system at home. The others are there only to send mail, not to receive it. So I need to set up the other systems to send all mail to eureka. That's in /usr/local/etc/postfix/transport:
# $Id: transport,v 1.1 2022/04/05 05:26:59 grog Exp $
* smtp:eureka.lemis.com
Then tell /usr/local/etc/postfix/main.cf about transport:
--- main.cf 2021/11/04 17:54:50 1.1
+++ main.cf 2022/04/05 01:34:06
@@ -675,3 +677,4 @@
inet_protocols = ipv4
meta_directory = /usr/local/libexec/postfix
shlib_directory = /usr/local/lib/postfix
+transport_maps = hash:/usr/local/etc/postfix/transport
=== root@dereel (/dev/pts/0) /usr/local/etc/postfix 175 -> /usr/local/sbin/postmap transport=== root@dereel (/dev/pts/0) /usr/local/etc/postfix 176 ->
And with that, dereel sends all mail to eureka, which sends it round the world and back to mail.lemis.com multiple times until the mail loop dies: eureka doesn't feel responsible for mail from dereel.lemis.com.
What can I do? According to the documentation, I should probably set up mail on dereel to send mail with origin @lemis.com. That's simple:
--- main.cf 2021/11/04 17:54:50 1.1
+++ main.cf 2022/04/05 01:34:06
@@ -116,6 +117,7 @@
#
#myorigin = $myhostname
#myorigin = $mydomain
+myorigin = $mydomain
Problem: it doesn't work. Why not? To be determined. In the meantime I can tell it to explicitly accept mail from dereel.lemis.com:
--- main.cf 2020/03/10 02:51:00 1.51
+++ main.cf 2022/04/05 01:27:18
@@ -163,7 +163,7 @@
mydestination = $myhostname, localhost.$mydomain, $mydomain,
mail.$mydomain, www.$mydomain, lax.$mydomain, ffm.$mydomain,
aussie-gw.$mydomain, ftp.$mydomain, pasocentral.org,
- w4.$mydomain, lagoon.$mydomain, teevee.$mydomain, localhost
+ tiwi.$mydomain, lagoon.$mydomain, dereel.$mydomain, teevee.$mydomain, localhost
That works, but it seems tacky. I also added tiwi, but what happens when I get more machines? And why doesn't myorigin work? And why can't I set something like .$mydomain or *.$mydomain to indicate that I'll accept anything in the domain?
|
Cameras from Ukraine
|
Topic: photography, opinion | Link here |
One of the lesser-known facts about Ukraine is that it was once one of the world's biggest camera producers, maybe the biggest one at one point. They started producing cameras in Kharkiv in 1934 (FED, a Leica II copy) and in Kyiv in 1947 (very close copies of the Contax I and II, initially using parts imported from Zeiss in Dresden). I have already had a Kharkiv-built FED (disguised as a Leica), but I have never had a Contax rangefinder. Now they're available in large quantities on eBay.
OK, which one? Clearly an old one that looks like a Contax, like this one from „Das Gesicht des Krieges“ (1941) that I discussed 6 years ago.
|
||||||||||||||||||||||
|
|
||||||||
That looks like a Contax II, though the advertising talks about the Contax III, the first camera with a built-in light meter.
OK, what are the corresponding Kiev models? Kiev 2 and Kiev 3, and except for the inscription they look pretty much identical to the corresponding Contax models. Spent a lot of time looking around, preferably for a model with only Cyrillic inscription, like this one (Kiev 2, pretty much identical to the Contax II above):
But that proved incompatible with the other specs I was looking for. Finally found a Kiev 3 A (with flash sync!) with Jupiter 8 50 mm f/2 (a clone of the Zeiss Sonnar 50 mm f/2 on the image from the book above), promised to be in functional condition, including the light meter:
But it took several hours. Why is it so difficult?
| Wednesday, 6 April 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
How much dog food?
|
Topic: animals, opinion | Link here |
I've been pondering the correct quantity to feed the dogs for some time, and I had gradually come to the conclusion that Larissa should get 180 g of pellets and Lena 240. Why the difference? Lara had frequently left food behind if we gave her more.
Now we're running out of pellets, and Yvonne has bought some more, along with feeding guidelines: for dogs between 30 and 40 kg, our category, they recommend 410 to 500 g. That's roughly double what we have been giving them.
OK, let's try. Today I gave Lara 250 g and Lena 320 g. They both ate them all. Could it be that Lara's health issues have lessened her hunger and skewed the results? To be observed.
|
Backup failure
|
Topic: technology, opinion | Link here |
Seen in the system logs this morning:
dump -2uf - / | nice zstd -T8 > /dump/lagoon-FreeBSD/2/root.bz2
DUMP: WARNING: should use -L when dumping live read-write filesystems!
DUMP: Date of this level 2 dump: Tue Apr 5 21:00:00 2022
DUMP: Date of last level 0 dump: Fri Apr 1 21:00:00 2022
DUMP: Dumping /dev/ada0p4 (/) to standard output
DUMP: mapping (Pass I) [regular files]
DUMP: mapping (Pass II) [directories]
DUMP: estimated 153867 tape blocks.
DUMP: dumping (Pass III) [directories]
DUMP: dumping (Pass IV) [regular files]
zstd: error 25 : Write error : cannot write compressed block
DUMP: Broken pipe
DUMP: The ENTIRE dump is aborted.
What's that? From /usr/include/sys/errno.h:
#define ENOTTY 25 /* Inappropriate ioctl for device */
So how did I get that in the middle of the dump? Repeated the dump and it ran fine. But then it happened again in the evening!
What's causing this? I at least need to consider:
So what can it be? There's nothing obvious. I hate problems, no matter how trivial, that can't be tracked down.
| Thursday, 7 April 2022 | Dereel | Images for 7 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More mail server problems
|
Topic: technology, opinion | Link here |
After getting dereel's mail system working properly, it's straightforward to migrate it to tiwi. After all, I installed both systems in the same way.
Surprise, surprise, it wasn't. I had assumed that it would be straightforward, so I didn't assign much time time it. But what I saw was anything but straightforward. After making the appropriate configuration changes,
Apr 7 09:44:45 tiwi sendmail[80071]: 236Nijid080071: from=grog, size=26, class=0, nrcpts=1, msgid=<202204062344.236Nijid080071@tiwi.lemis.com>, relay=grog@localhost
Apr 7 09:44:45 tiwi postfix/smtpd[80033]: connect from localhost[127.0.0.1]
Apr 7 09:44:45 tiwi postfix/smtpd[80033]: NOQUEUE: reject: RCPT from localhost[127.0.0.1]: 451 4.3.0 <grog@tiwi.lemis.com>: Temporary lookup failure; from=<grog@tiwi.lemis.com> to=<grog@tiwi.lemis.com> proto=ESMTP helo=<tiwi.lemis.com>
Apr 7 09:44:45 tiwi sendmail[80071]: 236Nijid080071: to=grog, ctladdr=grog (1004/1000), delay=00:00:00, xdelay=00:00:00, mailer=relay, pri=30026, relay=[127.0.0.1] [127.0.0.1], dsn=4.3.0, reply=451 4.3.0 <grog@tiwi.lemis.com>: Temporary lookup failure, stat=Deferred: 451 4.3.0 <grog@tiwi.lemis.com>: Temporary lookup failure
Apr 7 09:44:45 tiwi postfix/smtpd[80033]: disconnect from localhost[127.0.0.1] ehlo=1 mail=1 rcpt=0/1 data=0/1 rset=1 quit=1 commands=4/6
What's wrong there? Clearly a problem with postfix masquerading as sendmail, but here they both seem to play a part. I didn't have time to look for the causes today.
|
Magda and Nele visit
|
Topic: general, animals | Link here |
Now that COVID-19 is gradually becoming part of life, people are allowed to fly again, and Magda Delva is back in the country from Austria. Today she and daughter Nele came for lunch, not an ideal time for a visit, but the best that can be done. We would have preferred an evening visit.
After lunch, out to visit the horses:
|
||||||||||||||||||||||||
In the process we took Lena to that part of the property for the first time ever. She likes water—in fact, she plunged in the trough outside the lounge room on the way—but was very nervous about the “dam” and required much encouragement to even put a paw into it.
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
Paella again
|
Topic: food and drink, opinion | Link here |
For lunch we made a paella valenciana, the first time in 3½ years.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It was a success and well received, but somehow it's too much work.
|
Ukrainian cameras at a discount!
|
Topic: photography, politics, opinion | Link here |
My Kiev 3 camera is now on its way from Ukraine. Tracking tells me:
| 7 Apr 2022 11:19am | Відправлення прийняте |
Now I just need a Fed 1, made in Kharkiv. More searching today and found one at a discount:
|
|||||||||||||||||||||||
|
|
|||||||||
I was sorely tempted to buy it, but the condition wasn't good enough, so I bought a different one, slightly more expensive, but also from Ukraine.
| Friday, 8 April 2022 | Dereel | Images for 8 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Pointed capsicum?
|
Topic: food and drink | Link here |
On Wednesday Yvonne bought a bag of capsicums—she thought. But she found this in it:
|
|||||||||||||||||||||||
A relatively mild chili, and potentially something that we should investigate.
|
Servant of the People
|
Topic: politics, multimedia, opinion | Link here |
A number of politicians, even at the very top, have a background in film acting, and even without cynicism it's understandable why. Certainly the presentation skills are useful. Names that spring to mind are Ronald Reagan, Arnold Schwarzenegger and Donald Trump.
What were their films like? We watched Schwarzenegger as a he-man years ago. Not my kind of film. But then we watched a Reagan film because it had Reagan (cowboy) in. Not as good. And I even tried a Trump episode from The Apprentice, where he played himself, but it was so bad that I didn't finish.
Now there's a new contender: Volodymyr Zelenskyy from Ukraine. In his series “Servant of the People” he played a normal person who unexpectedly was elected to be the President of Ukraine. The series was aired between 16 November 2015 and 28 March 2019, after the Russians had annexed Crimea and supported the insurgents in Donbas.
The interesting thing about Zelenskyy is that he really fulfilled his role in the series. Less than 2 months after the last episode of the series, he really did become the President of Ukraine.
But what was his TV show like? It's now available on SBS, and today we watched the first episode. Surprise, surprise! It's good. Completely improbable, of course: a teacher runs for president, though he has no experience and campaign, campaigning mainly on social media, and wins by a landslide. All he has going for him is his honesty, something rare in that field.
And his real life campaign? He runs for president, though he has no experience and campaign, campaigning mainly on social media, and wins by a landslide. All he has going for him is his honesty, something rare in that field. That's where the comparison ends, I assume: I don't expect that anybody anticipated the Russian aggression.
|
Зеленський? Zelensky? Zelenskyy? Zelenskiy? Selenskyj? Selenski?
|
Topic: language, opinion | Link here |
what's the name of the Ukrainian President? Володимир Олександрович Зеленський, of course. But most people can't read that, so they transcribe into Roman letters. And there is absolutely no consensus on how to do that. The Ukrainians themselves recommend Zelenskyy, but it's not clear to me why. They first y corresponds to Cyrillic и and the second to й. If it were Russian, it would be transliterated Zelenskiy.
But this isn't Russian. Or is it? Зеленський is a native Russian speaker, though presumably he has spoken Ukrainian since infancy, and there are no Ukrainian letters in his name. It would be easy to say that it's clearly politically incorrect to use a Russian transliteration, but he has used the spelling Zelenskyy since long before the present situation. Peter Dickson of the Atlantic Council wrote an article on the subject only shortly after his election.
But that was only transliterations into “English”. What about other European languages? In France they spell it Volodymyr Zelensky, and the Germans have a completely different translation to match German prounuciation: Wolodymyr Selenskyj.
The real issue is a continuation of a problem I have had for nearly 60 years: how do you pronounce the letters for i in Russian? There's Ы, И and Й, considerably different from other vowels, which come in pairs, one iotified (Я, for example, and the corresponding non-iotified А). But that doesn't work with the is: Ы is not iotified, and I never learnt how to pronounce it properly.
But that's Russian. Ukrainian goes further: І and Ї. And it seems that Ы sounds different in Ukrainian. And that, it seems, is why they transliterate І as i and Ї. as ï, leaving y for both И and Й.
As if that weren't confusing enough, the Russian pair Е (iotified) and Э (not iotified) are almost reversed in Ukrainian, where Є (note direction) is iotified and Е is not.
So where do I find reliable information? Wikipedia is useless for this sort of thing. It's far too US-centric in pronunciations.
| Saturday, 9 April 2022 | Dereel | Images for 9 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
pkg problems
|
Topic: technology, opinion | Link here |
In the night's security run output today, found:
pkg: Warning: Major OS version upgrade detected. Running "pkg bootstrap -f" recommended
Huh? Oh well, it's straightforward enough:
=== root@dereel (/dev/pts/0) /usr/local/etc/postfix 180 -> pkg bootstrap -f
pkg: Warning: Major OS version upgrade detected. Running "pkg bootstrap -f" recommended
pkg(8) is already installed. Forcing reinstallation through pkg(7).
The package management tool is not yet installed on your system.
Do you want to fetch and install it now? [y/N]: y
Bootstrapping pkg from pkg+http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly, please wait...
pkg: Error fetching http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly/Latest/pkg.txz: No route to host
pkg is already installed. The package management tool is not installed. Clearly the error messages could be improved on. But what's this “No route to host”? That's a new one. Is it correct?
=== root@dereel (/dev/pts/0) /usr/local/etc/postfix 181 -> ping pkg.FreeBSD.org
PING pkgmir.geo.FreeBSD.org (140.113.168.171): 56 data bytes
64 bytes from 140.113.168.171: icmp_seq=0 ttl=47 time=322.764 ms
64 bytes from 140.113.168.171: icmp_seq=1 ttl=47 time=414.189 ms
^C
--- pkgmir.geo.FreeBSD.org ping statistics ---
2 packets transmitted, 2 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 322.764/368.477/414.189/45.712 ms
That looks fine. Try again.
=== root@dereel (/dev/pts/0) /usr/local/etc/postfix 182 -> pkg bootstrap -f
pkg: Warning: Major OS version upgrade detected. Running "pkg bootstrap -f" recommended
pkg(8) is already installed. Forcing reinstallation through pkg(7).
The package management tool is not yet installed on your system.
Do you want to fetch and install it now? [y/N]: y
Bootstrapping pkg from pkg+http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly, please wait...
pkg: Error fetching http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly/Latest/pkg.txz: Not Found
A pre-built version of pkg could not be found for your system.
Consider changing PACKAGESITE or installing it from ports: 'ports-mgmt/pkg'.
Well, at least it found the system, but not the file. Does pkg still work?
=== root@dereel (/dev/pts/0) /usr/ports 189 -> pkg upgrade firefox
pkg: Warning: Major OS version upgrade detected. Running "pkg bootstrap -f" recommended
Updating FreeBSD repository catalogue...
pkg: Repository FreeBSD has a wrong packagesite, need to re-create database
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/meta.txz: Not Found
repository FreeBSD has no meta file, using default settings
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/packagesite.pkg: Not Found
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/packagesite.txz: Not Found
Unable to update repository FreeBSD
Error updating repositories!
Nope, there seems to be something seriously wrong with the package database. OK, build pkg from ports and install
=== root@dereel (/dev/pts/0) /usr/ports/ports-mgmt/pkg 196 -> make install
===> Installing for pkg-1.17.5_1
===> Checking if pkg is already installed
===> Registering installation for pkg-1.17.5_1
Installing pkg-1.17.5_1...
...=== root@dereel (/dev/pts/0) /usr/ports/ports-mgmt/pkg 197 -> pkg upgrade firefox
Updating FreeBSD repository catalogue...
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/meta.txz: Not Found
repository FreeBSD has no meta file, using default settings
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/packagesite.pkg: Not Found
pkg: http://pkgmir.geo.freebsd.org/FreeBSD:14:amd64/quarterly/packagesite.txz: Not Found
Unable to update repository FreeBSD
Error updating repositories!
Nope. I'm dead in the water. How come I'm the only person to have this problem?
|
More air conditioner problems
|
Topic: Stones Road house, technology, photography, general, opinion | Link here |
It's been a month since I discussed our air conditioner problems with Tony from Atmos. I was to check the connections of the sensors to the main controller board.
No hurry: the machine is working fine. There don't seem to be any issues when using it to heat, though it produces more E25 statuses than before. But today it was warm again. Air conditioner off, change from “Heat” to “Cool”. Turn on. It turned itself off again, repeatedly.
OK, time to finally check the connectors. They're the ones at bottom left on the board:
|
|||||||||||||||||||||||
Somehow I had a recollection of a looser fit, like connectors on a computer board, which can get loose. These seem fairly foolproof:
|
|||||||||||||||||||||||
Even the different connectors make sense: that ensures that they don't get connected the wrong way round. The one on the left (red/white wires) is the ambient temperature sensor. Does the colour coding mean that it's polarized? I would have expected some thermoresistive sensor.
In any case, disconnected and reconnected the sensors. Power the machine back on again. The good news is that it worked again. The bad news is that the E25 are still there. So I'll have to call Tony again.
In passing, it's interesting to note these three LEDs:
|
|||||||||||||||||||||||
The PWR one is on red all the time, which makes sense. And COM1 and COM2? They blink green at about a 1 s interval. And despite all attempts, it's barely recognizable. It seems that I need to significantly reduce the exposure, here by 2.5 EV:
|
|||||||||||||||||||||||
|
Poulet basquaise revisited
|
Topic: food and drink, opinion | Link here |
We used to eat poulet basquaise relatively frequently, but gradually it got boring. It seems that the last time was Christmas 2015. Clearly time to do it again.
The biggest difference was that today we cooked deboned chicken thighs instead of drumsticks. Definitely a good idea. And I had forgotten how much it all is, resulting in a choice of a pan that was just too small:
|
|||||||||||||||||||||||
Even the replacement pot barely made it:
|
|||||||||||||||||||||||
| Sunday, 10 April 2022 | Dereel | Images for 10 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Kitchen computer?
|
Topic: technology, food and drink, opinion | Link here |
Decades ago I kept a laptop in the kitchen, mainly for access to recipes. That was before the days of the World Wide Web, but it was still useful. Since then I've taken laptops into the kitchen from time to time, but with graphical displays I've come to the conclusion that laptops are a pain to use. What I need is a Real Computer.
And of course, I have plenty of those. The closest is fwaggle, my Apple iMac that has been sitting in the dining room doing nothing for months. How about putting it closer to the kitchen?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, that works, but it still seems wrong. I need to access it standing up, and it's too low for that. I need to think more about what I really want. In the meantime I'll leave it there and try it from time to time.
|
Leica or Contax?
|
Topic: photography, history, opinion | Link here |
I've frequently marvelled at how Leica has come to its amazing reputation. It's been around for ever, of course. Well, real Leicas have (coupled) rangefinders, and the first Leica with a rangefinder was the Leica II, introduced in 1932. And then the Leica M3 came out 20 years later, so perfect a camera that its current successor, the M11, barely looks any different.
But that wasn't the only rangefinder camera. Since researching the Contax cameras, I'm baffled that Leica succeeded and Contax was forgotten.
To compare: the big issues in those days were the rangefinder (of course), the shutter, the lens and exposure measurement. The Leicas had separate viewfinder and rangefinder, and the base of the rangefinder (the distance between the two windows) was about 5 cm. On the Contax there was a single viewfinder with a 9 cm rangefinder base, making the measurements twice as accurate. The Contax also had a bayonet lens mount and things that didn't become mainstream until decades later, like metal shutter curtains, vertical travel shutter, single non-rotating shutter speed dial, exposure meter (only on the Contax III) and a self-timer. This page goes into more detail.
So why didn't the Contax survive? It took Leitz until 1954 to introduce the M3, and even then it didn't have some of the features that the Contax had 15 years earlier, like exposure meter and metal shutter curtains. As far as I can see, no film Leica ever had an exposure meter. It wasn't for lack of need: when I bought my Pentax SV in July 1965, the first thing I got for it was a half-coupled exposure meter. Leitz had clip-on exposure meters, but like the Contax III, they weren't coupled.
So why did the Contax die? The Kiev might be the clue: the Contax was built in Dresden, and most of the production equipment was shipped to Kyiv, where the camera went through a name change and suffered from the quality and innovation issues that plagued the Soviet Union. Leitz was in Wetzlar, comfortably in the American zone, where they could continue from where they left off.
But that alone doesn't seem to be the answer. Zeiss relocated to Oberkochen, also in the American zone, where they brought out cameras that I always thought were second-rate. Why? Maybe they sold better.
|
Ukraine: the Russian view
|
Topic: politics, opinion | Link here |
In any dispute there are two sides. In the case of the Ukraine everybody in the “West” was caught by surprise. People still are not in agreement about the reasons. Putain wanted to “own” Ukraine? Russia had serious security concerns with Ukraine joining NATO?
Today Yvonne came up with a Facebook link to Deutsche Welle article titled Ukraine war: How to spot the propaganda and fakes designed to manipulate us. It's quite interesting. One of the examples they chose was the Russian claim that Zelenskyy had fled Kyiv in the early days of the war.
That's clearly not the case, of course, so they've taken down the page, right? Wrong. It's still there:
Zelensky hastily fled Kiev, Russian State Duma Speaker claims
"He fled to Lvov with his entourage, where he and his assistance were provided with accommodation," the speaker said
MOSCOW, February 26. /TASS/. Ukrainian President Vladimir Zelensky hastily fled Kiev on Friday and is currently in Lvov with his entourage, Russian State Duma Speaker Vyacheslav Volodin stated in his Telegram channel Saturday."Zelensky hastily fled Kiev. He was not in the Ukrainian capital yesterday already. He fled to Lvov with his entourage, where he and his assistance were provided with accommodation," the speaker said, adding that all videos published in the social media were recorded in advance.
"The information comes from Verkhovna Rada lawmakers who tried to meet with [Zelensky] in Kiev. They were invited to Lvov for a meeting. Zelensky is currently under protection of neo-Nazis," he added.
The mind boggles. I'm reminded of the (surprisingly) old saying “There is no truth in Izvestia and no news in Pravda”. But the link to TASS was an interesting starting point, and it didn't take long to find this article: “What led up to and sparked the Ukraine conflict in 2022”.
It's full of lies, of course, but it's interesting reading anyway. The kind of lie gives me a better understanding of the real issues:
Russia called on NATO to refrain from creating situations posing a threat to security of both sides, to give up any military activity in Eastern Europe, the South Caucasus and Central Asia and also to stop the US-led military bloc’s further eastward expansion.The US-led North Atlantic bloc announced the expansion of its military presence in Eastern Europe amid mounting tensions.
“US-led military bloc” is a twist that I wouldn't have expected.
Apart from that, the article is full of lies and half-truths. I'm still trying to read between the lines.
The Russians have an advantage, of course: they control the news in Russia, and this kind of junk is almost the only thing the Russian population sees. Still, it's interesting to look at today's headlines. The English version has:
Kiev plotting provocations with mass killings of civilians to accuse Russian troops.
Apart from the sheer ludicrousness of the claim, it's completely unrelated to the Russian version:
В первом туре выборов президента Франции лидирует Макрон после обработки 95% голосов.
“Macron leads first round of French presidential election after processing 95% of votes”. Who are the English-speaking targets of the first headline? Do even Russians believe that the people of any country would kill their own civilians to blame an enemy?
I'm still horrified, and despite the lull in the war at the moment, I don't see any end in sight.
| Monday, 11 April 2022 | Dereel | Images for 11 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
For the cat
|
Topic: food and drink, language, opinion | Link here |
There's a German phrase „für die Katz'“ (“for the cat”), usually meaning it's not good enough or in some way unpalatable. This page has a long explanation that I suspect was invented later.
Yvonne has a horror of fat, and she throws away meat that seems perfectly good to me. Recently it was some of the less lean parts of chicken breast, which she was going to give to the cat, literally “for the cat”.
I disagreed. It's good enough for my breakfast, especially this piece:
|
|||||||||||||||||||||||
|
Tracing the pkg issue
|
Topic: technology, opinion | Link here |
So why can't I do anything with pkg any more on dereel? Conveniently the messages gave me the exact URLs of the failing download, such as http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly/Latest/pkg.txz. And of course a browser confirmed that the file doesn't exist. http://pkg.FreeBSD.org/FreeBSD:14:amd64/ has only one directory, latest. There are two main “mirror” sites and a number of “mirror” “mirror” sites, and they all agreed.
OK, mail to the ports mailing list:
I've checked both IPv4 sites, and
http://pkg.FreeBSD.org/FreeBSD:14:amd64/ only contains a "latest"
directory, not "quarterly".
Got an answer from a top-poster, appending my message unchanged:
Only latest pkg sources in current (now is 14).
Well, yes, that's what I said. Maybe he was just trying to show the dangers of top-posting.
Later, though, Mark Millard sent a clearer reply. Yes, there's only a latest directory. quarterly is for -STABLE releases. With a bit of investigation I discovered that the issue came from installing pkg on a -STABLE machine and then upgrading to -CURRENT. The upgrade to -CURRENT went without a hitch, but it seems that pkg waited until the new quarter to raise its ugly head.
The fix?
--- /etc/pkg/FreeBSD.conf 2019/11/01 04:31:45 1.1
+++ /etc/pkg/FreeBSD.conf 2022/04/11 23:39:21
@@ -8,7 +9,7 @@
FreeBSD: {
- url: "pkg+http://pkg.FreeBSD.org/${ABI}/quarterly",
+ url: "pkg+http://pkg.FreeBSD.org/${ABI}/latest",
mirror_type: "srv",
| Tuesday, 12 April 2022 | Dereel → Geelong → Dereel | Images for 12 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
KL Hokkien Mee: time to experiment
|
Topic: food and drink, opinion | Link here |
I've cooked KL Hokkien Mee on several occasions now, and I'm getting more confident. Time to experiment.
First, add chili. As it stands, it's one of the few dishes with no spice at all. And for some reason I had left some tauge out, along with pok choi instead of choi sam.
The chili was an advantage. The tauge was not, and the pok choi was not sufficiently different from the choi sam to be of interest, though I think I'll stick with choi sam next time. And once again I'm reminded that there's not enough salt. Should I add more light soya sauce or just salt? I'll try soya sauce first.
On a strictly procedural note, it's better to mix the broth and cornflour in a smooth white bowl rather than the plastic forms I have been using so far. That way it's easier to scrape the rest out with a spatula.
|
Kitchen computer pain
|
Topic: technology, food and drink, opinion | Link here |
While cooking the mee, I needed access to the recipe. Exactly the reason why I put the Apple (almost) in the kitchen on Sunday. So I used that.
It really isn't good. It started by using Safari instead of the firefox that I know, if not love. And that is just plain confusing.
OK, install firefox. That made things easier, once I discovered that c-f no longer works for “find”. It's m-f, though only because it's not a real Apple keyboard; otherwise it would be some silly Apple name. And this silly Apple mouse takes getting used to as well.
But I can put up with that. The real issue is accessing the computer standing up. People say that it's good to use them like that, but I can't see any advantage. Still, I can use it, so I'll try again and see if I can get used to it.
|
Periodontic checkup
|
Topic: health | Link here |
Off to Geelong today for my 6-monthly periodontic checkup, not made easier by COVID-19. Now Leela uses conventional tools to scrape around, rather than the high-speed drill-like devices: they produce too much spray. But the old devices are significantly more uncomfortable.
While I was there, showed him the remains of my old interdental brush. He confirmed what I have discovered: they seem to be going out of style. He couldn't explain why either. In the end I took a set of disposable brushes ($10 for 6, as opposed to $100 for 100 of the old interdental brushes), but they're not easy to use.
That's what I wrote, but I know that $100 is definitely wrong. But I no longer recall what it was. $10 sounds more plausible.
|
What pump?
|
Topic: general, opinion | Link here |
While in Geelong, also filled up my car for the first time this year:
|
||||||||||||||||||||||||
Where's the pump number? I couldn't see it anywhere. Only when I got to the cashier could I see it:
|
||||||||||||||||||||||||
|
Where have all the ATMs gone?
|
Topic: food and drink, general, opinion | Link here |
While in Geelong, off to Gourmet Asian Grocery for my Malaysian food topup. Since the Fruit Shack, I have needed less and less, and I was done quickly.
But I also needed to find an ATM to pick up some money for Nathan Delaney, who is coming tomorrow to mow the bracken. OK, where is there a Westpac ATM in Geelong? Their web site is too polite to worry me with such things, so I checked Google Maps. Bingo! 80 m from the GAG (bottom left, not marked here):
|
|
|||||||||
Off to take a look. I couldn't find it. Later the proprietor of the GAG told me that it had been closed for over a year, and that the nearest remaining ATM was some distance away. It reminds me of my issues this time last year, only that was with my old GPS navigator with no updates.
Never mind, I know where there's an ATM in Bannockburn, so I can pick up the money there.
But it's gone too! Why? A few years ago we saw bank branches closing down, presumably as the result of online banking. But ATMs? People still use cash. My best guess is that “cash out” at supermarket checkouts is replacing them.
|
Where are my cameras?
|
Topic: photography, technology, opinion | Link here |
I now have two cameras on their way from Ukraine. How are they progressing? eBay gives (usually somewhat out of date) tracking information. For the Kiev, after fixing the back-to-front representation, it's (conveniently hidden in a too-small window)
| Ukrposhta tracking #RH062990804UA | ||
| 6 Apr 2022 8:59pm | Tracking number provided | |
| 7 Apr 2022 11:19am | Відправлення прийняте: | |
| 8 Apr 2022 12:43pm | Надходження на сортувальний центр | |
| 9 Apr 2022 12:18pm | Надходження на сортувальний центр | |
| 10 Apr 2022 9:47pm | Надходження до установи обміну України | |
| 10 Apr 2022 10:15pm | Підготовлено для проведення митного контролю | |
| 11 Apr 2022 12:34pm | Випущено з митного контролю для подальшої доставки | |
| 12 Apr 2022 2:32am | Відправлення оброблено в установі обміну і передано для подальшого транспортування до країни призначення |
And for the FED it's:
| Ukrposhta tracking #RH063075495UA | ||
| 11 Apr 2022 4:30pm | Tracking number provided | |
| 11 Apr 2022 10:35am | Відправлення прийняте: | |
| 12 Apr 2022 5:57pm | Надходження на сортувальний центр |
What does all that mean? I can get Google Translate to tell me what it means, but how about Ukrposhta? Like other postal services, they give only an overview; you have to “click here” to get details:
| Acceptance: | 04214 KYIV 214 UKRAINE | 07.04.2022 | ||
| Dispatching | 04214 KYIV 214 UKRAINE | 07.04.2022 | ||
| Income | 03914 DOP №1 DOPP UKRAINE | 08.04.2022 | ||
| Dispatching | 03914 DOP №1 DOPP UKRAINE | 09.04.2022 | ||
| Income | 03909 DOPP KYIV UKRAINE | 09.04.2022 | ||
| Dispatching | 03909 DOPP KYIV UKRAINE | 09.04.2022 | ||
| Arrival at outward office of exchange | 03928 KYIV MMPO UKRAINE | 10.04.2022 | ||
| Item presented to customs: | 03928 KYIV MMPO UKRAINE | 10.04.2022 | ||
| Released from customs for further delivery: | 03928 KYIV MMPO UKRAINE | 11.04.2022 | ||
| The postal item has been processed at the office of exchange | ||||
| and handed over for further transportation to the destination country | 03928 KYIV MMPO UKRAINE | 12.04.2022 | ||
That's more intelligible, even if they're too polite to give the times of day. The interesting part, of course, is the location. Clearly 04214 and 03928 are post codes, but Google Maps can't identify them. Only MMPO clearly stands for the main post office.
The FED comes from elsewhere:
| Income | 53299 DOP NIKOPOL UKRAINE | 12.04.2022 | ||
| Dispatching | 53201 NIKOPOL 1 UKRAINE | 12.04.2022 | ||
| Acceptance: | 53201 NIKOPOL 1 UKRAINE | 11.04.2022 | ||
Nikopol is on the Dnieper, much closer to the conflict zone. I'll keep an eye on that for a number of reasons.
| Wednesday, 13 April 2022 | Dereel | Images for 13 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More port upgrade fun
|
Topic: technology, opinion | Link here |
So now I can finally upgrade my ports on dereel. I started yesterday, but the transition from 13-STABLE quarterly to 14-CURRENT latest certainly made its presence felt:
===== Tue 12 Apr 2022 11:47:35 AEST on dereel: pkg upgrade
Updating FreeBSD repository catalogue...
Newer FreeBSD version for package tdb:
To ignore this error set IGNORE_OSVERSION=yes
- package: 1400056
- running kernel: 1400053
Ignore the mismatch and continue? [y/N]: Processing entries............. done
FreeBSD repository update completed. 30988 packages processed.
All repositories are up to date.
Checking for upgrades (754 candidates): .......... done
Processing candidates (754 candidates): .......... done
The following 784 package(s) will be affected (of 0 checked):
Installed packages to be REMOVED:
ORBit2: 2.14.19_2
gindent: 2.2.12_1
policykit: 0.9_10
py37-dbus: 1.2.16
py37-evdev: 1.3.0
spidermonkey60: 60.9.0_7
...
zopfli-1.0.3_1 (ABI changed: 'freebsd:13:x86:64' -> 'freebsd:14:x86:64')
zsh-5.8.1 (ABI changed: 'freebsd:13:x86:64' -> 'freebsd:14:x86:64')
Number of packages to be removed: 6
Number of packages to be installed: 25
Number of packages to be upgraded: 217
Number of packages to be reinstalled: 536
The process will require 346 MiB more space.
2 GiB to be downloaded.
Proceed with this action? [y/N]: [1/778] Fetching zstd-1.5.2.pkg: .......... done
There's almost always something that gets REMOVED. In some cases it's important, like web browsers or (a particular favourite) Emacs. But why the kernel version mismatch? I couldn't explain this: after all, it is the latest. I upgraded again just to be sure, but this was the (same) result afterwards.
This is still a big upgrade, and I let it run overnight. When I came back I found:
===> Pre-installation configuration for squid-5.4.1
[698/784] Extracting squid-5.4.1: .......... done
pkg: Fail to rename /usr/local/etc/squid/errors/es-mx/.pkgtemp.ERR_ACCESS_DENIED.Qa5SqrmzMQEg -> /usr/local/etc/squid/errors/es-mx/ERR_ACCESS_DENIED:No such file or directory
What's that? NFS error? It was the only one, and when I repeated the upgrade, it didn't reoccur. In fact, the whole thing went remarkably smoothly. And suddenly the version mismatch was gone. Why?
|
ABC election
|
Topic: politics, opinion | Link here |
Scott Morrison, our fearless leader, has called what it seems is a general election (I would have called it a Federal election) for next month, finally a chance to get rid of him. The ABC promises to help with a site it calls Vote Compass, giving guidance as to which party would be most likely to fulfil my wishes of the government.
OK, sounds good. Fire up, enter post code as requested:
|
|
|||||||||
How do I know what “electorate” (the correct term is “electoral division”; I'm part of the electorate, but even the AEC mixes up the terms) I belong to? I had to go to the AEC site to find out: Ballarat.
OK, answer a whole lot of questions, press “Done” ... and got an empty screen. Repeatedly, and there were a lot of questions. Browser problems? Yes. It works on Microsoft, but not on FreeBSD, and for once it wasn't an out-of-date browser. How I hate this stuff!
|
Mowing the house forest
|
Topic: Stones Road house, general | Link here |
Nathan Delaney along today to mow the “house forest”, as planned. It was harder going than he (but not I) expected. Here in the middle of his action and after he had finished:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Yes, it looks like there's a lot left behind, but it's only about 10%. He couldn't get through there with the mower, so he'll be back with other tools. In the meantime we need to tidy away and burn the junk left behind.
|
More air conditioner issues
|
Topic: Stones Road house, general, opinion | Link here |
Before I went to bed this evening I turned the air conditioner from “cool” to “heat” in preparation for turning it on in the morning, when I expected it to be cool.
Things didn't go as planned. The wall controller for the lounge room displayed E5. OK, power cycle. E3 in the zone controller.
OK, power cycle from the zone controller. Not a real power cycle, but it should be a good start. It came back up indicating “cool” again, with the display 1 in the lounge room display, which I think is simply the zone number.
OK, turn to “heat” again. That worked, and there was no further issue.
From this chart,
|
|||||||||||||||||||||||
E5 means “Communication Error between Outdoor and Indoor units”, and E3 means “Room Sensor Error (open or short circuit)”.
Really? Clearly things are functioning again. I'd give the E5 more credence. But why did it display only on the wall controller for a specific zone? Why not in the zone controller, as the chart suggests? It's possible, of course, that the E5 displayed on all wall controllers; I'll have to check next time.
I now have three error conditions that look similar (all some kind of communication error between parts of the system), but are not localized. And my attempts on Saturday showed no indication that there was much wrong with the external sensor. Could we be looking in the wrong place?
| Thursday, 14 April 2022 | Dereel | Images for 14 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
igitt! Again!
|
Topic: technology, opinion | Link here |
It's time for the MFC of the commits to the FreeBSD repository that I made a couple of weeks ago. “MFC” (“Merge From Current”) is an old, worn-out magic TLA meaning moving the commits from the -CURRENT branch (thus the name) to the -STABLE branch, or whatever git calls it.
How does that work? In fact, it's quite straightforward, especially for git. Learn another redefinition. You don't need this to do the work:
|
|||||||||||||||||||||||
No, “cherry-pick” seems to be a US English redefinition of the word. OK, I can live with that. Find the hash, the random number that git uses to identify commits, and:
=== root@tiwi (/dev/pts/0) /usr/src 229 -> git cherry-pick -x 31e076d70b5f --edit
Auto-merging sys/dev/usb/usbdevs
hint: Waiting for your editor to close the file... Waiting for Emacs...
[stable/13 d45fa9f42406] Add devices and quirks for some Olympus cameras.
Date: Wed Mar 23 12:40:28 2022 +1100
2 files changed, 48 insertions(+), 1 deletion(-)=== root@tiwi (/dev/pts/0) /usr/src 230 -> git push
The authenticity of host 'gitrepo.freebsd.org (96.47.72.44)' can't be established.
ED25519 key fingerprint is SHA256:lNR6i4BEOaaUhmDHBA1WJsO7H3KtvjE2r5q4sOxtIWo.
Matching host key fingerprint found in DNS.
This key is not known by any other names
Are you sure you want to continue connecting (yes/no/[fingerprint])? yes
Warning: Permanently added 'gitrepo.freebsd.org' (ED25519) to the list of known hosts.
agent key RSA SHA256:S7sZHLcY4dgw53/rF70vrScdPuGef3enHdJzuYA1WDo returned incorrect signature type
Enumerating objects: 15, done.
Counting objects: 100% (15/15), done.
Delta compression using up to 4 threads
Compressing objects: 100% (8/8), done.
Writing objects: 100% (8/8), 1.37 KiB | 117.00 KiB/s, done.
Total 8 (delta 6), reused 1 (delta 0), pack-reused 0
To gitrepo.freebsd.org:src.git
7472cae2e098..d45fa9f42406 stable/13 -> stable/13
Yes, there are strangenesses there. What's the incorrect signature type? I'm guessing that it relates to my old ssh installation. But on the whole it worked quite nicely.
Twice. But I had three MFCs to do, and the third one failed badly:
|
|||||||||||||||||||||||
Aargh! This horrible half-coloured output again. Die! Die!
But I had already told git that it had a dumb terminal, and it didn't want to know. Parallel to the main problem discussed this one with Daniel O'Connor, who finally came up with this content for ~/.gitconfig:
[color]
pager = false
ui = false
advice = false
push = false
transport = false
And that finally shut the bloody thing up, at least for this specific output.
And the problem? With a sane view, it's:
=== root@tiwi (/dev/pts/0) /usr/src 229 -> git cherry-pick -x 3c9f304b383b --edit
Auto-merging bin/date/date.1
CONFLICT (content): Merge conflict in bin/date/date.1
error: could not apply 3c9f304b383b... Describe modern time zone handling.
hint: After resolving the conflicts, mark them with
hint: "git add/rm <pathspec>", then run
hint: "git cherry-pick --continue".
hint: You can instead skip this commit with "git cherry-pick --skip".
hint: To abort and get back to the state before "git cherry-pick",
hint: run "git cherry-pick --abort".=== root@tiwi (/dev/pts/0) /usr/src 234 -> echo $TERM
xterm-color=== root@tiwi (/dev/pts/0) /usr/src 235 -> TERM=xterm=== root@tiwi (/dev/pts/0) /usr/src 236 -> TERM==== root@tiwi (/dev/pts/0) /usr/src 237 -> git cherry-pick -x 3c9f304b383b --edit
error: Cherry-picking is not possible because you have unmerged files.
hint: Fix them up in the work tree, and then use 'git add/rm <file>'
hint: as appropriate to mark resolution and make a commit.
fatal: cherry-pick failed
What's that? After some searching, it proved that bin/date/date.1 really did have a conflict marked in the source:
-<<<<<<< HEAD
-The
-.Fl d , f , I , j , r , t ,
-and
-.Fl v
-options are all extensions to the standard.
-=======
With the exception of the
.Fl u
option, all options are extensions to the standard.
->>>>>>> 3c9f304b383b (Describe modern time zone handling.)
But what's that nonsense? I know conflicts like this when two people have committed to the same file, but that's not the case here; the HEAD version is directly replaced by my version. HEAD's log starts (in reverse sequence, of course):
commit 014ae00ef6edca2687d618e0bda138086a1e1230
Author: Mateusz Piotrowski <1mp@FreeBSD.org>
Date: Wed Nov 3 12:59:49 2021 +0100
What does that hash mean? Nothing at all. It's not obviously related to the commit sequence in CURRENT (sorry, main), here truncated to the headers:
commit 8deef8862883f2bb8e6a2ab9666b92d66738ec62
Author: Greg Lehey <groggyhimself@FreeBSD.org>
Date: Wed Mar 30 14:55:28 2022 +1100
commit 8e4b069f4555abd1906423f48440f63145f3c579
Author: Greg Lehey <groggyhimself@FreeBSD.org>
Date: Wed Mar 30 14:42:14 2022 +1100
commit 3c9f304b383bd378757ac33393b02f457947bc50
Author: Greg Lehey <groggyhimself@FreeBSD.org>
Date: Mon Mar 28 14:57:37 2022 +1100
commit c537bf9d5903d0689321f83691341c93b5919172
Author: Mateusz Piotrowski <1mp@FreeBSD.org>
Date: Wed Nov 3 12:59:49 2021 +0100
But they forgot to remove the date, so clearly 014ae00ef6edca2687d618e0bda138086a1e1230 in stable/13 is the same commit as c537bf9d5903d0689321f83691341c93b5919172 in head, and there is no conflict.
More to the point, how do I get rid of it? With subversion I could just remove the file and do svn up, but git doesn't want to know:
=== root@tiwi (/dev/pts/0) /usr/src 265 -> igitt up
This seems to be a src clone
...
Already up to date.=== root@tiwi (/dev/pts/0) /usr/src 269 -> git cherry-pick -x 3c9f304b383b --edit
Auto-merging bin/date/date.1
CONFLICT (content): Merge conflict in bin/date/date.1
And the funny thing is that it didn't even reinstate the file.
OK, the error message makes a couple of confusing suggestions: run git rm or git add (what?) or git cherry-pick with various verbose options. Somewhere there was also a git reset. Tried some of them:
=== root@tiwi (/dev/pts/0) /usr/src 270 -> git reset 01f095a35560
Unstaged changes after reset:
M bin/date/date.1
It seems that the M at the beginning of the line means the same thing that it does in svn: “File is modified”. And it's not even there! Which copy does it mean? OK,
=== root@tiwi (/dev/pts/0) /usr/src 271 -> git cherry-pick --continue
error: no cherry-pick or revert in progress
fatal: cherry-pick failed=== root@tiwi (/dev/pts/0) /usr/src 272 -> git cherry-pick -x 3c9f304b383b --edit
error: Your local changes to the following files would be overwritten by merge:
bin/date/date.1
Please commit your changes or stash them before you merge.
Aborting
fatal: cherry-pick failed
Somehow the more I flail around, the worse things get. What happens if I just check out a pristine source tree? It could work, though there's no reason to expect it: this was, in principle, a pristine tree, just with my first two MFCs added. Are they the issue? But if I do check out a new tree and things work, I may never understand what caused the current problems. So: onto the tuit queue for now.
|
45 GB throwaway data
|
Topic: technology, multimedia, general, opinion | Link here |
Volodymyr Zelenskyy made a total of 51 episodes of Servant of the people, of which I have 23 or 24: the first season should be 24 episodes, but I only got 23, the first of which is double length. So probably I have the complete first season.
OK, there are sources with better quality and all episodes. 45 GB! But when the National Broadband Network works, it works well, and I have unlimited data, so I was able to download them overnight at about 3 MB/s.
No subtitles, only original Ukrainian!
In fact I think it's Russian.
I can't do anything with that. Well, yes, I can: remove them again.
Looking back to the last millennium, it's interesting to compare things. That speed of 3 MB/s is roughly 4 times the maximum transfer rate of an IBM 3330. The 45 GB would fill 28 of these 1.8 m high cabinets, over 100 m in length if they were put end to end:
And if I had downloaded it with my network capabilities of 30 years ago (9.6 kb/s, 0.45 DM per kilobyte) it would have taken 15 months and cost a little over 20,000,000 DM, corresponding to about USD 25,000,000 nowadays.
And I threw it all away again. How times change.
|
Zoom: modern communications
|
Topic: technology, opinion | Link here |
I have to connect with a friend who uses Zoom. Why can't he use Whats App? That at least has a FreeBSD connection.
OK, install Zoom on my phone. It doesn't make me fall in love with it. First, I get:
|
|||||||||||||||||||||||
There's so much wrong with that!
OK, sigh and continue.
|
|
|||||||||
Another thing for illiterate people, though at least the right way round. But what have I done to the Zoom people that makes them want me to slide my finger over the phone until I get the year of my birth? No way.
And then it confirms:
|
|||||||||||||||||||||||
By chance that can't be misinterpreted, though probably I should have set something like 4 March 2001, which they could display as 4/3/1 or 3/4/1 or 1/3/4 with equal validity.
After that, I made the mistake of hitting “continue” before setting up the email ID that I had thought out for them:
Apr 14 10:22:54 eureka postfix/smtpd[70821]: NOQUEUE: reject: RCPT from www.lemis.com[45.32.70.18]: 550 5.1.1 <z00mer@lemis.com>: Recipient address rejected: User unknown in local recipient table; from=<bounces+15570388-5372-zoom=lemis.com@bounce-sg.zoom.us> to=<zoom@lemis.com> proto=ESMTP helo=<lax.lemis.com>
OK, no worries, “resend message”. Yes, it says, done. But it lies. I had to create and enter a new email ID before it would send me anything else.
OK, next a password. Enter. Press “Continue”. Dead in the water.
Another system that feels abandoned without Microsoft? OK, do the bloody thing all over again on distress. And how about that, with the same version of the same browser it produced popups telling me how it wanted to lessen the burden of people trying to crack the password: a certain length, specific characters and so on.
Finally I was done, and I got a meeting URL, for a purpose that hasn't been adequately documented. Now to contact the friend and see where we go from here.
Later I got a notification (or is that a message?) on my phone:
|
|||||||||||||||||||||||
What's that? Who sent it? While trying to confirm or deny that it was Zoom, I copied the screen shot to a real computer. And look, it's in the file name: Screenshot_2022-04-14-13-18-02-223_us.zoom.videomeetings.jpg. That's actually quite a useful way to work around this kind of documentation breakage in modern apps.
|
Stinkengriller, one last time
|
Topic: food and drink, opinion | Link here |
We still have some frozen „Schinkengriller“ sausages in the deep freeze. Helena Mirčić had brought along a potato salad, so it sounded like a good accompaniment.
|
|||||||||||||||||||||||
That's the last time. I have never been able to cook them without them all bursting (the second from the right, which I gave to Yvonne, had burst on the underside). The two remaining ones will come into some lentil dish.
It's interesting to note that ALDI has never repeated this item.
| Friday, 15 April 2022 | Dereel | Images for 15 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Greg Lehey, famous author
|
Topic: technology, general, opinion | Link here |
I continually get bombarded with emails from Academia asking me if I have written some paper or book. Usually I have, and I've already told them several times—there aren't many. And then there are these messages:
Subject: “G Lehey”: The name “G Lehey” was mentioned in a paper recently found by Academia written by Antti Kantee
Subject: The name "Greg Lehey" is mentioned by a Political Science researcher in a paper uploaded to Academia
Subject: “Greg Lehey”: 1 Mention in a paper uploaded to Academia by someone in Algeria
In every case, for more details I need to sign up for the full programme, something that isn't interesting enough. I would almost do it to check the veracity of these improbable claims, but so far I have resisted. My guess is that they're counting on false positives, but as it happens I appear to be the only Greg Lehey known to the web.
But today I got one that topped them all:
Date: Thu, 14 Apr 2022 17:19:26 +0000
From: Presentation and Evaluation of the IPSL“CM6A”LR Climate Model <updates@academia-mail.com>
Subject: Are you the Greg Lehey who wrote "Presentation and Evaluation of the IPSL“CM6A”LR Climate Model"?
Greg, is this publication yours? Help us keep your profile up to date.
Presentation and Evaluation of the IPSL“CM6A”LR Climate Model
By Olivier Boucher, Jérôme Servonnat, Anna Lea Albright, Olivier Aumont, Yves Balkanski, Vladislav Bastrikov, Slimane Bekki, Rémy Bonnet, Sandrine Bony, Laurent Bopp, Pascale Braconnot,Patrick Brockmann, Patricia Cadule, Arnaud Caubel, Frederique Cheruy, Francis Codron, Anne Cozic, David Cugnet, Fabio D'Andrea, Paolo Davini, Casimir Lavergne, Sébastien Denvil, JulieDeshayes, Marion Devilliers, Agnes Ducharne, Jean-Louis Dufresne, Eliott Dupont, Christian Éthé, Laurent Fairhead, Lola Falletti, Simona Flavoni, Marie-Alice Foujols, SébastienGardoll, Guillaume Gastineau, Josefine Ghattas, Jean-Yves Grandpeix, Bertrand Guenet, Lionel, E. Guez, Eric Guilyardi, Matthieu Guimberteau, Didier Hauglustaine, Frédéric Hourdin,Abderrahmane Idelkadi, Sylvie Joussaume, Masa Kageyama, Myriam Khodri, Gerhard Krinner, Nicolas Lebas, Guillaume Levavasseur, Claire Lévy, Laurent Li, François Lott, Thibaut Lurton,Sebastiaan Luyssaert, Gurvan Madec, Jean-Baptiste Madeleine, Fabienne Maignan, Marion Marchand, Olivier Marti, Lidia Mellul, Yann Meurdesoif, Juliette Mignot, Ionela Musat, CatherineOttlé, Philippe Peylin, Yann Planton, Jan Polcher, Catherine Rio, Nicolas Rochetin, Clément Rousset, Pierre Sepulchre, Adriana Sima, Didier Swingedouw, Rémi Thiéblemont, Abdoul KhadreTraore, Martin Vancoppenolle, Jessica Vial, Jérôme Vialard, Nicolas Viovy, Nicolas Vuichard
Now what kind of nonsense is that? It's an area in which I have no expertise whatsoever, and they've mentioned the names of every one of the plethora of authors. How many people did they send a similar message to?
|
Ukraine war: state of play
|
Topic: politics, language, opinion | Link here |
It's been over 7 weeks since Russia invaded Ukraine. The exact logic of the attacks remains a mystery, but it's clear that Putain greatly miscalculated. And if he had planned to occupy the country, you'd think that he would have taken more care not to completely destroy whole cities like Mariupol.
But where are things going now? Off to the east to destroy even more cities? Certainly they're taking their time, and in the process giving the Ukrainians time to do their own preparations. I'm sure I wasn't the only one filled with Schadenfreude when the Ukrainian sank the Moskva, Russia's flagship. As the ABC reported (my expansion):
Ukrainian authorities said Moskva featured in one of the landmark early exchanges of the war in Ukraine, when Ukrainian border guards on Snake Island, a small outcrop in the Black Sea, told the ship to "go fuck yourself" after it demanded they surrender.
Maksym Marchenko, the Ukrainian governor of the region around the Black Sea port of Odesa, claimed the Moskva had been hit by two Ukrainian-made Neptune anti-ship cruise missiles, which caused "very serious damage".
"It has been confirmed that the missile cruiser Moskva today went exactly where the border guards on Snake Island told it to go," he said in an online post.
Somehow the references don't quite tie up; my guess is that the translations were not as accurate as they might have been. Wikipedia quotes the original as “Русский военный корабль, иди на хуй”. Wikipedia translates that as “Russian warship, go fuck yourself”, and Google Translate (currently) translates it as “Russian warship, fuck you”, independently of whether it's told that the source language is Russian or Ukrainian.
And what is it really? I can read enough Russian to understand that иди на хуй means “go to ??”. And strangely, хуй isn't in my dictionary. Alone, Google Translate translates it from Russian as “fuck” and from Ukrainian as “cock”. Openrussian.org tells me that it means “dick, fuck, fucking”. So somehow I'm still missing the connection with what Maksym Marchenko claimed. Unfortunately the original of his quotation doesn't help much:
Підтвердили, що ракетний крейсер «Москва» сьогодні пішов саме туди, куди його послали наші прикордонники на острові Зміїний!
It has been confirmed that the missile cruiser "Moscow" today went exactly where it was sent by our border guards on Snake Island!
But apart from a certain humour that we've seen elsewhere amongst the Ukrainians—something that I wouldn't expect from the Russians—this is a slightly different story from what the Russians claim, that the damage was the result of an explosion on board.
Of course, there doesn't have to be any contradiction there. If you shoot a couple of missiles into a warship containing munitions, there's liable to be a fire. And really the Russians can't win: either they were hit by missiles, or they were incompetent enough to sink their own ship. And their dementi don't explain the fact that the other ships in the Black Sea retreated to a safe distance from the coast, nor that they bombed a factory in Kyiv where, apparently, the missiles were being made.
So where do we go from here? The Russians are clearly the stronger side. But what do they gain by destroying city after city? And their morale must surely be even lower now that their flagship has been sunk apparently without them being able to defend themselves.
Wouldn't it be time for Putain to produce another lie, say on 9 May: agree with Ukraine for cessation of hostilities in return for the remains of Mariupol and then proclaim a victory to the Russians who still believe him? Then the Russians could rebuild the city and the Ukrainians could move back.
Looking from a more global perspective, the question is whether any country should have veto power on the United Nations Security Council.
| Saturday, 16 April 2022 | Dereel | Images for 16 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Piccola and Melanie
|
Topic: animals | Link here |
Seen at breakfast:
|
|||||||||||||||||||||||
|
Laksa Kedah?
|
Topic: food and drink, opinion | Link here |
On Tuesday I bought some new noodles:
|
|||||||||||||||||||||||
“Laksa beras” means “rice laksa”, something to try out, though the detailed description gives the lie: it also contains sago starch, (sweet?) corn starch and tapioca starch.
But on the rear it has a quite detailed recipe for “Laksa Kedah”. Kedah (Darul Aman) is a Malaysian state on the mainland opposite Penang. I hadn't heard of Laksa Kedah before, but from the recipe it doesn't seem obviously different from Penang laksa.
|
|||||||||||||||||||||||
Wikipedia (currently) does know it:
Kedah Laksa (Laksa Kedah) is similar to Penang Laksa. The soup is usually made with eel instead of mackerel and quite differs, by the use of asam Gelugur instead of asam Jawa that is commonly used in Penang Laksa. As the main rice-producing state in Malaysia, Kedah Laksa uses rice flour to make laksa noodles. Sliced boiled eggs are usually added to the dish.
But the differences aren't obvious in the packaging. The noodles do contain other starches than rice, the fish is tenggiri (usually translated as mackerel or (possibly correctly) Spanish mackerel) and asam jawa, though it also includes a little asam gelugur. The recipe also doesn't include boiled eggs, though the photo on the packaging does.
Still, that's academic. I have no intention of making either Penang laksa nor Kedah laksa from scratch. The noodles are off-white, which may mean that they have a different taste.
|
Jane Ashhurst visits
|
Topic: general | Link here |
Jane Ashhurst, a friend of Yvonne's, is visiting us for a few days. We didn't have very big flowers to welcome her:
|
|||||||||||||||||||||||
| Sunday, 17 April 2022 | Dereel | Images for 17 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Baked beans: how many?
|
Topic: food and drink, opinion | Link here |
Bacon and eggs for breakfast today, and Jane wanted baked beans along with them. Not a problem. Oh yes, problem: how big is a serving? The obvious thing to do would be to look at the tin, but there is none. OK, after a bit of discussion on IRC (Andy Farkas: “I eat a 220 g tin”), asked Google. The hits were encouraging:
A 1/2-cup (130-gram) serving of baked beans...
According to NHS guidelines, a portion of 80g (cooked weight) beans and pulses,
approximately 3 heaped tablespoons, makes up one of your five-a-day.
There are 86 calories in 100 g of Heinz Baked Beans. Nutrition Facts. Serving
Size. 100 g. per serve
There are 110 calories in 1 serving (130 g) of Heinz Baked Beans
Australian made SPC Baked Beans Rich Tomato have 2 serves of vegies per 220g can
Average serving size is estimated to be 1/4 cup of uncooked beans (56.70 grams)
per person (approx. 1/2 cup or 113.40 grams of cooked beans). Based on this
serving size, one 110 lb. bag (50kg) of uncooked beans will yield approximately
880 servings.
I particularly like the last one, converted without thinking from Avoirdupois measurements (2 and 4 oz). But clearly my weights of 120 or (currently) 140 g are not to far off.
Well, maybe:
|
|||||||||||||||||||||||
That was 40 g left over of 140 g. I think I'll base further portions on 100 g. And, as I discovered too late, that's what I wrote in the recipe some time ago.
|
Removing power lines
|
Topic: photography, technology, animals, opinion | Link here |
Four years ago I was lucky enough to get an almost perfect scene from our lounge room window:
|
||||||||||||||||||||||||
But only almost perfect. Yes, both kangaroos have clearly visible joeys in their pouches. But that fence! How do I get rid of it? Over the months that followed I tried inpaint and movavi, and then a number of other products. They all had the same issue: identifying what you want to remove. So I gave up, waiting for a better product.
Now Skylum has come out with a new product, Luminar Neo, which claims to be able to do almost exactly what I'm looking for. They call it “remove power lines”, but clearly the software doesn't make that kind of distinction. Here from their advertising (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But somehow their web site is hard to understand. They seem to have two products, Luminar AI and Luminar Neo. It's not clear what the difference is, though it seems that Luminar Neo is newer, and it also seems that the line removal feature is only available in Neo. There also doesn't appear to be a free trial for Neo, only for AI.
OK, how about tutorials? The video they provide was singularly useless, advertising rather than description. But there are a number on YouTube, both for AI and Neo. They really didn't make the difference clear.
OK, download AI and see what I can do. Basically the same features that I had found in the other packages I looked at years ago. On first impression, navigation is just plain terrible. How do I tell which of these photos is Kangaroos-47 (my favourite)?
|
|||||||||||||||||||||||
Running the cursor over the images (which, like in all such software, get built up at a snail's pace) doesn't help. You have to click. That's a real pain.
OK, pick my favourite photo and remove the fence post. That went well. Here the original, the result with inpaint and then Luminar:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
|
|||||||||
|
|
||||||||||
It's hard to compare, but there's more evidence of the missing post in the image from inpaint: the Luminar results look better. But we're not done yet. Now remove the fence lines in front of the horses:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
In more detail:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Yes, the line in front of the horses is gone, but the replacement is very bad. Luminar AI was not able to recognize the lines by themselves: I had to mark them, and that's very difficult to do accurately. Can Luminar Neo do better? I get the impression, but I haven't seen any evidence. The overall level of documentation and ease of use is not good enough for me to spend $140 odd on the off chance. Do they have a free trial? It's hard to say, potentially another documentation issue.
That's a pity, because Skylum is based in Kyiv, and I would have been happy to buy another Ukrainian product.
|
More air conditioner insights
|
Topic: Stones Road house, general, opinion | Link here |
It was relatively warm today, and I had the air conditioner on in my office. And somehow it seemed particularly cool. Looking at the wall controller, it showed 23.4°, although it was set to 24°. And the indicator showed that it was cooling strongly. OK, turn it off, and of course it stopped.
But how could that happen? Are there two different temperature sensors? Is maybe that the meaning of the E25 errors? An ambient temperature sensor in the house rather than in the external unit? That would go along with the fact that the connectors outside made no difference, that the temperature settings in the house are inaccurate—something that I ascribed to inadequate heating or cooling capability—and the E3 and E5 status codes that I experienced on Wednesday.
| Monday, 18 April 2022 | Dereel | Images for 18 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More Luminar investigations
|
Topic: photography, technology, opinion | Link here |
Yesterday's investigation of Luminar didn't produce any spectacular insights. Is Luminar Neo so much better than Luminar AI? Should I buy it on the off chance? That way madness lies.
Off searching on the web and came up with The complete Photofocus guide to Luminar Neo, which certainly gave me some interesting insights. It addresses my “power line” issue:
This tool is effective when the power lines are up against the sky, but if there are lines in front of architecture, foliage or other intricate details, the results can be unpredictable. Overall, I found that Remove Power Lines lessened the time I’d spend manually removing power lines, but didn’t eliminate the need for some manual brushwork.
That seems reasonable, but not a silver bullet. And the comments were less reassuring:
The power line removal tool yields inconsistent results at best – even with power lines against clear sky.
What about the manual? Off to look for it. I wasn't the first. From a forum post titled Lunimar [sic] Neo user manual:
Where is the manual? Not where it should be i.e. "Luminar Neo" "Help" "Users Manual". Somebody knows?
A couple of responses, clearly from Skylum. The most recent, from Katherine Nelson at the beginning of the month, reads:
Let me assure you that we are doing our best to add a user manual as soon as possible. But we are also doing everything we can to make sure every member of our team is safe, and for that reason, at this time, we do not announce any exact dates. We appreciate your understanding!
Yes, I can appreciate that the current situation in Ukraine is concerning, but this product was released, without a manual, before the hostilities began.
Then I found this video by Scott Davenport:
And that's actually quite good. It shows that Luminar Neo really can remove most power lines automatically. But that's “most”. What about the rest? If I understand the video, there's no way to go back and retouch things. In general, though I haven't had enough experience with the product, it seems that the approach is wrong: lots of “tools”, but with Luminar AI you can only apply each of them once per image. Luminar Neo has the great advantage that you can apply them more than once That sounds strange to me, and a good reason not to buy AI.
The whole situation smells of a product released in a hurry and before it's ready. I'll keep an eye on it; it has potential.
And yes, of course I looked for other software that can do the same thing. Things don't seem to have changed much in the last three years.
|
“The Royals”: parallel history
|
Topic: history, opinion | Link here |
Last month I started watching a series Royals: Keeping the Crown, and at the time I commented that history was presented in a very different way from my understanding.
Today I started watching the third episode, Post-War Amnesia. The summary is illuminating:
World War II may be over, but it still casts a long shadow over royal families, particularly those with links to fascism; in Britain, King George VI works to keep his brother's closeness to the Nazis away from the British people.
They mention the Marburg Files, discovered not in Marburg, but in a little village nearby. Well, that's what they say. In fact the papers were discovered in Degenershausen Estate (parse Degener - s - Hausen), near Falkenstein, about 280 km away. But what's that to an Englishman? Almost exactly the same distance as London to Exeter.
But was Edward VIII really a Nazi? What I saw of the papers suggested that at the most he had sympathies, and the Wikipedia page Operation Willi agrees.
But wait, that's not all. After all that, Princess Elizabeth fell in love with a foreigner, a Greek prince with many Nazis in his family. Round about here I begin to wonder how tenable the claims are. Yes, Philip of Greece was a Greek prince, but he was also Elizabeth's third cousin. Like Elizabeth he was of primarily German descent.
And the Nazis? His sister was a member of the Nazi party before the war. But she died in 1937. So why should there be any concern about Philip's connection with the Nazis in the late 1940s? Ah, but it sells documentary films.
All of this is conjecture. But when they talk about the British Empire falling apart, they produce this map:
|
|||||||||||||||||||||||
India (cross-hatched) leaves the Empire, leaving Pakistan (red) behind. Where did they get that idea from? They also talk about a civil war in Malaya, the “Garden of the East”, the first time I had heard that term. And Malaya was also cross-hatched, although that didn't happen for another 10 years, long after the time frame of this episode.
So: what should I think of the series? Clearly it's highly inaccurate, and I suspect it intends to be sensationalist.
| Tuesday, 19 April 2022 | Dereel | Images for 19 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Upgrading tiwi
|
Topic: technology, opinion | Link here |
In preparation for another Mediathekview port attempt today, built a new system on tiwi. First, of course, git to update the source tree:
=== root@tiwi (/dev/pts/1) /usr/src 3 -> git pull --ff-only
fatal: unsafe repository ('/eureka/home/src/FreeBSD/git/stable-13' is owned by someone else)
To add an exception for this directory, call:
git config --global --add safe.directory /eureka/home/src/FreeBSD/git/stable-13
Huh? What does that mean. /usr/src is a symlink:
=== root@tiwi (/dev/pts/1) /usr/src 4 -> ls -l /usr/src
lrwxr-xr-x 1 root wheel 26 28 Mar 15:43 /usr/src -> /src/FreeBSD/git/stable-13
Previously I had updated it from dereel, and that still works:
=== root@dereel (/dev/pts/0) /usr/stable-13 25 -> git pull --ff-only
remote: Enumerating objects: 244, done.
...
12 files changed, 72 insertions(+), 20 deletions(-)
What's the point of restricting it? And how does it tell? Clearly it can be fixed, but I wish I understood this stuff.
|
MediathekView again
|
Topic: technology, multimedia, opinion | Link here |
Another attempt at getting the latest version of MediathekView to work on FreeBSD. Last time things failed with an initialization error:
Exception in thread "main" java.lang.ExceptionInInitializerError
...
Caused by: java.lang.NullPointerException: Cannot read the array length because "this.screens" is null
OK, that was with the FreeBSD Java installation. It should be compatible, but who knows? MedithekView comes with its own Java system, and maybe there's a good reason for this. Run that instead? How do I do that? It's a Linux program, so first install Linux compatibility. How do I do that? Took a look in the handbook and discovered that since I last looked at this, they have changed the base to CentOS 7. pkg search? Comes up with nothing. Ah, fool, you should know that the package is called linux-c7, one of the 109 packages that start with that name.
Install, then run java:
/usr/ports/multimedia/mediathekview/work/MediathekView/jre/bin/java: error while loading shared libraries: libjli.so: cannot open shared object file: No such file or directory
Why not? It's there. Presumably java is looking in the wrong place. It's in /usr/ports/multimedia/mediathekview/work/MediathekView/jre/lib, matching the executable /usr/ports/multimedia/mediathekview/work/MediathekView/jre/bin/java. Off looking for solutions, and found amazingly stupid answers, including setting symlinks to where the executable might expect to find it. Surely there's an environment variable? The search wasn't very helpful, but there is a variable JAVA_HOME. OK, try this::
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 40 -> export JAVA_HOME=`pwd`/MediathekView/jre/=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 41 -> $JAVA_HOME/bin/java -jar MediathekView.jar
. Configuring for non-portable mode
Exception in thread "main" java.lang.ExceptionInInitializerError
...
at java.desktop/java.awt.GraphicsEnvironment.getLocalGraphicsEnvironment(GraphicsEnvironment.java:106)
at java.desktop/sun.awt.X11.XToolkit.<clinit>(XToolkit.java:224)
at java.desktop/sun.awt.PlatformGraphicsInfo.createToolkit(PlatformGraphicsInfo.java:40)
at java.desktop/java.awt.Toolkit.getDefaultToolkit(Toolkit.java:599)
at javafx.embed.swing.JFXPanel.initFx(JFXPanel.java:218)
at javafx.embed.swing.JFXPanel.<init>(JFXPanel.java:258)
at mediathek.Main.initializeJavaFX(Main.java:432)
at mediathek.Main.main(Main.java:336)
Caused by: java.lang.NullPointerException: Cannot read the array length because "this.screens" is null
at java.desktop/sun.awt.X11GraphicsEnvironment.initDevices(X11GraphicsEnvironment.java:235)
at java.desktop/sun.awt.X11GraphicsEnvironment.rebuildDevices(X11GraphicsEnvironment.java:216)
at java.desktop/sun.awt.X11GraphicsEnvironment.<init>(X11GraphicsEnvironment.java:201)
at java.desktop/sun.awt.PlatformGraphicsInfo.createGE(PlatformGraphicsInfo.java:36)
at java.desktop/java.awt.GraphicsEnvironment$LocalGE.createGE(GraphicsEnvironment.java:93)
at java.desktop/java.awt.GraphicsEnvironment$LocalGE.<clinit>(GraphicsEnvironment.java:84)
... 8 more
===
Exactly the same error! Do I really have the new Java libraries? How can I tell. How about asking the executables?
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 21 -> $JAVA_HOME/bin/java -version
openjdk version "17.0.1" 2021-10-19
OpenJDK Runtime Environment Temurin-17.0.1+12 (build 17.0.1+12)
OpenJDK 64-Bit Server VM Temurin-17.0.1+12 (build 17.0.1+12, mixed mode)=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 22 -> java -version
openjdk version "17.0.1" 2021-10-19
OpenJDK Runtime Environment Temurin-17.0.1+12 (build 17.0.1+12)
OpenJDK 64-Bit Server VM Temurin-17.0.1+12 (build 17.0.1+12, mixed mode)
Exactly the same? That sounds like I did something wrong. Ran a ktrace to see what it was really looking at, and yes, it really was looking at those libraries. Of course! JAVA_HOME was still set. Without it we get:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 55 -> JAVA_HOME= /usr/local/bin/java -version
openjdk version "17.0.2" 2022-01-18
OpenJDK Runtime Environment (build 17.0.2+8-1)
OpenJDK 64-Bit Server VM (build 17.0.2+8-1, mixed mode, sharing)
Still very close, but different. So whatever the problem is, it's not related to the Java version. Where do I go from here?
|
British Royal Family, revisited
|
Topic: history, multimedia, opinion | Link here |
Yesterday I ranted about the perceived inaccuracy of “Royals: Keeping the Crown”. And mainly by coincidence, today I saw an episode of another series, The Royal House of Windsor, covering exactly the same period in time.
What a difference! The concerns of the British people weren't about the perceived Naziism of the Royal Family or the Duke of Edinburgh, who suddenly was no longer Greek but Elizabeth's cousin: the people were concerned about the austerity measures and the cold winter, and the king was mainly concerned about maintaining the relevance of the crown. Again, I'm not sure how accurate it is, but it sounds a whole lot more plausible than “Royals”. I'm not convinced that it's overly accurate, but what a difference!
|
Lamb roasting times?
|
Topic: food and drink, opinion | Link here |
Roast lamb for dinner this evening, a somewhat late nod to Easter. We don't cook it very often, and I'm still not sure about cooking times. Last time I noted that the joint should start at room temperature, and I did that today. But it also had a bone in it; how does that affect things? It weighed 1.4 kg, and based on previous experience I expected it to be done (60°) in 75 minutes in a fan-forced oven at 180°.
But the thermometer went off round 50 minutes! Did I misposition it? Pushed it in a little and watched the temperature drop to 52°. Again it went off before I was ready. Pushed it in a bit more. Still only showing 57° after 75 minutes!
Clearly measuring temperature is not the way to go here. I took it out, left it to rest for 30 minutes, and it was done, in fact possibly a little overdone. It tasted fine, and all were happy with the results—except me. I think this might really be a case where temperature measurements just don't cut it. Of course, the alternative is to buy a joint without the bone, and that's what we'll do next time.
|
Academia revisited
|
Topic: technology, general, opinion | Link here |
Chris Bahlo around for dinner this evening, and since she's a real live academic, I asked her about Academia.edu. Yes, she too gets the same kind of mail from them, though not as bad as I had last week, and it irritates her too. But she did have some insight: she once got a mail asking if she was the author of a paper co-authored by Tony Papenfuss. No, no such paper. But she's not the only Bahlo in academia, and her sister Melanie Bahlo is married to Tony. So clearly Academia is mixing up first names. And that could explain some of the bombardment. Suddenly false positives become a possibility. And it seems that Chris has at least one: the two hits I see there are both in Switzerland and could be the same person. And despite the unusual surname, Tony has a Tony Papenfuss Doppelgänger in the USA.
| Wednesday, 20 April 2022 | Dereel | Images for 20 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Interesting PV status
|
Topic: Stones Road house, general, opinion | Link here |
Seen in the PV inverter output today:
tstamp Pac Status SOC VBat PacGrid PacBat PacPV FromPV W1
2022-04-19 15:40:10 0 Internal Bypass Closed 28 265 NULL 0 0 0 11902
...
2022-04-19 15:40:51 0 Internal Bypass Closed 28 265 NULL 0 0 0 11591
2022-04-19 15:40:56 9 Internal Bypass Closed 28 265 NULL -218 227 0 11304
What's that? The clue is in the last column (W1), the power from the grid: up to nearly 12 kW! That may not seem like that much, but neither the inverter nor the grid can handle that much (a maximum of 11.5 kW in each case). So now I know what the inverter does under such conditions; only a little more and presumably the main switch would have tripped.
Why so much power? Air conditioner, washing machine, two dishwashers, dryer... In Germany there are continuous flow water heaters that consume as much as 33 kW. Even small ones consume 18 kW. But we knew of the restrictions when we built the house, and it's surprising that it happened at all; at least it's the first time since we installed the PV system three years ago.
|
Garden flowers in mid-autumn
|
Topic: gardening | Link here |
It's a month after the equinox, time for the monthly garden photos. The year's coming to an end, and there's much less to see in the garden. But since the attention it has been getting from many people, it's beginning to look less like a disaster area. Here the middle bed to the north of the house in October and now:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, there was much more in October, but much of it was weeds. And gradually the garden is getting past the destruction of Bryan Ross. Here the Persicaria odorata (daun kesum) in October and now:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
And the volunteer tomatoes are fruiting, though far too late in the year to make much difference:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The Abutilon cuttings that I took in January last year have done well, particularly after I planted one of them:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Until then, the other one was growing better, but it's way behind now:
|
|||||||||||||||||||||||
I'll have to find a place to plant it.
In the east garden, the trees are doing well enough. The Schinus molle that gave me concern a year or so ago is not exactly growing furiously, but it's still alive:
|
|||||||||||||||||||||||
And the Grevillea robusta that looked half dead last summer has recovered to some extent; at least it's now larger than when we planted it (first photo):
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
But that has been nearly 3 years. Still, that area seems to be recovering:
|
|||||||||||||||||||||||
Apart from the many volunteer wattles, there's also another Schinus molle, a birch and a struggling Salix babylonica that Magda Delvar gave me over 5 years ago.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And the Buddleja x weyeriana that almost died have recovered well, and the first flowers are showing:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Even the tiny twigs that I cut off are now growing well, and the first bud is showing:
|
|||||||||||||||||||||||
Only 2½ months ago they looked like this:
|
|||||||||||||||||||||||
| Thursday, 21 April 2022 | Dereel | Images for 21 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Animal destruction
|
Topic: general, animals | Link here |
Puppies love to chew, and Larissa and Lena are no exception. Here the current state of Yvonne's bed:
|
|||||||||||||||||||||||
But there's more. We've already seen that the once flat surface of the dog run has developed a number of craters up to 50 cm deep:
|
|||||||||||||||||||||||
And yesterday Yvonne had the idea of putting an old mattress in their kennel, since it was wet. Great idea, thought the dogs:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Fortunately they're growing up.
|
Mariupol: liberated!
|
Topic: history, politics, opinion | Link here |
I get the impression that Putain is changing his tune. After setting not one, not two, but three deadlines for the last Ukrainian militants remaining in Mariupol to surrender, he has now decided no, there's no need: we have liberated Mariupol. Leave the militants where they are.
Liberated? Yes, liberated. The city is largely free of undamaged buildings and population. How many are left? Who knows? Some say 100,000, others only 10,000. Either way it's only a fraction of the 2021 census estimate of 431,859. And according to the press the city has been reduced to rubble.
So: what happens now? The Russians have taken possession of an unpopulated scrap heap with known militants waiting in the wings. What are they going to do? My guess is that Putain will celebrate the “liberation” on 9 May, but then? The Russians rebuild Mariupol, then the Ukrainian population comes back and makes life unbearable for the Russians, so the Russians leave again?
The other thing of interest is that he has taken a leaf out of Kim Jong-Un's copybook and done a (successful, of course!) missile test. Yes, a little bigger than Kim's toys—in fact, it seems, the biggest ICBM ever built—but that's literally overkill. Who cares?
But where do we go from here? What Putain has achieved is the rupture of ties with the West, including Nord Stream 2. That will haunt them for decades after the present conflict is over. And getting NATO away from their borders? So far the only change I see is that Sweden and Finland are seriously considering joining.
And what does he get in return? Some have suggested that he might commit suicide, either by firing a number of bullets into the back of his head, jumping out of a high storey window after carefully locking it from inside, or both.
| Friday, 22 April 2022 | Dereel | Images for 22 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Academia surpasses itself
|
Topic: technology, general, opinion | Link here |
I've commented a number of times recently about Academia.edu, most recently on Tuesday. But they're not letting up. Today I received a message that surpassed all I have had so far:
|
|
|||||||||
What does that mean? About the only thing that's clear is that the title has nothing to do with me. Why do they come up with this stuff?
|
Another fence removal?
|
Topic: photography, technology, opinion | Link here |
Mail from Kevin Koster today:
Your recent diary entries talking about removing wire fences from a photo reminded me of playing with a free open-source program called Seam Carving GUI many years ago. There doesn't seem to have been any new development since, with the last version (from 2009 I believe, based on file dates) mirrored here: https://github.com/tari/seam-carving-gui.
Most of the stuff about seam carving on the web talks about resizing images, but the program also has the function of removing highlighted objects. You mark the areas to remove with "mark area for removal" option selected on the right, then click on the "Remove" button.
He included a couple of attempts that showed both the advantages and disadvantages of the software. The advantages were that it (almost) worked, The disadvantage was that it almost worked, as he pointed out in advance. Here two steps in the process (run the cursor over an image to compare it with its neighbour):
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
As Kevin says,
Obvious problems are that the roo on the right gets its tail chopped short, and the horses are getting squashed down. Maybe avoidable with more experimentation, maybe not.
Comparing the images (with mouseover in the HTML version), it seems that seam carving simply removes slices of the image and joins the rest together. Under those circumstances, it's surprising that the result looks so good. But, once again, it's probably not what I'm looking for.
| Saturday, 23 April 2022 | Dereel | Images for 23 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another pump failure!
|
Topic: Stones Road house, general, opinion | Link here |
Got up this morning to be told by Yvonne that we had had a “partial power failure”: a circuit breaker had tripped. Yes, and it stayed tripped.
We've seen that before: the old el-cheapo drainage pump that I replaced with a much more expensive one 6 months ago. And now, it seems, it too has failed. Disconnect it and all was well. At least it wasn't wet, and though it—typically—failed on early Saturday on a long weekend, I can wait until Tuesday to get it fixed.
|
Flowers in sunshine
|
Topic: photography, gardening, opinion | Link here |
Taking photos of the garden flowers is something that works best in overcast weather. When the sun is shining, the contrast is too high, and things don't look as good.
But of course I have a method for handling high contrast: HDR. How does that compare? By default I take three images with exposure offsets +4.7 EV, +1.7 EV and -1.3 EV. Today I tried that, but also a set with +3, 0 and -3 EV. The middle one of that set is, of course, a perfectly normally exposed image.
The results? First the normally exposed image, then the 3/0/-3, then the 4.7/1.7/-1.3 (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's amazing! There's almost no difference between the 3/0/-3 and a single image! What have I done wrong?
Looking at these images a year later, it's clear that they're not what I thought they were. The first two images appear to be the same image, at 1/640 s and f/5.6, the third at 1/250 s and f/5.6 (about 1.4 EV more exposure), and apparently the missing image from the pair below. And that's what they look like.To do this right, I not only need the correct images, but also the correct postprocessing: none. In this case, they look like:
Image title: Gazanias unprocessed3 3 1EV Dimensions: 5273 x 3888, 2999 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 23 April 2022, thumbnails All images taken on Saturday, 23 April 2022, small Diary entry for Saturday, 23 April 2022 Complete exposure details
Image title: Gazanias unprocessed3+0EV Dimensions: 5273 x 3888, 6707 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 23 April 2022, thumbnails All images taken on Saturday, 23 April 2022, small Diary entry for Saturday, 23 April 2022 Complete exposure details
Image title: Gazanias unprocessed3 3+1EV Dimensions: 5273 x 3888, 4369 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 23 April 2022, thumbnails All images taken on Saturday, 23 April 2022, small Diary entry for Saturday, 23 April 2022 Complete exposure details
Now how did I make that mistake?
The 4.7/1.7/-1.3 looks better, and there the middle image is not as good:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But somehow I get the feeling that the shadow detail is wanting. By chance I ran into the same problem 10 years ago, but since then I thought that my technique had improved. Certainly my software has changed.
The other issue I have here is that I don't know the point I was trying to make.
| Sunday, 24 April 2022 | Dereel | Images for 24 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Walking with George again
|
Topic: animals, opinion | Link here |
Peggy Naumov and George along today for the first walk with our dogs since Larissa's arthritis problems.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
George is much more relaxed, and he approached both me and the dogs with less reserve than previously. When we got back we let him loose with Lena, but they were obviously too tired:
|
|||||||||||||||||||||||
In the evening, though, Lara showed some signs of joint pain. We shouldn't overdo these things. Today she got a shot of Meloxicam; hopefully it'll be a one-off.
|
MediathekView: done!
|
Topic: technology, multimedia, opinion | Link here |
After my lack of success with MediathekView, most recently 19 April, I identified a forum for MediathekView (or MV, as it proves they call it) developers and asked for help.
FreeBSD? Who cares about that? But the message in the middle of the verbose stack trace is clear:
Caused by: java.lang.NullPointerException: Cannot read the array length because "this.screens" is null
at java.desktop/sun.awt.X11GraphicsEnvironment.initDevices(X11GraphicsEnvironment.java:235)
That's Java's way of saying “No screens found”. Clearly a FreeBSD problem.
More discussion with little of use. “Which version of FreeBSD? < 13?”. Presumably the version is more of an issue with Linux, but clearly it's not the case here. More to the point: I was building the port on tiwi but (trying to) display it on eureka. That works fine with the old version of MV, but could they have broken things? Let's start X on tiwi:
/usr/local/bin/X: Only console users are allowed to run the X server
I've seen that before. Ah, the instructions: add entries to /etc/X11/Xwrapper.config. Oh, no, old worn-out magic path. Make that /usr/local/etc/X11/Xwrapper.config:
# $Id$
# Make X do what it should
allowed_users=anybody
needs_root_rights=yes
Can't access it from eureka! Ah, there's the old issue with -listen-tcp. Try that:
startx -- -listen tcp
And we got beyond the screens issue:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 41 -> java -jar MediathekView.jar
. Configuring for non-portable mode
Apr 24, 2022 12:42:49 PM com.sun.javafx.application.PlatformImpl startup
WARNING: Unsupported JavaFX configuration: classes were loaded from 'unnamed module @6d195aa5'
. Command line parse error:
java.lang.UnsupportedOperationException: FreeBSD is not supported
Huh? This is the FreeBSD version of Java! OK, try the version delivered with MV:
=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 42 -> JAVA_HOME=/usr/ports/multimedia/mediathekview/work/MediathekView/jre/=== root@tiwi (/dev/pts/0) /usr/ports/multimedia/mediathekview/work/MediathekView 43 -> java -jar MediathekView.jar
. Configuring for non-portable mode
Apr 25, 2022 12:44:34 PM com.sun.javafx.application.PlatformImpl startup
WARNING: Unsupported JavaFX configuration: classes were loaded from 'unnamed module @1d119efb'
GLib-GIO-Message: 12:44:34.350: Using the 'memory' GSettings backend. Your settings will not be saved or shared with other applications.
. Programmstart: 2022-04-25T12:44:34.386848035
. Version: 13.8.1
...
And as far as I could see, it works. First I need to configure it, but nothing seems to have gone wrong. “It compiles! Ship it”.
Well. There's a lot to look at. Why does the new version of MV no longer work across a network? Why does it claim not to work on FreeBSD? More questions for the developers, but I have a feeling that I'll have to answer them myself.
|
DxO virtual copies
|
Topic: photography, technology, opinion | Link here |
DxO PhotoLab has a strange feature called “virtual copy” of an image source file. What's it for? I don't know, and I don't really care. How do you get one? I don't know, and I only care because I sometimes make them accidentally. How do you get rid of them? I don't know, but I really care.
Processing yesterday's photos, I somehow managed to create virtual copies of every one of 144 source images. How do I get rid of them? I can select the rubbish bin icon on each image, but as they say in German, I can rivet pancakes to my ears and claim to be a Dachshund. Removing all files from the directory didn't help. Stopping the program didn't help. Somehow DxO is keeping track of them, but it isn't in the .dop file.
Off searching, bringing little of help beyond the suggestion to re-sort the images so that the copies were together. Then mark them all and delete them. How? I suppose I should RTFM, but my recollection is that it's particularly painful.
| Monday, 25 April 2022 | Dereel | Images for 25 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More Academia strangenesses
|
Topic: technology, opinion | Link here |
Somehow Academia.edu has been particularly active lately, and it has been the source of some surprises. Today I got a message asking “Are you the Greg Lehey who wrote "Closed Source Fights Back"?”.
Am I? Possibly. How do I find out? It proved to be about the SCO fiasco, but I didn't have anything in that directory that matched. It took quite some time to find out that it was an article that I wrote for ACM Queue in 2003. So yes, my article. But how much time it took to find!
Of course, Academia is sure that I've written other stuff too. It then asked me if I had written any of these papers:
|
|||||||||||||||||||||||
Clearly “The Complete FreeBSD” is mine. The subtitle “Documentation from the source” was thought out by O'Reilly and Associates. And clearly “IPSL-CM5A-MR model output prepared for CMIP5 RCP8.5, served by ESGF”, whatever that is, is not by me. The list included a number of titles containing the string “ipsl”, such as “cmip5 output1 ipsl ipsl cm5a lr decadal1960 mon aerosol aero r3i1p1 v20111119 sconcso4”, “cmip6 cmip ipsl ipsl cm6a lr” and “Improved near surface continental climate in IPSL‐CM6A‐LR by combined evolutions of atmospheric and land surface physics”. What does that say about their recognition algorithms?
But what about “Evolution of a free software project”? That could well be by me, but possibly also by many other people. How do I find out? I don't. I have the option of clicking on “This Is Me [sic]”, but that's all. Clicking on the title does nothing obvious.
While searching for the authorship of “Closed Source fights back”, came across a number of other things that I had forgotten about. Ageing AUUG faces to disband: “The Australian Unix and Open Systems Users Group (AUUG) today [14 September 2006] faced a call from past president Greg Lehey to dismantle the 31-year-old organisation.”. That sounds like me, and the date helped me find it. Well, almost. My reference to the letter was http://www.auug.org.au/pipermail/auug-announce/2006/000241.html, the same as in the ZDnet article. It's dead, Jim, and it took me a long time to find where I had saved my local copy: in auug-announce.
And how about that, I had forgotten all about it. But reading it again, I think it was correct to send the message, especially as there were almost no replies. Also interesting was a proposal by Jeff Waugh, suggesting merging AUUG and Linux Australia. A good idea, but somehow nothing came of it.
The real question, though: why have I forgotten all this?
| Tuesday, 26 April 2022 | Dereel | Images for 26 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Tortilla composition, yet again
|
Topic: food and drink, opinion | Link here |
When I started making my own tortillas, I had a lot of trouble with the composition, and on 7 October 2011 I started a detailed page on the topic.
Over the years the masa:water relationships changed greatly, between 1:1.29 and 1:1.8. And then on 16 February 2021 I started the first of 5 bags of « El Maizal » masa and discovered that it required a ratio of 1:1.45. But only the first bag. The next one wanted 1:1.6. And so did the one after that.
Why? I gradually came to the conclusion that the exact quantity wasn't that important. So today, when I opened another bag, I used 1:1.6 and waited for it to become firm. It didn't. With the experience of the last decade, I was able to make some tortillas anyway, but they showed the problem:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
In the second image I left the paper on the tortilla until it had cooked slightly, in the fulfilled hope that it would then not disintegrate as easily. But clearly I need to change the composition yet again. 1:1.5?
|
Fixing the pump
|
Topic: Stones Road house, general | Link here |
Called up UPI today about the sump pump. They're another modern company who make you search for the phone number, though they slipped up and showed a photo of their delivery vans with the phone number. For reference, Google does better.
Explained the issue to Kylie, who called Nick and didn't give him the details. He'll send somebody out “later this week”. Fortunately no rain is forecast.
|
Recovering from dog damage
|
Topic: animals, opinion | Link here |
Yvonne has decided that the mattress that she gave the dogs is a danger to them in its present condition. She could be right:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It'll go to the dump tomorrow.
|
Core team again?
|
Topic: technology, opinion | Link here |
Nearly a quarter of a century ago I nominated myself for the FreeBSD Core Team and was elected. But after a few years, along with my other work, it seemed too much trouble, and I stepped back in 2004.
Now elections are coming up again. I have more time, and there are surprisingly few nominations. Should I stand again? I have until 4:00 on Thursday morning (so effectively tomorrow evening) to decide.
|
Next Academia surprise
|
Topic: technology | Link here |
After yesterday's acknowledgement of ownership of “Closed Source Fights Back”, it wasn't surprising that I got a message from Academia telling me:
Someone confirmed that your paper, "Closed Source Fights Back" mentions them.
Yes, that's the “pay us money and we'll tell you who” message that I continually see. But no, this time it was straightforward and presumably correct:
Peter H. Salus recently confirmed that your paper Closed Source Fights Back mentions them.
|
Updating dereel
|
Topic: technology, opinion | Link here |
So now I have MediathekView running on tiwi. Before I commit it, I need to have it running on FreeBSD-CURRENT. That's what dereel is for. But first it needs to be up to date. OK, pkg upgrade:
===== Tue 26 Apr 2022 15:05:31 AEST on dereel: pkg upgrade
Updating FreeBSD repository catalogue...
Fetching packagesite.pkg: .......... done
Processing entries:
Newer FreeBSD version for package rbllookup:
To ignore this error set IGNORE_OSVERSION=yes
- package: 1400056
- running kernel: 1400053
Ignore the mismatch and continue? [y/N]:
Huh? I have a -CURRENT kernel built earlier this month. The OSVERSION can't have incremented that much in that time, can it? In any case, ignored it and continued.
Next, Linux compatibility. I've already established that the port is called linux-c7. More time wasted:
===== Tue 26 Apr 2022 17:35:40 AEST on dereel: pkg install linux-c7
Updating FreeBSD repository catalogue...
FreeBSD repository is up to date.
All repositories are up to date.
The following 83 package(s) will be affected (of 0 checked):
...
The process will require 897 MiB more space.
222 MiB to be downloaded.
...
That will take its time. Mañana.
| Wednesday, 27 April 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
dereel: next upgrade
|
Topic: technology, opinion | Link here |
Yesterday I started installing Linux compatibility on dereel. This morning I read:
...
[83/83] Fetching linux-c7-alsa-plugins-pulseaudio-1.1.6_1.pkg: ...... done
Checking integrity... done (0 conflicting)
[1/83] Installing linux_base-c7-7.9.2009...
Cannot install package: kernel missing 64-bit Linux support
pkg: PRE-INSTALL script failed
Now isn't that clever? First download all the stuff, then complain that the kernel can't install it. Clearly it wants:
=== root@dereel (/dev/pts/7) /home/grog 40 -> kldload linux
OK, next time should be faster. But no, for some reason it wanted to download everything all over again. Why? These things get stored somewhere, but I don't recall where. While waiting for the download, checked in the man page pkg(8):
FILES
See pkg.conf(5).
Somehow these pages are all a can of worms. OK, check pkg.conf(5). No FILES section! Another man page to fix.
And at the end?
[1/83] Installing linux_base-c7-7.9.2009...
Cannot install package: kernel missing 64-bit Linux support
pkg: PRE-INSTALL script failed
FOOL. You loaded linux.ko, not linux64.ko. Try again. Just to confuse me, it worked this time. But there's another bug: what good is linux.ko if it's not enough to run Linux compatibility?
|
Testing MediathekView on dereel
|
Topic: technology, multimedia, opinion | Link here |
So finally I have dereel up to date and with linux64 compatibility. Test MediathekView:
=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 49 -> export JAVA_HOME=`pwd`/jre=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 51 -> $JAVA_HOME/bin/java -jar MediathekView.jar
/usr/ports/multimedia/mediathekview/work/MediathekView/jre/bin/java: error while loading shared libraries: libjli.so: cannot open shared object file: No such file or directory
Huh? That worked on tiwi. Yes, I've seen this message before, but it went away when I used the native Java.
But looking at what happened since then, that seems to almost be a coincidence. Now I find:
=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 53 -> unset JAVA_HOME=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 54 -> java -version
openjdk version "1.8.0_322"
OpenJDK Runtime Environment (build 1.8.0_322-b06)
OpenJDK 64-Bit Server VM (build 25.322-b06, mixed mode)=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 55 -> export JAVA_HOME=`pwd`/jre=== root@dereel (/dev/pts/7) /usr/ports/multimedia/mediathekview/work/MediathekView 56 -> $JAVA_HOME/java -version
/usr/ports/multimedia/mediathekview/work/MediathekView/jre/bin/java: error while loading shared libraries: libjli.so: cannot open shared object file: No such file or directory
What a mess this stuff is!
|
Fixing the sump pump
|
Topic: Stones Road house, general, opinion | Link here |
Call from Jarrod from UPI today today. He was planning to come by and look at the pump. Did he have a replacement pump? No. What use is it coming? Probably the thing now works.
To be sure, went out and checked. Yes, no longer trips the circuit breaker. But Jarrod said he could come anyway and check the electrics. OK, as long as he knows what he's letting himself in for.
He came along and tested the cable. Yes, low resistance between ground and active. OK, change the pump. He did have one with him (“I'm not as stupid as I look”). So why did he say that he didn't have one? Before he put it in, he drilled a couple of holes in the underside, to spray a little of the pumped water back out into the sump water to mix up the sludge. He seems to think that that will somehow reduce the likelihood of electrical problems, though he wasn't able to explain how.
While they were there, they checked the old pump, which I still haven't tidied away. Yes, it too has a short circuit. Why does that happen? Jarrod says that it isn't common; could it be something to do with the location? It would be interesting to know whether the problem is in the pump or the cable. Maybe cable damage round the frame of the cover? I should investigate the one I have left over.
| Thursday, 28 April 2022 | Dereel | Images for 28 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Fixing corrupt videos
|
Topic: multimedia, technology, opinion | Link here |
Lately I've been running into an issue with streams of Al Jazeera news downloaded from YouTube: I can't search back, getting messages like
AV: 00:00:57 / 00:18:05 (5%) A-V: -0.000
[ffmpeg/video] h264: Missing reference picture, default is 0
[ffmpeg/video] h264: decode_slice_header error
AV: 00:01:00 / 00:18:05 (5%) A-V: -0.000
[ffmpeg/video] h264: Missing reference picture, default is 0
[ffmpeg/video] h264: decode_slice_header error
AV: 00:00:49 / 00:18:05 (4%) A-V: -0.000
That's new, and clearly there's some format corruption there. Can I fix it by recoding? It took quite a while to find out, but this page proved useful. It's as simple as:
ffmpeg -i input.mp4 -c copy output.mp4
There are other incantations which look at the images in more detail, something that might address some of the issues that I have with programmes from the ZDF, but for Al Jazeera that's enough, and it's something that I can incorporate into my download script.
Of course, mpv isn't without fault. It's another of these stupid programs that wants to set a foreground text colour without knowing the background colour. Trying to display the same programme on eureka (across the network) gave me:
|
|
|||||||||
Sanitizing, this says
The legacy option syntax ('-vo value') is deprecated and dangerous.
Please use '--vo=value'.
(+) Video --vid=1 (h264 1920x1080 30.000fps)
(+) Audio --aid=1 (aac 2ch 44100Hz)
[vo/x11] Warning: this legacy VO has bad performance. Consider fixing your graphics drivers, or not forcing the x11 VO.
AO: [oss] 44100Hz stereo 2ch s32
VO: [x11] 1920x1080 yuv420p
[vo/x11] Shared memory not supported
[vo/x11] Reverting to normal Xlib
[vo/x11] Shared memory not supported
[vo/x11] Reverting to normal Xlib
[vo/x11] Shared memory not supported
[vo/x11] Reverting to normal Xlib
[vo/x11] Shared memory not supported
[vo/x11] Reverting to normal Xlib
AV: 00:00:37 / 00:18:05 (3%) A-V: -0.000 Dropped: 54
I hadn't even noticed the first two lines before, and now that I do, I have my serious doubts about them. What's dangerous about an option format that has been with us for decades? And how does changing it to something more “modern” improve anything?
|
More git strangeness
|
Topic: technology, opinion | Link here |
Bringing dereel up to date again today:
=== root@dereel (/dev/pts/0) /usr/src 72 -> igitt up
fatal: unsafe repository ('/eureka/home/src/FreeBSD/git/main' is owned by someone else)
To add an exception for this directory, call:
git config --global --add safe.directory /eureka/home/src/FreeBSD/git/main
I've seen that before. But that was with -STABLE, and this is -CURRENT, which I have only used on dereel. Why has git decided that the repository “is owned by someone else”? And who is this someone? And what does it mean by “owned”?
| Friday, 29 April 2022 | Dereel | Images for 29 April 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
AusPost: scammers?
|
Topic: general, technology, opinion | Link here |
A few weeks ago Yvonne received a package transported by Australia Post:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Despite the “Do not bend”, it had been bent and badly damaged. On 21 April Yvonne entered a formal complaint and demanded compensation. OK, receipts please. Yvonne sent them last week.
No reply. Lost in the post, maybe? On Wednesday she went in person to the appropriate place, and later received an email:
Date: Wed, 27 Apr 2022 02:08:13 +0000 (GMT)
From: Australia Post <customerservice@auspost.com.au>
Subject: Australia Post Investigation - Case: 48079138
Message-ID: <Ut0tQ000000000000000000000000000000000000000000000RAZ6YS00K0FjMGtWT8CaroQgfb5jlw@sfdc.net>
We recently contacted you regarding your enquiry, and need some additional information in order to proceed with your case.
Please reply to this email to provide more information and we will be happy to assist you.
[ ref:_00D301GGce._5002v34899P:ref ]
~~~
Previous Communication :
Help & Support
Reference 48079138
Update for Investigation - 48079138
Hi Yvonne,
RE: 99706618700601000830904
Item: painting
Thank you for lodging a case with us regarding a damaged 'painting' which you received damaged. I can only imagine how upsetting it is to wait for a delivery only to find the item damaged and I extend my apologies for this. I wish to discuss this matter further with you today.
I called you earlier on 0417397062 but I was unable to reach you.
Yvonne, as the item is over $100, we will require a proof of value such as ; an invoice, receipt or an eBay sale screen showing the item, the value and the sale details and postage costs.
+Once this is received, we will be able to call you to discuss this matter further.
Did you know Australia Post has a mobile App? Our free App is like having a Post Office in your pocket. It makes it even easier to access your favourite Australia Post tools, services and
+information. You can track your deliveries, find delivery time estimates, pay bills and much more. Our free App is available for IOS and Android, for more info visit,
+https://auspost.com.au/about-us/about-our-site/australia-post-app?fm=search-organic:australia%20post%20app:1.
We look forward to your response in due course.
Thanks for your patience and ongoing support at this difficult time.
That's almost identical to the message she received last week! And look at how much is wrong with it:
Clearly the person who sent the message was confused. I thought that some of Yvonne's reply might confuse him further. For example, instead of re-attaching the invoices, she bounced the original message. So I followed up with a clarification. DSN confirmed that both messages were received:
Apr 29 04:35:32 lax postfix/qmgr[73953]: 1FFB428034: from=<yl@lagoon.lemis.com>, size=3644, nrcpt=1 (queue active)
Apr 29 04:35:34 lax postfix/smtp[71573]: 1FFB428034: to=<customerservice@auspost.com.au>, relay=mxa-00312501.gslb.pphosted.com[185.132.181.201]:25, delay=2, delays=0.21/0/1.1/0.77, dsn=2.0.0, status=sent (250 2.0.0 3fprs5et07-1 Message accepted for delivery)
...
Apr 29 05:16:36 lax postfix/qmgr[73953]: 102A9280CE: from=<groggyhimself@lemis.com>, size=866042, nrcpt=1 (queue active)
Apr 29 05:16:40 lax postfix/smtp[71786]: 102A9280CE: to=<customerservice@auspost.com.au>, relay=mxb-00312501.gslb.pphosted.com[185.183.29.208]:25, delay=9.4, delays=5.8/0/0.78/2.9, dsn=2.0.0, status=sent (250 2.0.0 3fprsb6xam-1 Message accepted for delivery)
But then I got a strange message:
Date: Fri, 29 Apr 2022 05:17:11 +0000 (GMT)
From: Australia Post - No Reply <noreply@auspost.com.au>
To: "grog@lemis.com" <grog@lemis.com>
Subject: Australia Post - Auto Reply - Please visit Help & Support
Content-Type: text/plain; charset=ISO-8859-1
Australia Post
Sorry, we couldn\222t find a match!
This means that:
you\222ve sent us a new request, which cannot be supported via this email address
the original email you replied to has been modified or cropped
your enquiry has been closed
If you have any additional enquiries, please visit Help & Support to:
contact our support teams
find answers to many commonly asked questions
Thank you,
Australia Post
[ref:_00D301GGce._5002v34UUSR:ref]
This email was sent by Australia Post. Australia Post does not represent, warrant or guarantee that the integrity of this email communication has been maintained nor that the communication is free of errors, viruses or interference.
Help & support | Terms & conditions | Privacy policy
Huh? What's wrong with this message?
OK, go looking for “Help & Support”. The HTML version helps there, including proof that the character set is broken and confirming my guess at interpreting the message:
|
|
|||||||||
The URL I'm looking for is http://auspost.com.au/help-and-support?ilink=ccc-edm-cases-help#/. But it contains nothing useful. How about a human? No offer. Is this message even from Australia Post? All indications, including the terminal stupidity, suggest that it is. But how about reporting it as a scam? Maybe they'll react more efficiently then. And how about that, with only moderate searching, found this page:
|
|
|||||||||
OK, followed the link to the feedback form:
|
|
|||||||||
Huh? That has nothing to do with Australia Post. Try again:
|
|
|||||||||
What is this stuff? Maybe this is a scam. Where did I get this mail from?
From noreply@auspost.com.au Fri Apr 29 15:17:14 2022
Return-Path: <noreply@auspost.com.au>
Received: from lax.lemis.com (www.lemis.com [45.32.70.18])
by eureka.lemis.com (Postfix) with ESMTP id 1A82E2635BE
for <groggyhimself@lemis.com>; Fri, 29 Apr 2022 15:17:14 +1000 (AEST)
Received: from smtp15-hnd-sp1.mta.salesforce.com (smtp15-hnd-sp1.mta.salesforce.com [101.53.172.238])
by lax.lemis.com (Postfix) with ESMTP id 393782815B
for <groggyhimself@lemis.com>; Fri, 29 Apr 2022 05:17:13 +0000 (UTC)
Not from Australia Post! But when I bounced my message, I got the same response. How did salesforce.com intercept Australia Post's mail?
That's enough for today, but clearly it warrants following up.
|
Bomb Guterres!
|
Topic: politics, opinion | Link here |
So António Guterres was in Moscow earlier this week to talk to Volodomyr Putain about ending their aggression in Ukraine. Very little was reported, but it's clear that Putain did not agree, presumably blaming Ukraine for the aggression.
Today Guterres was in Kyiv. And barely had his meeting with Volodymyr Zelenskyy ended that a number (reportedly 2 to 5) of Russian missiles hit targets in Kyiv.
I'm dumbfounded. What was the purpose of that? Thumb their noses at Guterres? Are they trying to destroy the credibility of the United Nations? I don't understand what earthly good that can do for Russia, apart from possibly motivating other Russians to get rid of Putain.
In passing, it seems that the Russian use of the word “Nazi” is well-established. From episode 1-6 of Servant of the People, first aired in 2015:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The background is that president Goloborodko is back in the classroom and has asked one of his pupils to write an essay on the subject.
|
More pump fun
|
Topic: Stones Road house, general, opinion | Link here |
Call from Jarrod of UPI again today. He had found the cause of last weekend's pump problems, a damaged cable, and he wanted to return to put it back in again.
Interesting. I had wondered about that, specifically whether the damage had occurred round the frame. And indeed, so it was. But looking at the gap under the frame—bigger than I had expected—and the damage, it seems less likely:
|
|||||||||||||||||||||||
That looks like spade damage, which is quite plausible. But is it worth replacing the old pump? The “new” pump is no longer brand new, and the work is almost as expensive as the pump. Still, not a question, and he came along and replaced it, in the process spending a surprising amount of time pulling the cable through an outer protective tube:
|
|||||||||||||||||||||||
He also installed a non-return valve in the output pipe, which seemed rather unnecessary, but he says that it will reduce backflow, which makes sense. Confirmed that it worked, and he was off.
Later we had some heavy rain—something like 18 mm in an hour. Came outside to find the entrance flooded and the pump submerged:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Called up Jarrod, 16:45 on a Friday afternoon. Some hope. But he did answer, talked me through a number of things to try, none of which worked, but which confirmed that the pump was trying to run. So he came out and checked through things himself, without finding anything. But after removing the pump and all bits and pieces, it worked, and continued working after he reassembled it, finally pumping the sump dry:
|
|||||||||||||||||||||||
That was at 17:30. After he left, it started raining again. Outside at 18:00 to see another flood! Called up Jarrod, who could barely have got back by that time, left voice mail, and out to check. Oh, no, pump running, just so much water came down that it couldn't get into the sump with the cover on. Another call to call him off, but he didn't hear it before calling me again. He must be heartily sick of this place by now.
|
Apple in kitchen?
|
Topic: food and drink, technology, opinion | Link here |
I still have an Apple computer in the kitchen. Now I have removed the upper part of the cupboard where I had it before, so I can access the keyboard more easily:
|
|||||||||||||||||||||||
And how is it? The more I try these things, the more I get the impression that there's only one good way to access a computer, and that's sitting in front of a good keyboard. In days gone by I used laptops, but they're barely acceptable, and you really need to be sitting to have good control of a keyboard.
No wonder I can't stand mobile phone data entry.
| Saturday, 30 April 2022 | Dereel | Images for 30 April 2022 |
| Top of page | ||
| previous day |
|
Understanding AusPost exploit
|
Topic: general, technology, opinion | Link here |
So what went wrong with the Australia Post interaction yesterday? How did salesforce.com get wind of my messages to Australia Post? Looking at the original message to Yvonne, I found:
