|
| Greg's diary |
| May 2021 |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Saturday, 1 May 2021 | Dereel | Images for 1 May 2021 |
| Top of page | ||
| next day | ||
| last day |
|
Battery calibration: so nice, so nice, we do it twice
|
Topic: Stones Road house, general, opinion | Link here |
I was pleasantly surprised by the brevity of yesterday's battery recalibration. But it wasn't over. Today it happened again. Once again it only charged to 100%, and this time (later in the morning) it took only 84 minutes.
Where will we go from here? We'll see. The great advantage of not discharging the batteries is, of course, that no power from the PV array gets lost. It's still not completely free of charge, of course, because the system charges from the grid when it wouldn't otherwise be necessary.
|
Gold?
|
Topic: Stones Road house, general, animals, opinion | Link here |
Rick, the gold digger who came past a couple of weeks ago, showed up this morning just as we were preparing breakfast. He wanted to “dig” for gold (which I think means “sniff rocks with a metal detector”), but the time was inconvenient. I asked him to come back later, but he didn't. I wonder if Pedro worried him: he came up with an effusive ”hello”, but it seems that Rick isn't a dog person.
|
Frijoles de la olla again
|
Topic: food and drink, opinion | Link here |
Somehow the last few days have been busy in the kitchen, and they show no signs of letting up: mi udang, chicken pieces for breakfast, frijoles de la olla, bread...
The frijoles are fairly simple: cook the beans with onion, garlic, chili, epazote and lard, adding salt a little later. I have plenty of frozen epazote from last year, but also plenty of fresh epazote from this year. It's going to seed, so it's a good time to harvest a few branches.
But how many? For 1 kg of beans I had written 72 g of epazote. But in the meantime I had had difficulties with getting the leaves out. How about putting whole twigs in, so that I can find them more easily after cooking? They look like this:
|
|||||||||||||||||||||||
How much do you put in? Those twigs appeared to be about 150 g, and I decided to put in 120 g of them. After cooking, they were easy enough to find:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But a number of leaves fell off, of course. I'll have to decide whether it's worth putting the twigs in—after I have eaten it, which will take several months.
The other issue was the chili. I had written “25 g chili", but today I had different chilis. Tasted one: very strong. So I used less than 10 g.
|
French chicken rice
|
Topic: food and drink | Link here |
Hainan chicken rice is a world-famous dish, but my attempts to recreate it have been less than spectacular. Yvonne has a different dish: poule au riz, literally “chicken with rice”. It has a number of similarities with Hainan chicken rice, including the steeping of a whole chicken:
|
|||||||||||||||||||||||
It's less spectacular in appearance when served:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
| Sunday, 2 May 2021 | Dereel | Images for 2 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Bread experiments
|
Topic: food and drink, opinion | Link here |
I've been baking bread for 12 years now, and for much of that time the proportions have been the same, modulo some experimentation with the amount of water. In particular, the proportions are 9 parts rye, 4 parts wheat, for reasons lost to history. Recently Yvonne asked why not all rye? Answer “nobody does that”.
But that's not a good answer, and in particular a little more rye might be interesting. How about 11 parts rye, 2 parts wheat? That will require less water, of course. Normally I use 975 ml of water total, but this time I reduced it to 935 ml. And then I discovered that I didn't have enough rye flour: I only had a total of 949 g, making a ratio of 949:221 instead of the 970:160 I wanted. Yes, I had another sack that I could have opened, but this was an experiment, so it seemed reasonable just to document it.
The dough seemed a little on the dry side, but apart from that things seemed OK. Rose faster (under 4 hours instead of the more typical 5), but that could be a coincidence. And it was noticeably darker. The rest will come when we eat it.
|
Reinheitsgebot reinterpreted
|
Topic: food and drink, opinion | Link here |
I've ranted in the past about the German Reinheitsgebot, the so-called purity decree for beer. It was issued in Bavaria, a German state, on 23 April 1516, potentially to explain the doubling of the price of beer on that day (from one Pfennig to 2 Pfennigs per measure, approximately a litre). That was to stay that way until Michaelmas, when the price would revert to 1 Pfennig.
There was a second clause: brew beer with only barley (and, implicitly, no wheat) and hops (and, implicitly, no other herbs). No mention of yeast, because it wasn't understood at the time. But as I complained 11 years ago, this doesn't stop people from brewing beer with wheat and claiming that it conforms to the Reinheitsgebot. And that's the point: the Reinheitsgebot is obsolete, having been replaced by laws from the European Union.
But Dan Murphy has copywriters that know it even better:
For many beer lovers, Germans do it best. Thanks, in no small part, to Reinheitsgebot, the German Purity Law which allows beer to be brewed with malt, hops, and water, with yeast being added only after fermentation. Dating back to 1516, this rule is the oldest, still enforced food regulation law in the world. Some people might see it as restrictive, but it’s behind the extraordinarily high quality beer produced by Germany.
Adding yeast after the fermentation? Where did he get that idea? Presumably it's an explanation for Hefeweißbier (Yeast White Beer, where there is considerable yeast in the beer as served).
|
ALDI: Soup for geniuses
|
Topic: food and drink, opinion | Link here |
Yvonne bought some pre-packaged pumpkin soup on Wednesday, and we were to consume it today:
|
|||||||||||||||||||||||
But how do we open it?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I couldn't work it out. Into the office to take a photo, where it gradually became clear:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It had looked like the “Open” tab had to be pulled up, but in fact it was just protecting the lid, which was not joined to the rim as one might have expected, and the intention was to pull it out, as shown. Yvonne had been all for returning it to ALDI and saying that we couldn't open it.
| Monday, 3 May 2021 | Dereel | Images for 3 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More gardening
|
Topic: gardening | Link here |
Nathan Hiscock, the gardener, along today to mow the lawn. I hadn't realized that Bryan Ross wasn't going to be here. Each have their own specialties, and I really needed Bryan to tidy up the garden beds.
|
Another dead kangaroo
|
Topic: animals | Link here |
A couple of days ago Yvonne told me about a dead kangaroo in the sheep paddock across the road. Like so often, it had got a leg caught in the wires of a fence, and died upside down.
But now there's less:
|
|||||||||||||||||||||||
I don't know what it was, but something removed nearly all of the kangaroo and ate a lot of it.
|
|||||||||||||||||||||||
It seems that the kangaroo was a big male Eastern grey, so whatever took it apart must have been pretty big too.
|
Backup pain
|
Topic: technology, opinion | Link here |
The beginning of the month is the day when I do my level 0 dumps. There's a lot of data to transfer, and they run for up to 15 hours. But yesterday things didn't end quite the way I had hoped:
dump -0uf - /home | bzip2 > /dump/lagoon-FreeBSD/0/home.bz2
DUMP: WARNING: should use -L when dumping live read-write filesystems!
DUMP: Date of this level 0 dump: Sat May 1 22:05:27 2021
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/ada0p5 (/home) to standard output
...
DUMP: dumping (Pass IV) [regular files]
DUMP: 0.59% done, finished in 14:04 at Sun May 2 12:15:09 2021
DUMP: 1.25% done, finished in 13:10 at Sun May 2 11:26:44 2021
...
DUMP: 55.19% done, finished in 5:36 at Sun May 2 10:37:44 2021
bzip2: I/O or other error, bailing out. Possible reason follows.
bzip2: Permission denied
Input file = (stdin), output file = (stdout)
DUMP: Broken pipe
DUMP: The ENTIRE dump is aborted.
Dumping completed at Sun May 2 05:04:14 AEST 2021
What went wrong there? An NFS problem? I should follow up, but first I should upgrade my systems. Simpler just to repeat the dump.
12 hours? Much of that would be caused by bzip2. So I ran just a simple
=== root@lagoon (/dev/pts/2) /eureka/home/grog 6 -> dump -0uf - /home > /dump/lagoon-FreeBSD/0/home
DUMP: WARNING: should use -L when dumping live read-write filesystems!
DUMP: Date of this level 0 dump: Sun May 2 11:50:30 2021
...
DUMP: dumping (Pass IV) [regular files]
DUMP: 6.72% done, finished in 1:09 at Sun May 2 13:05:17 2021
...
DUMP: 94.09% done, finished in 0:04 at Sun May 2 13:05:18 2021
DUMP: DUMP: 267076646 tape blocks
DUMP: finished in 4492 seconds, throughput 59456 KBytes/sec
DUMP: level 0 dump on Sun May 2 11:50:30 2021
DUMP: DUMP IS DONE
Now isn't that much faster? OK, so then I started the bzip:
=== root@eureka (/dev/pts/4) /dump/lagoon-FreeBSD/0 14 -> time bzip home
real 1588m47.974s
user 1576m6.408s
sys 4m58.371s
1588 minutes! That's amazing! That's well over 26 hours, longer than it would have taken normally. It wasn't until later that I discovered I had run (arguably) the wrong program: instead of the new, improved bzip2, I had run the old, mouldy bzip.
OK, let's try again with bzip2. I'll find out tomorrow, In the meantime, looking at the results, I wonder if it's even worth the trouble compressing:
=== root@eureka (/dev/pts/22) /dump/lagoon-FreeBSD/0 9 -> l
-rw-r--r-- 2 root wheel 273,486,479,360 2 May 13:06 home
-rw-r--r-- 1 root wheel 223,600,494,258 2 May 13:06 home.bz
|
Arena cover: the pain
|
Topic: Stones Road house, general, opinion | Link here |
The council still wants us to use Colorbond® for our riding arena cover. And we don't want to. David Rowe doesn't seem very interested in helping resolve the problem, so started writing a hopefully convincing letter to the council. It's like pulling teeth.
And then I wanted to refer to the plan in the application. I don't have it! I saw it once in the advertisement, but that's no longer available. All I have is the application, and that doesn't give access to the plans (any more).
It's beginning to look more and more like David is a big part of the problem. Now to write him a friendly but firm email asking for details.
| Tuesday, 4 May 2021 | Dereel | Images for 4 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Compressing dumps, try 2
|
Topic: technology, opinion | Link here |
Yesterday's attempts at compressing backups were suboptimal because I used the old, worn-out bzip instead of the new, better bzip2. So I tried again:
=== root@eureka (/dev/pts/4) /dump/lagoon-FreeBSD/0 17 -> time bzip2 home
real 815m16.263s
user 804m19.745s
sys 3m54.405s=== root@eureka (/dev/pts/4) /dump/lagoon-FreeBSD/0 18 -> l
...
-rw-r--r-- 1 root wheel 273,486,479,360 2 May 13:06 home
-rw-r--r-- 1 root wheel 222,389,811,395 2 May 13:06 home.bz2
-rw-r--r-- 1 root wheel 223,600,494,258 2 May 13:06 home.bz
How about that. Still no ball of fire, but it's much faster, 13.6 hours instead 26.5 hours, almost exactly half the time. Interestingly, it also compresses marginally better. Of course, there's a third option: pbzip2, which by default uses all available cores.
|
24 years in Australia
|
Topic: history, general, opinion | Link here |
Read in my calendar mail today:
May 4 Greg returns to Australia, 1997
That's almost a third of my life ago. How time flies!
|
Pedro vomits blood?
|
Topic: animals, opinion | Link here |
Pedro vomited today, choosing the nearest carpet and not taking no for an answer. Finally he was done:
|
|||||||||||||||||||||||
That's normal enough, but what's this?
|
|||||||||||||||||||||||
It happened a minute or two later, after I had chased him round the carpets. Is that blood? Is it serious?
|
Descaling: So nice, so nice, we do it twice
|
Topic: food and drink, opinion | Link here |
Last week I had trouble descaling my coffee machine. Despite following the instructions (almost) to the letter, the red “descale” LED didn't go out.
What could be wrong? Yes, I didn't use descaling agent, because I don't need it: the whole idea of descaling a device that uses rain water is stupid. But surely that couldn't be the reason?
Another possibility could be that I had got the wrong (machine readable) instructions. Today I found the printed version that I got with the machine. And oh wonder, it was different! After descaling, rinse the machine twice! And put the container under both the cappuccino pipe and the coffee outlet. Did that, only about an hour, and it worked!
There's also a water hardness test strip attached to the instructions. I should plan to take a video of the test.
|
Eternal planning permit issues
|
Topic: Stones Road house, general, opinion | Link here |
As planned yesterday, sent a mail message to David Rowe asking for documentation. The response was almost immediate, but not what I expected:
Date: Tue, 4 May 2021 06:57:39 +0000
From: David Rowe <david@bondhomes.com.au>
Subject: RE: Shed in Dereel, Planning Permit P20390
Message-ID: <ME3P282MB14286AC9BEB9719B4D3661448A5A9@ME3P282MB1428.AUSP282.PROD.OUTLOOK.COM>
Hi Greg,
You have sent me an email with no content, where are you up to regarding the colour for your building
...
-----Original Message-----
From: Greg 'groggy' Lehey <groggyhimself@lemis.com>
Sent: Tuesday, 4 May 2021 3:47 PM
To: David Rowe <david@bondhomes.com.au>
Cc: Yvonne Lehey <yvonne@lemis.com>
Subject: Re: Shed in Dereel, Planning Permit P20390
On Tuesday, 13 April 2021 at 8:38:50 +0000, David Rowe wrote:
> In amending plans I would need to specify the colorbond colour. If you
> want me to contact council and say that you will want to go to VCAT I
> can ask them what the process is. Other than that if you want to have
> a site meeting, this would be between yourself and council , I rather
> not get involved.
OK, I understand. Under the circumstances, it's best if I continue the discussion with council. For this, I need from you:
- A copy of the application. I asked you for this before (30
November), but I didn't receive any reply. My neighbours showed me
details of the application from the advertisement of the
application, but it's no longer available.
- Copies of any written correspondence between you and council. In
particular, I'd be interested to see how you presented the
information I sent you on 29 March, and if possible how they
responded.
Somehow I'm getting the impression that he doesn't want to cooperate. What do we do now?
| Wednesday, 5 May 2021 | Dereel | Images for 5 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Lens sharpness again
|
Topic: photography, opinion | Link here |
Somehow I'm getting less than satisfactory results with my new Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3 super zoom lens. I've been trying one particular kind of shot, with Yvonne and the dogs in the forest. And they keep coming out unsharp!
|
|||||||||||||||||||||||
Yes, the “thumbnail” images are quite acceptable, but what's this?
|
|||||||||||||||||||||||
Why is that? Is there something wrong with the lens? I know it's not the sharpest one in the stable—in fact, it's the least sharp—but I had exactly that issue with my first Olympus M.Zuiko Digital ED 14-150 mm f/4.0-5.6 seven years ago. Could it be the same issue? If it is, it's not as bad as last time, but is it good enough?
Off today to compare lenses. The focal lengths that interest me are clearly beyond 100 mm, which my standard M.Zuiko Digital ED 12-100 mm f/4.0 IS PRO can do better. In the end I compared 6 different lenses:
From left to right, the 12-100, the 12-200, the 14-150 (the replacement for the defective one), the Zuiko Digital ED 35-100 mm f/2.0, the Leica DG Vario-Elmar 100-400 mm f/4.0-6.3 and the old Hanimex 300 mm f/5.5 telephoto that I have had for nearly 20 years:
|
|||||||||||||||||||||||
Why the Hanimex? It's M42 and manual, and I don't use it any more. But by chance Kev Russell of the Facebook M43 Tech Talk has been comparing 300 mm lenses, so it seemed worth a go.
The results? I'll need some time to analyse them, but it seems that the 12-200 really needs its optical corrections; without them CA is quite pronounced. In fact, the Hanimex does better. But it seems that there's nothing wrong with this lens; presumably the problems I had were due to the slow shutter speed (1/20 s), whether it's camera shake or Pedro's obvious motion.
|
The daily dump compression
|
Topic: technology | Link here |
So today my pbzip2 finished:
=== root@eureka (/dev/pts/4) /dump/lagoon-FreeBSD/0 30 -> time pbzip2 home3
real 309m3.275s
user 1771m35.488s
sys 28m39.588s
As expected, at only 5 hours it was the fastest. But look at that user time: not quite 30 hours! That's far longer than bzip (1576 minutes/26.3 hours) or bzip2 (804 minutes/13.4 hours). I hadn't expected that.
The compression was good, but not as good as bzip2:
| Compressor | User time | User time | Size | |||
| minutes | hours | |||||
| bzip | 1576 | 26.3 | 223,600,494,258 | |||
| bzip2 | 804 | 13.4 | 222,389,811,395 | |||
| pbzip2 | 1772 | 29.4 | 222,482,452,587 |
| Thursday, 6 May 2021 | Dereel → Ballarat → Dereel | Images for 6 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Navigating without GPS
|
Topic: health, technology, opinion | Link here |
Into town this morning for a blood test. For once I didn't take my GPS navigator: after all, I know my way round Ballarat. Turn right on Leith St—don't I? Yes, of course I do. Down Barkly St and turn left towards Bakery Hill, and then I'm there.
But where do I turn left? Main Road? Or at the roundabout before. Chose before. Then another 3-way roundabout. Where now? Chose left again and didn't recognize anything until I was at the police station: Eastwood Street, going in exactly the wrong direction.
I was able to find my way back, of course, but the real issue is that I relied too much on the navigator. Is that good or bad?
|
Compression, for real
|
Topic: technology, opinion | Link here |
Daniel Nebdal read my articles about compressing dumps. It seems that I'm a rank amateur: he has tried many others with better results than the ones I had:
xz has been in base for mumble years, and often gives a better compression/time tradeoff than the bzip family. The default settings are single threaded and slow, but try xz -T4 -1 (four threads, +level 1) - at least for me, it's both faster and smaller than bzip. The speed depends on how many cores you can give it, of course.
I think lz4 has been included for a few years, too. It's much faster than gzip but typically doesn't compress as well, which is still a useful niche. It's popular for disk compression in ZFS, since the time you save reading less data is often more than you spend decompression. I've also used it for network transfers.
The new toy in 13.0 is zstd, from the same guy who wrote lz4. Its party trick is that it scales very well - it has good compression per CPU cycle all the way from "faster than gzip -1" to "smaller output than bzip2 -9", sort of an everything for everyone. It's in archivers/zstd if you want to test without upgrading to 13.
As a test, here's my /usr/src sent through some different alternatives. I think zstd -1 and lz4 ran into the disk read bandwidth. This was about 2580 MB of rather compressible input (though the .git folder is already compressed). YMMV, of course; this is an intel i5 3570 (a 2012 3.4GHz 4-core) running 13.0.
method MB out time sec input MB/s xz 1400 923 2.8 xz -T4 1410 247 10.4 zstd -12 1450 130 19.8 xz -T4 -1 1460 134 19.3 bzip2 1470 293 8.8 Zstd -5 1480 37 69.7 gzip 1510 68 37.9 Zstd -1 1540 17 151.8 gzip -1 1570 49 52.7 lz4 1670 17 151.8
So why didn't I know this? I think it's a case of “if it ain't broke, don't fix it”. I've probably been doing backups my way for over a quarter of a century, and those were the tools available at the time.
| Friday, 7 May 2021 | Dereel → Ballarat → Dereel | Images for 7 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
COVID-19 vaccination with UFS
|
Topic: health, opinion | Link here |
Into Ballarat today to get the first AstraZeneca vaccine shot against SARS-CoV-2. We had been given consent forms (PDF) to fill out, and of course I put mine off until the last minute.
What a form! It asks for my name 7 times, my Medicare number (please fill out the boxes) 6 times. It asks questions that no person can reasonably be expected to answer, like “Do you have a mast cell disorder?”, and others that are so vague that they can only be answered with “yes”: “Have you been sick with a cough, sore throat, fever or are feeling sick in another way?”. Well, yes, I have been sick with a cough, most recently on 31 July 2019. So clearly I had to answer yes, and prepared myself for the response.
We had already heard about long waiting times for the vaccination, so we were prepared for the worst. Arrived at 11:47:
|
||||||||||||||||||||||||
Inside there were two patients and two bureaucrats, one of which typed in our forms. No comment about my incomplete answers or Medicare numbers. The other took bookings for second rounds—in 12 weeks! My understanding was that it should be 3 weeks. I was about to cancel the booking until they could complete it in the normal time. But no, it seems that they have a general policy of giving the second dose in 12 weeks:
|
||||||||||||||||||||||||
I later found an article in the British Medical Journal which confirmed the time. I wonder when they changed that.
Once the bureaucracy was over, out to wait:
|
||||||||||||||||||||||||
That wasn't long either. I didn't even have time to open my book. Soon Maxine, apparently the only nurse, called us in and asked yet more questions. The first was Yvonne's first initial: Y. Maxine got quite flustered. She didn't have a good answer, and it turned out that she had understood “why?”. Then all sorts of questions that could equally well have been on the first form. And she, too, didn't ask about the coughing or the mast cell disorders. Why do we have to fill out the forms?
Then the “jab”. Almost instantaneous, so fast that I didn't have time to get a photo. It wouldn't have helped: for some reasons they don't want photos taken. Had mine done, so high on the arm that I had to take off my shirt, and we were done. None of this horrible dart throwing that we've seen in the news: as gentle an injection as I have had. Maxine told us that the people on TV don't just throw darts, they also insert the needle in the wrong place.
Well, no. We had another 15 minutes to wait to see if we had any reactions (we didn't). And that was probably the longest period of the lot. On return home, I checked the newspapers: the delays related to another mass vaccination hub. You certainly can't call this one mass vaccination; my guess is that they vaccinate about 5 people per hour.
And the reactions? People have made a lot of noise about rare side effects of the vaccine, but what's normal? About 2 hours later I felt a slight stiffness in my arm, but it went away almost immediately. And after another 3 hours it happened again. I wonder if I would have paid any attention had I not just had the vaccination.
|
Bloody Guinness!
|
Topic: food and drink, opinion | Link here |
Putting a couple of slabs of Oettinger into the cupboard, still more cans of Guinness took offence and split:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
This is really unacceptable. Should I go to the trouble of ringing them up and complaining?
| Saturday, 8 May 2021 | Dereel | Images for 8 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Hypochondria?
|
Topic: health, opinion | Link here |
Yvonne wasn't feeling the best this morning, a vague sense of nausea. Consequence of the COVID-19 vaccination? It cleared up during the day.
I wasn't feeling as bright as normal, but it wasn't serious, and I wondered if it wasn't due to the particularly dreary weather. My arm felt vaguely stiff, but less so than after the influenza vaccination last year.
So: real reactions, or imagined? This is the first time we've thought about them. Over 50 years ago I had a cholera vaccination. That was painful, and I have a clear recollection of visiting Jock Ferguson, a retired (and obviously Scottish) engineer. When I told him about the pain in my arm, he clapped me on the arm. Ow! You bastard! But he said that the pain would go away quickly, and it did.
That was much more painful than the current COVID-19 vaccination, but I didn't even mention the incident in my diary of the time. It must have been in the first week of July 1970. But it relativizes my careful observations now.
|
Panorama convergence revisited
|
Topic: photography, opinion | Link here |
What's the angle of view of a fisheye lens? The canonical answer is 180°, diagonally for a “full frame” fisheye and vertically for a “circular” fisheye.
And the practice? 3½ years ago I did some measurements and came to the conclusion that the lens (the Olympus Zuiko Digital ED 8 mm f/3.5) didn't even do 180° diagonal. I'm not convinced that my measurements were accurate, but how does it compare with the Olympus M.Zuiko Digital ED 8mm f/1.8 Fisheye PRO that I received only 2 weeks later? Today was house photo day, so I decided to do one of the panoramas twice, once with the old lens, once with the new.
What did I expect? Maybe different coverage, potentially more flare from the old lens, less good matching of control points with the old lens, since DxO PhotoLab doesn't have a correction profile for it.
As expected, the final images are almost indistinguishable:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
You need to look at the Exif data to see that the second image was the one taken by the old lens. But to my surprise, the old lens had better results from the control point detector: average error 3.3 pixels, maximum 7.7. The new lens, despite optical corrections, had an average of 5.3 and a maximum of 12.8. Why is that? It could be the wind, but while checking my entrance pupil adjustments for the old lens, I noted a calculated scale position of 82 mm, along with a comment of using 97.5 mm. But that was based on a calculated position of 94 mm, so the values could be out of date. And this panorama is really not critical from that point of view.
And the image size?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
(run the cursor over an image to compare it with its neighbour) Again the first image was taken with the new lens.
It's interesting that the roll is different between the two images. Hugin shows that the roll for the new lens was a constant 0.2°, while for the old lens it was between 0.4° and 0.6°. More significantly, the roll was a constant 8.8° for the new lens, and between 9.5° and 10.2° for the old lens. I can't make up my mind what that signifies. Is it the lens? Hugin? The tripod?
In any case, it's relatively clear that the new lens has a wider field of view than the old one: the left side of the image is pretty much the same for both, but the right side is more considerably wider for the new lens.
| Sunday, 9 May 2021 | Dereel | Images for 9 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
What's this breakin attempt?
|
Topic: technology, opinion | Link here |
Overnight my external web server received round 2000 breakin attempts. The server tells me:
Date: Sat, 8 May 2021 18:14:24 GMT
From: World Wide Web Owner <www@lemis.com>
To: groggyhimself@lemis.com
Subject: FAILURE: /grog/diary-nov2017.php/%29%3BCREATE%20OR%20REPLACE%20FUNCTION%20SLEEP%28int%29%20RETURNS%20int%20AS%20%27%2Flib%2Flibc.so.6%27%2C%27sleep%27%20language%20%27C%27%20STRICT%3B%20SELECT%20sleep%2832%29-- <-
http://www.lemis.com:80/grog/diary-nov2017.php/
Referrer: http://www.lemis.com:80/grog/diary-nov2017.php/
Referenced URL: http://www.lemis.com/grog/diary-nov2017.php/%29%3BCREATE%20OR%20REPLACE%20FUNCTION%20SLEEP%28int%29%20RETURNS%20int%20AS%20%27%2Flib%2Flibc.so.6%27%2C%27sleep%27%20language%20%27C%27%20STRICT%3B%20SELECT%20sleep%2832%29--
Request URI: /grog/diary-nov2017.php/%29%3BCREATE%20OR%20REPLACE%20FUNCTION%20SLEEP%28int%29%20RETURNS%20int%20AS%20%27%2Flib%2Flibc.so.6%27%2C%27sleep%27%20language%20%27C%27%20STRICT%3B%20SELECT%20sleep%2832%29--
Remote IP: 45.146.166.229
Client: Mozilla/5.0 (Windows NT 6.2) AppleWebKit/536.3 (KHTML, like Gecko) Chrome/19.0.1061.1 Safari/536.3
Clearly this is a URL crafted to attempt some kind of request, but what? Resolving the %s, I get:
);CREATE OR REPLACE FUNCTION SLEEP(int) RETURNS int AS '/lib/libc.so.6','sleep' language 'C' STRICT; SELECT sleep(32)
So what is it trying to do? By itself this snippet seems harmless, especially since /lib/libc.so.6 doesn't exist. Is it relying on all 2000 requests being honoured?
|
Much cooking
|
Topic: food and drink, opinion | Link here |
Spent a lot of time cooking (and taking photos of it) today. Started with a relatively satisfying bacon and eggs dish:
|
|||||||||||||||||||||||
It's been nearly two weeks since Lorraine Carranza brought us a couple of bags of chestnuts. What can we do with them? Yvonne doesn't like chestnuts, and I can't think of what to do with them. But it's clear that if we don't cook them soon, they will go bad. So first step: roast them.
How? The canonical way is in the oven, 20 minutes at 200°. But what about the air fryer? 20 minutes at 180°? Started off like that, and round the 10 minute mark it began to look as if they were done. Off to get my camera, and while I was doing that, heard a bang from the kitchen. One chestnut exploded:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Apart from that, though, they appear to have been cooked. I have found (and then lost) a recipe for minced chicken and chestnut balls that looks like it could be good and even entice Yvonne to have a try. We'll see then how well it worked.
In the evening, then, chicken tanduri, done Yvonne's way. But I still need to cook them. A year ago I tried it in the air fryer, and they didn't come out badly. But neither did the ones I cooked in the microwave oven first.
How about trying that hybrid cooking method? Today I tried 7 minutes at 600 W in the microwave oven, then some time in the air fryer at 230° until they were nicely browned:
|
|||||||||||||||||||||||
The “some time” proved to be 15 minutes, longer than I had expected, and as long as I had taken last time without the microwave pre-cooking. It was cooked properly, though we must cut it thinner in future, but how about 10 minutes in the microwave oven, then maybe only 10 minutes in the air fryer?
| Monday, 10 May 2021 | Dereel | Images for 10 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Ring fried egg again
|
Topic: food and drink, opinion | Link here |
Nasi lemak for breakfast this morning, along with a fried egg.
The egg had to go on top of the rice. How do you keep it compact? We have rings, which we have tried and discarded. Try again?
The real issue is that it takes so long! Today I fried for nearly 9 minutes, by which time the white still had not completely congealed:
|
|||||||||||||||||||||||
And independently of that problem, the egg stuck to the non-stick ring:
|
|||||||||||||||||||||||
The result is still not really acceptable:
|
|||||||||||||||||||||||
Apart from the appearance, the bottom of the yolk was cooked through. How can I improve on it? One obvious solution would be a wider ring. Or a frying pan dished in the centre. Out looking for the latter, with no resounding success.
|
More riding arena pain
|
Topic: Stones Road house, general, opinion | Link here |
We've been waiting for months for the planning permit for the riding arena, and it's becoming clearer that David Rowe, whom we entrusted with the application, is a big part of the problem. After another mail exchange, discovered that he had done things that he hadn't told me about and not done things that he had promised to do. It all sounds something like an Anglican confession.
Clearly I have to take things into my own hands. Spent some time researching what had happened and writing a long and hopefully not too acrimonious letter about the situation. But that's painful, so decided to call Sarah Smith of the Golden Plains Shire Council. Once again I didn't get through—this time, also not for the first time, I didn't even manage reception. Left a message, but what is going on there?
|
Spathiphyllum photos
|
Topic: photography, gardening, opinion | Link here |
The Spathiphyllum flower that I observed last month is now in full flower:
|
|||||||||||||||||||||||
Time for some photos, of course. But how? Last time I used my “Viltrox JY-670 Macro Ring Lite”, and got surprisingly good results. But what about the other two units, both with TTL flash measurement: the Olympus STF-8 and the mecablitz 15 MS-1? Clearly a case for comparisons.
First, attaching the flashes. The “Viltrox JY-670 Macro Ring Lite” is a bit clumsy on top of the camera, and the cable insulation has torn. I kept the ring on the (Four Thirds) Zuiko Digital ED 14-42 mm f/3.5-5.6 that came with my second Olympus E-30, so all I had to do was attach the lens to the camera and then the ring flash to the lens:
|
|||||||||||||||||||||||
The mecablitz 15 MS-1 has its own ring, of course, but the trick here is that I need a second unit to trigger it:
|
|||||||||||||||||||||||
And the Olympus STF-8 is a real pain to mount. I leave the flash heads on the ring and then screw the lens into the ring, and only then do I attach the assembly to the camera:
|
|||||||||||||||||||||||
The Viltrox didn't do as well this time. As before, I set it to ½ power. I used the flash exposure meter, which told me f/22, considerably less than the f/11 I used last time. OK, compromise on f/16. No, the meter was right. Why? What has changed? But after stopping down to f/20, all seemed OK:
|
|||||||||||||||||||||||
The mecablitz produced an underexposed image:
|
|||||||||||||||||||||||
That was also at f/20, and in expectation I had already set the flash exposure compensation to +⅓ EV. Too weak? I changed the exposure to f/10, with no difference. Only when I came closer did it look marginally better.
And the STF-8? Same again. On reflection, it's clear that this is a poor test for TTL exposure: it aims for averages, not bright white, so that's what it delivers. From my point of view, it still makes me wonder why I use TTL rather than the Viltrox.
Independently of that, which flash gives the most even illumination. Here the three after postprocessing:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I don't understand why the TTL units have such dark backgrounds—not necessarily a disadvantage. But particularly the shadow from the STF-8 (3rd image) is unpleasant, somehow like the whole unit.
|
New COVID-19 site
|
Topic: health, opinion | Link here |
I've been following Covidly for stats about the pandemic, but it seems to be running out of steam. For example, there's no information about the state of vaccinations. Recently Al Jazeera aired a clip correlating vaccination rates with infection rates, in particular talking about the current situation in India (400,000 infections per day) with the vaccination rate (round 12%).
Clearly that's comparing apples and oranges: percentages and absolute numbers, which particularly in India are inaccurate. But Covidly can't help me there, because it doesn't track vaccinations. While looking for something else, came across this page, apparently not the only thing that the site offers. With a bit of messing around, some of the information seems quite usable. And of course it shows that Australia has an even lower vaccination rate than India.
|
More grid power problems
|
Topic: Stones Road house, general | Link here |
More grid power fluctuations this afternoon, a total of 4, depending on how you count them. The longest was 7 seconds, and before the inverter could time out on it (one minute), another 19 second outage occurred, to be followed some time later by a couple of brief 1-second outages.
| Tuesday, 11 May 2021 | Dereel | Images for 11 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Why 25 Mb/s is enough
|
Topic: technology, opinion | Link here |
For once in my life I'm on the trailing edge of Internet link speed. Since signing up with the National Broadband Network over 7 years ago, my link speed has been 25 Mb/s down, 5 Mb/s up. But as I discovered last year, that tariff is no longer available. Now the only speed they offer is 75 Mb/s down and (I think; they don't like to divulge such important information) 10 Mb/s up. When I wanted up upgrade my traffic quota in October, the only option with more traffic was unlimited 75/10 for $14 more. I managed to negotiate (and forget to note in my diary) the old 25/5 for $10 less because I was a “legacy” customer.
But why not pay the extra $10 for potentially 3 times the download rate? I don't need it! I do download a lot of videos, notably from Austria, Germany and Switzerland, using MediathekView, which limits the number of concurrent downloads. And downloading from Germany is slow, typically round 3 Mb/s.
To be expected from the other end of the world? Well, no. The downloads from Switzerland are round 10 times as fast. Here a bandwidth window from MediathekView:
|
|
|||||||||
The first half shows one download from Switzerland and one from Germany. The second half shows only the download from Germany. The sawtooth pattern is typical, but I don't understand why.
Under these circumstances, it's clear that a 25 Mb/s link is enough. iptop shows that the link occasionally maxes out at 25 Mb/s, but it's so seldom that it's not worth mentioning. So at least until I find a way to request more data, the 25/5 link should be enough.
|
Chicken and chestnut balls found
|
Topic: food and drink | Link here |
Found the chicken and chestnut balls recipe that I had lost. Now to make it some time soon.
|
The origins of SARS-CoV-2
|
Topic: health, opinion | Link here |
SARS-CoV-2 was first detected in Wuhan, and Donald Trump made much capital of the fact that it was a “Chinese disease”. But there have always been doubts, and recently I've found this article, which shows the results of investigations of a large number of samples, and comes up with some interesting insights, including this family tree:
There are at least three significant things that are visible from this tree:
I had a severe and for me completely atypical lung infection in July 2019. Could that have been related? From the point of view of time frame, it's quite possible. It might also explain why I had as good as no reaction to the vaccination last week. I wonder if they will ever trace the virus further back. But certainly this is something worth following, as is the Medical News site.
|
More flash considerations
|
Topic: photography, opinion | Link here |
Apart from the close-up photos of the Spathiphyllum flower that I took yesterday, I also took a general photo:
|
|||||||||||||||||||||||
What's the gradation like? This one was taken with natural light and +0.7 EV exposure compensation. As usual, it was processed with DxO PhotoLab using my particular parameters.
How about flash? In fact, I tried that first, bouncing the mecablitz 58-AF-2 off the ceiling. Here the flash photo before and after DxO processing, followed by the version taken with natural light before and after DxO (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Clearly DxO improves the detail. But somehow the flash version looks dead by comparison with natural light.
|
House lighting after incandescent
|
Topic: general, opinion | Link here |
Once upon a time, house lighting was simple: you put in a light globe, used it until it burnt out, and then replaced it.
But incandescent lighting is passé, and nowadays we have multiple alternatives. In the
pantry we had an LED globe and a fluorescent globe. One of them burnt out
failed. At a guess it would be the LED.
Took off the cover to look. No, the fluoro. Had it really failed, or had the fitting failed? Not a question we asked in the last millennium, but it has proven to be an issue in this one. Find a new fluoro, coincidentally a brand new one out of a package of 6, from which the other one also came. Put in. Dead.
Damn, what's wrong with the fitting? Tried the two in a standard lamp in the office. Both dead. On closer examination, I found:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
How did that happen? It seems that the construction has something to do with it: the adhesive was on all of them. I can't imagine that the damage came from external influence.
OK, try another one, one of the many old IKEA globes. Works:
|
|||||||||||||||||||||||
Yes, it should be marginally dimmer, but that's ridiculous. The IKEA globes take a while to warm up, but they won't get the chance in the pantry, where I seldom stay more than one or two minutes. OK, look for another LED globe... found! And it works.
What a pain! In the Good Old Days it would have been done right away.
|
What's that popup?
|
Topic: technology, opinion | Link here |
While working with distress, my Microsoft 10 box, I somehow conjured up this popup:
|
|
|||||||||
How did I get it? What produced it? What is it? What good is it? It clearly seems to be produced by Microsoft, not a third party application. I'll probably never understand this system, but then, even if I did, by that time it would have been replaced by something even more bizarre.
| Wednesday, 12 May 2021 | Dereel | Images for 12 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another cutting
|
Topic: gardening | Link here |
On the way back from our COVID-19 vaccination last week, Yvonne broke off a twig from a Salvia plant outside the Base Hospital. No need to let it go to waste: put it in water, and how about that, it flowered. So today it went into a pot:
|
|||||||||||||||||||||||
We'll see how that works.
| Thursday, 13 May 2021 | Dereel → Ballarat → Dereel | Images for 13 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Epicondylitis!
|
Topic: health, animals, opinion | Link here |
Off into town today to see Heather Dalman about a pain in my arm. Clear case: Epicondylitis, also known as Tennis Elbow. Gave me various things to do, which may or may not fix the problem.
Also talking about Pedro, who tends to lie on my legs in the evening while we're watching TV:
|
|||||||||||||||||||||
Heather was enthralled, thus the photo.
| Friday, 14 May 2021 | Dereel | Images for 14 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Council: time to give in?
|
Topic: Stones Road house, general, opinion | Link here |
I've been growling about the process of the planning permit application for the arena cover for months now, and precious little good has it done. My calls to Sarah Smith of the Council have once again gone unanswered, but now David Rowe has forwarded me a message that she sent to him during the time I was waiting for a call back:
The application will likely go one of two ways.
I will issue the permit with no endorsed plans and a condition that the colours be changed. Then that can be appealed to VCAT.
Other option is refusal based on not meeting the objectives of the zones and overlays. Then option again will be appeal the decision at VCAT.
Theres about a 6-8 month wait on appeals. Its likely we will go with option 1 as we support the shed in its context, just not the colour zinc.
Spelling is original.
What do I do? As a matter of principle, I would take it to VCAT, but I can't make Yvonne wait that long. Give in?
All this would have been much easier if David and Sarah had pulled their weight. But at some point I have to admit that I'm beaten. Today Yvonne went to Eureka sheds and came back with a set of Colorbond® cards. Now to find a way to get through to Sarah; I think we've had as much of David as we can handle.
| Saturday, 15 May 2021 | Dereel | Images for 15 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More mice!
|
Topic: animals, opinion | Link here |
Lately we've seen a couple of mice round the garage and laundry. I moved the mouse trap there to catch them, but it didn't catch any.
Why not? Did they steal the bait and get out again? Or is it just too old? Time to drag out that tube of mouse bait paste, which they can't completely remove. Bingo! Next time I looked, there was a dead mouse in the trap. Took it out, turned it on again, and an hour or so later there was another one, still warm. Piccola and Pedro were respectively grateful.
And no more? Maybe that's all there was. The trap is waiting.
|
Processing chestnuts
|
Topic: food and drink, opinion | Link here |
I've been dragging my feet about the chestnuts that Lorraine Carranza brought us. But they need at least cooking, and I had also bought meat to make the chicken and chestnut balls recipe that I had identified.
Do I really need to roast them? A quick search brought this page from Chestnuts Australia, which told me that I could cook them too, 15 to 20 minutes from bringing the cold water to the boil. That sounded a whole lot better than roasting them, but first I needed to halve them, which must have taken 15 minutes.
After boiling and cooling down, the fun wasn't over. It seems that the exact nature of the cut makes a big difference. Some of the chestnuts just fell out of the shell, others took a lot of work and even scraping out:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
After 30 minutes I had processed about half of them. Did I do something wrong? Does it depend on the chestnuts? Comparing the photo on the web site, it seems that I should have cut the chestnuts horizontally, not vertically:
When boiling, cut the chestnuts in half across the width of the chestnut before cooking.
And they should just have fallen out of the shell:
But even if I had done that, it would have taken the same 15 minutes at the beginning. I think that chestnuts, if I can persuade Yvonne to eat them, are best bought pre-processed.
| Sunday, 16 May 2021 | Dereel | Images for 16 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Chicken and chestnut balls
|
Topic: food and drink, opinion | Link here |
Spent much of the afternoon preparing my planned chicken and chestnut balls. The first issue was finishing shelling the chestnuts. Had I maybe not cooked them enough? Put the rest back in the pot and boiled for another 13 minutes, after which, sure enough, more chestnuts fell out of the shells. But not all:
|
|||||||||||||||||||||||
On closer examination, there were two kinds of chestnut flesh:
|
|
|||||||||
|
|
|||||||||
The yellow ones seem normal, and they fell out of the shell. The darker ones didn't. Are they too old? After all, I waited a couple of weeks before cooking them. That could explain a lot, including the fact that they didn't have much flavour.
Making the balls is a pain! First mince the meat and the chestnuts, then mix them in a mixer. What a lot to clean! Then make the balls, which were particularly sticky. And then, when it was too late, I realized that the recipe made no mention at all of salt!
The recipe wants the balls to be chilled for 20 minutes, presumably to make them easier to handle. But there's a simpler solution for that: warm them to 60°-70° in a microwave oven. For the balls in the second image, I used 5 minutes at 500 W.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
After that, they are firm and can be handled easily. For reasons I'm not sure about, I did them in a steel pan, and they stuck nicely:
|
|||||||||||||||||||||||
After that it was plain sailing, except that I forgot the spinach which, according to the recipe, I was to toss the dish through.
And the results? Not bad. But not much taste of chestnuts. Next time I will definitely use pre-prepared chestnuts. And what a mess!
|
|||||||||||||||||||||||
| Monday, 17 May 2021 | Dereel → Ballarat → Dereel | Images for 17 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Rain gauge strangeness
|
Topic: gardening, general, opinion | Link here |
I've had two rain gauges next to each other for nearly 6 years now:
|
|||||||||||||||||||||||
And I've also observed that the readings are inconsistent: the left-hand one almost invariably recorded less than the right one. And I couldn't work out why.
But now things have changed. Now the right-hand one reads less. Here from the table I made up 6 years ago, updated:
| Date | Gauge 1 (mm) | Gauge 2 (mm) | discrepancy | |||
| 13 July 2015 | 7.5 | 9.6 | 1.9 | |||
| 14 July 2015 | 1.7 | 3.2 | 1.5 | |||
| 15 July 2015 | 4.0 | 5.5 | 1.5 | |||
| ... | ||||||
| 11 May 2021 | 6.7 | 6.3 | -0.4 | |||
| 16 May 2021 | 8.3 | 7.3 | -1.0 | |||
| 17 May 2021 | 7.5 | 6.7 | -0.8 | |||
What causes that?
|
More anniversaries
|
Topic: history, opinion | Link here |
More anniversaries in my calendar today:
May 17 Greg starts working at Karstadt, 1976
May 17 Greg starts working at Tandem, 1982
May 17 Greg last flew in an aeroplane, 2006
There's nothing special about the day 17 May, and it was sheer coincidence that I started two jobs on the same day, especially in Germany, where employment always starts at the beginning of the month. But apart from the fact that I have gone 15 years without flying, it's interesting to note that it's a third of the time since I joined Karstadt. Looking back in increments of 15 years, we have:
Date Age Activity
@19610517 12 Greg in Melbourne, preparing for school in England
@19760517 27 Greg starts working at Karstadt
@19910517 42 Greg in last year at Tandem
@20060517 57 Greg last flew in an aeroplane
That's just about all my life. The real surprise is that I've spent nearly a quarter of it in Dereel.
|
Blood test results
|
Topic: health | Link here |
Into town today to hear the results of my blood test at the beginning of the month. Nothing of interest, except that the MCV is stubbornly high at 104 fl. It's been over the normal range since I've been measuring it, and since there are no other issues, we're beginning to wonder if it's some genetic thing. I suspect I have interested Paul in the matter enough that he'll go off and investigate.
|
Planning permit!
|
Topic: Stones Road house, general, opinion | Link here |
In the evening, received email: the planning permit (PDF) for the arena cover shed. Finally things are moving. Planned to go into Ballarat tomorrow: it seems that John wants some documents signed.
| Tuesday, 18 May 2021 | Dereel → Sebastopol → Dereel | Images for 18 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Disintegrating mobile phone
|
Topic: general, technology, opinion | Link here |
I've recently discovered some strange behaviour with my Nokia 3 mobile phone. The rear cover started peeling off, and now it's extreme:
|
|||||||||||||||||||||||
I can push it back in again—there are clips holding it in place—but it doesn't stay. What's causing that? Took a look inside, which showed little of interest:
|
|||||||||||||||||||||||
It seems that something inside is swelling, presumably the battery. Is this a result of it being on almost permanent charge? The charging circuitry should regulate that, but after all, this is a cheap phone.
I wouldn't object to replacing it—if I could find something that annoys me less. For the moment, I suppose I'll put up with it until it stops working or ALDI comes up with a new special offer.
|
Eureka: Yes! No! Yes! No!
|
Topic: Stones Road house, general | Link here |
John Hofman of Eureka Garages had told Yvonne her we would need to come in and sign some documentation. OK, we can combine that with a haircut for me. She called up John to make an appointment directly afterwards, but was told that we didn't need to come after all. Still later I spoke to him personally. Yes, we needed to come. And I had barely sat down when he called back to say no, no need, he found the document that he had mislaid.
|
Haircut: end of an era
|
Topic: general, history, opinion | Link here |
I had made an appointment for a haircut this afternoon in the assumption that I would combine it with a visit to John Hofman of Eureka Garages, but that was canceled, so I had to go in just for the haircut. It's something like the end of an era: until I moved to Dereel, I had made a point of having no two consecutive haircuts on the same continent for about 10 years. Now that I no longer travel, I have been having my hair cut by Kerry in Sebastopol for nearly 14 years. He wasn't young then, but he's gradually getting old and feeble, so at Yvonne's recommendation I went to Arabella, hairdressers and not barbers, about 60 m further north.
Yvonne has been going there for some years, and had recommended Lynne, who remembers Yvonne well—surprising, for a single client who comes every three months. She was pleased to hear about the planning permit, and asked whether we had had to accept Colorbond® or not. I wonder how many hairdressers could remember that much detail.
| Wednesday, 19 May 2021 | Dereel | Images for 19 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Smart phones: the pain
|
Topic: technology, opinion | Link here |
It proves useful to have both this diary and IRC contact. After posting yesterday's article about my phone, a lot of people were up in arms: don't use it, it could catch fire.
Oh. Yes, I knew about mobile phone batteries catching fire, but they were Sλmsung, not Nokia. On reflection, yes, potentially it could. But if that were the case, there would be many more warnings about it.
OK, nothing for it, take the thing apart and look inside:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
In these photos, the “top” of the phone is on the right.
Yes, no doubt about it: the battery has swollen significantly. What should I do about it? Is it worth replacing? Rang around in town and finally found one repairer who was prepared to repair Nokia. They wanted $89 for the job. And that for a phone that cost me $98.
OK, off to eBay to see what they had. How about that, they're available for $14.88, apparently a bad number nowadays. In my day 1488 was an RS-232 driver chip. Now it appears to be a hate symbol, presumably not related to dislike of RS-232.
So how do I get the battery out? There's a little flat band cable that goes down under a cover, which appears to be in two parts: the one on the right (bottom on the photo above) covers the connector and is held in place with two screws. Here seen the right way round, with the top up:
|
|||||||||||||||||||||||
Remove them and... discover that the cover is in a single part, with a ruled line to confuse people like me. OK, remove the other 4 screws. Still won't come off. The chrome screw at bottom left isn't an adjusting screw, just another one. Why is it chrome?
Daniel O'Connor pointed me at this instruction page, which shows what I had already discovered, but this time the chrome screw was in a different place:
OK, remove it too. But how? I couldn't get purchase on the screw, and I thought that I had stripped it. Once again a camera to the rescue:
|
|
|||||||||
The screw is tiny (but undamaged). But what's that gunk? And how do I get it out? Used the smallest flat-blade screwdriver I could find and pressed a bit harder, and finally got it out. But why do people do this?
|
|||||||||||||||||||||||
Underneath is not encouraging. Yes, there's some kind of plug on the battery, but the contacts look like this (bottom right):
|
|
|||||||||
Am I going to be able to reconnect it? Maybe I should just give up the whole thing as a bad job. I have better ways to spend my time.
|
A use for old macro lenses
|
Topic: photography, opinion | Link here |
Taking the photos of the mobile phone wasn't easy. Many reflective surfaces, and big differences between the (almost) white of the case and the black of the innards. Once again the light tent that I bought a couple of months ago came in handy. But that means taking the photos through the hole in the top, a (more or less) fixed distance from the subject. What I need is a zoom macro lens.
I don't have that, of course, but I do have five different macro lenses, with focal lengths 30 mm, 35 mm, 50 mm (twice) and 60 mm. And with exception of the SMC Macro-Takumar 50 mm f/4, I used all of them today. I never thought I'd use the Zuiko Digital 35 mm f/3.5 Macro or the Zuiko Digital ED 50 mm f/2.0 Macro again.
|
Fish cakes: not our kind
|
Topic: food and drink, opinion | Link here |
Yvonne was obviously hungry when she went shopping today, and came back with no fewer than three pre-prepared meals that weren't on our weekly menu. The one she wanted to eat today looked particularly dubious: “Home made salmon patties” that looked nothing like fish.
How do you prepare them? Frying pan? They appeared to be mainly potato and turmeric, and they were falling apart even before I tried to put them in the pan. With Yvonne's help, managed to heat them and serve them:
|
|||||||||||||||||||||||
Ugh! Yes, there's some fish in there, but how can you tell if it was salmon? Looking carefully, all I can see is what could be cauliflower and either onion or celery:
|
|
|||||||||
|
|
|||||||||
The flavour was completely boring, and that for $10!
|
|||||||||||||||||||||||
For that price we could have had two good steaks. Yvonne was in agreement that it wasn't worth eating, so she'll take it back next week. I'll be interested to hear what they have to say for themselves. Until now we liked the shop.
| Thursday, 20 May 2021 | Dereel → Napoleons → Dereel | Images for 20 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Garden in late autumn
|
Topic: gardening | Link here |
Today was a month before the solstice, time for the monthly garden photos.
I've been watching the Buddleja weyeriana for some time. It's still flowering, but not looking overly happy. Here last month and this month:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, it's still alive and flowering, but the leaves are looking unhappy. It might be the weather, but it's about time I took some cuttings and also repotted the main plant.
For what seems the first time, the trees in the front garden (birch and cherry) both look like autumn:
|
|||||||||||||||||||||||
It may have been a bad year for the garden, but the north garden (up against the north side of the house) has really filled out, in particular the Alyogyne huegelii and the Cannas:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
As a result, the Hibiscus rosa-sinensis is well protected and flowering happily, a far cry from the way it looked in the spring (first photo):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
There, too, I need to make some cuttings.
And practically for the first time, our lime bush has borne more than one or two fruit:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It's still not much to show for nearly 13 years.
On the other hand, the tomatoes have been an absolute catastrophe this year. The ones I planted in soil are still not ripe:
|
|||||||||||||||||||||||
And the ones I left, neglected, in a pot, have only now started to ripen:
|
|||||||||||||||||||||||
One surprise was a half-circle of mushrooms in the front garden:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
My understanding is that these circles grow outwards, but this is the first time I've seen them.
|
What's a mobile phone good for?
|
Topic: technology, general, opinion | Link here |
Once again I'm spending far too much time on a topic that doesn't really make me happy: mobile (“smart”) phones. I've decided to cut my losses in one direction: I won't repair it. I clearly also won't pay $89 to replace a battery on the phone when I can buy a brand new one for $69.
But do I want to buy a new Nokia 3? I think it has been particularly painful (thus the “smart”), and if I'm going to buy a new phone I should probably get something with an interface that is closer to the norm. Sλmsung maybe? Asked for advice on IRC and got answers like “no phone is good enough for you”, which might be correct.
But what do people use phones for? Asked around and came up with (after rearranging and interpreting):
What's missing from this picture? Telephone calls! And people agreed that that came very far down on their list. Also photos, though high-end phones can take quite decent photos if you're not overly picky.
So there are three main uses. “Mobile terminal” seems to cover many of them, including web access, social media, text chat. All of these require using the horrible keyboard substitute. Callum Gibson relativized it, though:
<cgibson> Unless you want to use horrible voice recognition is there an alternative if you
are out and about and not near a computer?
Then there are podcasts and “Movies, Netflix”. Somehow the approach is schizophrenic: at one end of the scale people sell bigger and bigger TV displays. My TV has a 75" (diagonal) screen, an area of 2.4 m². My Nokia 3 has a reasonably typical 6.3 x 11 cm display, 69 cm² or about 2.9% of the TV size. Why should people prefer that, or even put up with it?
And finally there are games. Maybe mobile phones really are a good choice for games, but I don't play games.
The real issue is that they're a workaround for people's lifestyle issues: people are out and about all the time. That's a pre-COVID-19 mentality, and it may change afterwards. But there's another issue: interoperability. I have my environment with me when I carry my mobile phone with me. I don't when I access the net via anonymous Microsoft computers. So at least part of the issue is interoperability, which mobile phones have destroyed almost completely.
|
New camera
|
Topic: photography, opinion | Link here |
Over the weekend I bought a Praktica Nova camera on eBay, mainly because it was so cheap: $5. It arrived today, so off to Napoleons to pick it up.
What can I say about it? The best I can say is “How are the mighty fallen”. Praktica is the successor of one of the earliest modern SLRs, but compared to the Pentax of the day it looks really rough. I thought that this one had a light meter, but it seems not; in any case, the versions with light meter weren't coupled with the camera itself. Very primitive by comparison.
I'll take photos some other time.
| Friday, 21 May 2021 | Dereel | Images for 21 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More power failures?
|
Topic: Stones Road house, general, opinion | Link here |
Two more grid power failures today. How long? The first one was presumably a one-second job, but for the second my log says:
2021-05-21 09:49:22 1760 Off-grid 45 268 NULL -828 2588 0 0
2021-05-21 09:49:26 1416 Waiting to connect to On-grid 45 267 NULL -1183 2599 0 0
...
2021-05-21 09:50:26 959 Waiting to connect to On-grid 45 268 NULL -1690 2649 0 0
For some reason, I missed status messages for 3 seconds at the beginning. But the last one was at 9:50:26, and I know that the inverter waits for exactly 60 seconds before trying to reconnect. So I'm guessing that the missing messages all had an Off-grid status, and that this was thus a 5 second outage. But why did the messages go missing?
|
What's that clunk?
|
Topic: technology, general, opinion | Link here |
While downloading videos this morning, heard a repetitive “clunk-a-clunk-it” coming from eureka about once every 2 seconds. Must be the disk. Why are these things so noisy? But it wasn't always so, and it continued unabated after a couple of syncs.. It also continued after the downloads were finished.
What's accessing the disk? Off to look with top, which showed random access to the root file system, and iostat, which showed relatively continuous access to the disk on which it was located. But what's causing it? Stopped likely processes like the four firefox instances, MySQL and MediathekView. No change.
Network? Disconnected the local network cable. No change.
Gradually it began to look like a dying disk. Yes, I have backups, most recently:
Dumping completed at Thu 20 May 2021 21:40:18 AEST
But recovering from backup would require three passes (level 0, 1 and 2 dumps or tar backups) for two different file systems. It would be simpler to copy the data to the new disk before the old one dies.
Which disk was it, anyway? The case is open, so I could just put my hand on one of the disks:
|
|||||||||||||||||||||||
And how about that, I felt the clunk rhythm right away. Tried another disk. Same thing! I only really have 2 disks on the system, the system disk with root and /home, and the photo disk with, well, photos. The system disk gets much more activity, so I'd guess that it's the one that's dying.
New disk? Do I have one? Surprisingly, yes, a 6 TB drive that was originally the system disk for teevee until it overflowed only a year later. But it's almost too big for eureka, which has:
=== grog@eureka (/dev/pts/24) ~ 7 -> df
Filesystem 1048576-blocks Used Avail Capacity Mounted on
/dev/ada0p4 39,662 27,156 9,333 74% /
/dev/ada0p2 39,662 13,447 23,042 37% /destdir
/dev/ada0p5 3,705,520 2,199,692 1,209,386 65% /home
But it was the only one over 1 TB, and it's only 3 years old, so it should be OK. How do I use it? It's partitioned almost the same as the disk on eureka. OK, put it in eureso, create new file systems and copy the data across.
Oh.
ld-elf.so.1: Undefined symbol "rl_filename_rewrite_hook" referenced from COPY relocation in /usr/local/bin/bash
I've seen this before, in fact 3 months ago and put it on the tuit queue. I think it has aged enough now: the whole system isn't worth keeping.
Back to the present problem: how do I even log in? toor is your friend, if you have a password. I didn't, so I had to reboot single user and set one.
OK, create the file systems and start copying the data across. 2.2 TB will take a while, so set to preparing the root file system for reboot on eureso as, well, eureso:
--- rc.conf 2021/05/22 02:02:28 1.42
+++ rc.conf 2021/05/22 02:03:05
@@ -2,7 +2,7 @@
-FORREAL=1
+FORREAL=0
if [ $FORREAL -eq 0 ]; then
hostname="eureso.lemis.com"
hostip=192.109.197.192
...
And then wait. The data came across at about 70 MB/s, quite acceptable for a 1 Gb/s link, but it still meant waiting all day. By evening I had transferred 1.3 TB.
| Saturday, 22 May 2021 | Dereel | Images for 22 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Disk recovery, continued
|
Topic: technology, opinion | Link here |
Yesterday's disk recovery went smoothly enough, but by evening I had only transferred 1.3 TB or so. Came in this morning, and... the file system still had only 1.3 TB in it.
Looking around, discovered:
May 21 18:03:58 eureka kernel: em0: TX Queue 0 ------
May 21 18:03:58 eureka kernel: em0: hw tdh = 543, hw tdt = 769
May 21 18:03:58 eureka kernel: em0: Tx Queue Status = -2147483648
May 21 18:03:58 eureka kernel: em0: TX descriptors avail = 790
May 21 18:03:58 eureka kernel: em0: Tx Descriptors avail failure = 0
May 21 18:03:58 eureka kernel: em0: RX Queue 0 ------
May 21 18:03:58 eureka kernel: em0: hw rdh = 110, hw rdt = 109
May 21 18:03:58 eureka kernel: em0: RX discarded packets = 0
May 21 18:03:58 eureka kernel: em0: RX Next to Check = 110
May 21 18:03:58 eureka kernel: em0: RX Next to Refresh = 109
May 21 18:03:58 eureka kernel: em0: link state changed to DOWN
May 21 18:04:01 eureka kernel: em0: link state changed to UP
Why did that happen? It was recoverable, but presumably not for tar.
Damn. Could it happen again? If so, restarting the tar didn't sound like a good idea. Time for rsync.
But first, let's boot from the new disk. Set the details in /boot/loader.conf and boot. It still came up with the old system!
OK, disconnect the first disk cables and boot. It tried a net boot, and when I aborted that, told me that there was no “Operating system”. At a guess, the system (a ThinkCentre M93p) isn't set up to boot from anything except the first disk. Given that it expects a DVD drive as the second disk, that's marginally acceptable. But by the time I thought of that, I had rebooted with the old disk in place, so it'll have to wait. And only much later did it occur to me that I had modified the wrong /boot/loader.conf: the first disk had two root file systems, and I had been running from the second, but you need to set the partitions in the loader.conf in the first partition, regardless of which you want to boot. And of course I had modified the second.
Fired rsync up and watched it go. Much slower than tar. By evening it still wasn't done.
On the positive side, the clunking noise was gone. Why? That depends on what caused the clunking, of course, but it's still a good idea to have a hot backup ready to go if the disk does die.
Another strangeness: the display on the monitor looks broken:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What causes that? I don't think it's a defect in the monitor itself. It seems to happen after booting, when the system changes the font (why?). I've probably never noticed it before because I almost always run X.
|
Computer failures gang up
|
Topic: technology, opinion | Link here |
Checking my weather information today showed a surprise: nothing! Did I do something wrong with my recovery efforts yesterday? They shouldn't have affected the weather station, which is connected to teevee.
Much searching. No, MySQL still running on eureka, weather software still running on teevee. But the weather station didn't want to talk to it. Disconnect and reconnect the USB connection, and it started again. A random issue, unrelated to the other problems.
|
High-resolution panoramas
|
Topic: photography, technology, opinion | Link here |
House photo day today, and bright sunshine. Time to try out a couple of things I've been thinking about for a while. First, how about a high-resolution panorama? Currently I take 4 photos at 90° angles and get a panorama with up to 90 MP, depending on cropping. I had tried before with longer focal length lenses, but getting them to align correctly proved to be more pain than I wanted to go to. So how about using the Olympus OM-D E-M1 Mark II's high resolution shot feature, 80 MP per image instead of 20?
First problem: I take my images using the in-camera HDR bracketing feature, 3 images offset by 3 EV. But that doesn't work in conjunction with Hi-Res. So I had to bracket manually, more work than I had expected. And then a problem: I had to hold my hand in front of the sun for one of the shots to the north. And in Hi-Res mode the viewfinder turns off, so I couldn't keep an eye on my hand. In the end took three photos until I got one right, in the process forgetting the one without the hand:
|
|||||||||||||||||||||||
And for the fourth photo, Pedro kept running round, and I needed no fewer than 4 images.
OK, process the photos. distress and DxO PhotoLab got quite upset about the size of the images, it seems. I couldn't tell, because it stopped responding to keyboard and mouse input. OK, disconnect the rdesktop session and reconnect. No, sorry, don't want to know:
=== grog@eureka (/dev/pts/20) ~ 11 -> /home/local/bin/dordesktop distress 2510x1390+0
User groggyhimself sword Scabbard
Connecting to distress
Connection established using SSL.
disconnect: Unknown reason.
Disconnected from distress, status 63
For 11 images it took nearly half an hour:
-rwxr--r-- 1 grog wheel 484,079,030 22 May 14:20 45229691_DxO.tif
-rwxr--r-- 1 grog wheel 484,078,944 22 May 14:22 45229690_DxO.tif
-rwxr--r-- 1 grog wheel 484,078,706 22 May 14:23 45229689_DxO.tif
-rwxr--r-- 1 grog wheel 484,078,882 22 May 14:24 45229692_DxO.tif
-rwxr--r-- 1 grog wheel 484,078,930 22 May 14:38 45229695_DxO.tif
-rwxr--r-- 1 grog wheel 484,077,184 22 May 14:41 45229693_DxO.tif
-rwxr--r-- 1 grog wheel 484,077,248 22 May 14:41 45229694_DxO.tif
-rwxr--r-- 1 grog wheel 484,078,944 22 May 14:46 45229696_DxO.tif
Normally DxO converts about 5 images a minute.
Round about here I called it a day. Mañana.
|
Lens flare revisited
|
Topic: photography, opinion | Link here |
My oldest really wide angle lens is the Zuiko Digital ED 9-18 mm f/4.0-5.6, which I got over 12 years ago. Like all lenses, it suffers from flare and reflections in the sun. But when I got the Olympus Zuiko Digital ED 8 mm f/3.5 fisheye lens, I noted that it was much better.
OK, I now have no fewer than 8 lenses with a focal length of 12 mm or less (corresponding to 24 mm on a 35 mm camera). And today the sun was shining. How about comparing them (run the cursor over an image to compare it with its neighbour)?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
In order, the images were taken with the Olympus M.Zuiko Digital ED 8 mm f/1.8 Fisheye PRO, the Olympus Zuiko Digital ED 8 mm f/3.5 fisheye lens, the M.Zuiko DIGITAL ED 7-14 mm f/2.8 PRO, the Zuiko Digital ED 9-18 mm f/4.0-5.6, the Zuiko Digital ED 12-60 mm f/2.8-4.0 SWD, the Leica DG Vario-Elmarit 12-60 mm f/2.8-4, the M.Zuiko Digital ED 12-100 mm f/4.0 IS PRO and the Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3. I have (exceptionally) allowed DxO to “defish” the fisheye images.
The results weren't quite what I expected: the spots in the image look like dust, though I can't see much on the lenses. And there aren't that many reflections, though clearly the 8 mm f/1.8 really does the best,
| Sunday, 23 May 2021 | Dereel | Images for 23 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Disk recovery, continued
|
Topic: technology, opinion | Link here |
Finally my disk copy is done! Of course, it was already out of date, since the original disk had been updated. Time to run another rsync.. To my surprise, it ran for over 2½ hours and transferred 22 GB! It seems that I update quite a bit of stuff on a regular basis.
But there are still no obvious signs that the old disk is dying. In fact, there's not even a reason to believe that the problem is with that disk; it could also be the photo disk, or something else altogether, though that sounds less likely. Still, I'm receiving some new hardware next weekend, so maybe the easiest thing would be to do regular rsyncs until then.
|
Recycling kangaroos
|
Topic: animals, general | Link here |
Three weeks ago I found the remains of a dead kangaroo in the sheep paddock across the road, once again the victim of a wire fence:
|
|||||||||||||||||||||||
The leg disappeared soon after, but today Nikolai found it:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That kept him busy for a while, and it's a whole lot safer than chasing kangaroos.
|
Still more technology pain
|
Topic: technology, opinion | Link here |
Somehow everything seems to be running at a snail's pace lately. I might suspect that it has something to do with the disk issues on eureka, but that doesn't explain why distress, my Microsoft box, seems to be running at 100% disk so often, apparently in system processes such as “Malware removal”. I'm reminded of an ancient German story about maintaining system reliability with a program called „Oremus“ (“let us pray”). It doesn't even seem worth looking for it online: it would drown in false positives. If I recall correctly, it kept the computer from crashing by using up all CPU time. Why should that help? I don't recall. In any case, my photo processing on distress was even slower than normal today.
|
History and yet more technology pain
|
Topic: history, technology, opinion | Link here |
Jerry Dunham sent me an interesting historical photo today:
|
|||||||||||||||||||||
That was a photo of the Tandem Computers employees in the Ben-Gurion-Ring facility done by Fritz Joern round 1987, and he had posted it on a Tandem alumni group on groups.io. Why am I not a member? Jerry told me how to sign up, and of course I had to give an email address and wait for a message.
After some time, none had come. OK, what do we have email logs for? I don't know:
=== root@eureka (/dev/pts/4) /var/log 194 -> ls -l maillog
-rw-r----- 1 root wheel 60 23 May 00:00 maillog
That's not much. What's in it?
=== root@eureka (/dev/pts/4) /var/log 201 -> cat maillog.hang
May 23 00:00:00 eureso newsyslog[4083]: logfile turned over
Had postfix stopped running? No, I had been receiving mail as usual, but it hadn't been logged. Restart postfix? Nothing logged, not even the restart. Config changed? No. Where do I specify the log file, anyway? I don't!
It took me a while to realize that postfix (presumably like all MTAs) logs via syslog, so the real entry is in /etc/syslogd.conf:
mail.info /var/log/maillog
OK, what does logger say? Nothing!
But syslogd seemed to be working for other facilities. Wake up? No difference. I had to restart it to get it to listen again.
In passing, it seems that both eureka and eureso had been logging to the same file. There's nothing wrong with that, of course.
And groups.io? Just slow. Got the message, signed in and applied for membership in the group. No reply today.
|
The high resolution panorama, finally
|
Topic: photography, opinion | Link here |
Stitching yesterday's high-resolution panorama was more of a problem than I had expected. Firstly, I had to use Photomatix in manual mode (normally I use the batch facility) to convert the images, and despite using default settings they came out with strangely exaggerated colours and contrast. Maybe that was why Hugin didn't want to align them, though the normal resolution photos taken from exactly that position worked fine.
Finally managed to align them by adding images one at a time, starting with the last, which looked particularly strange. Here the normal resolution one and the high-resolution one, both merged by Photomatix:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Finally got round to stitching the panorama. 319 megapixels!
Oh:
Blending images...
enblend: info: loading next image: house-from-n-High-res-3-house-from-n-High-res-40000.tif 1/1
enblend: info: loading next image: house-from-n-High-res-3-house-from-n-High-res-40001.tif 1/1
enblend: No space left on device
enblend: an exception occured
enblend:
enblend: error writing to image swap file.
Most likely cause: No space for temporary files.
Make sure that there is enough space in the temporary directory
Now isn't it nice of it to say what file name, or what file system? Presumably the root file system, but it had 15 GB free. Some searching, in which I discovered that the Hugin preferences for temporary files were set to “use system default” (there isn't one). Presumably it had used /tmp. Set to /var/tmp, and all was well.
|
|||||||||||||||||||||||
And the results? Really not good. Here a crop from the normal resolution panorama and the high resolution panorama:
|
|
|||||||||
|
|
|||||||||
If anything, the normal resolution image looks sharper. There's a clear reason for this: Photomatix didn't align the images correctly. So I have learnt something, but the results aren't worth having. If I try again I'll use single images.
| Monday, 24 May 2021 | Dereel | Images for 24 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Welcome to TandemAlumni
|
Topic: general, technology, opinion | Link here |
Mail message today: my application for membership of the groups.io TandemAlumni group was accepted. But why? It seems that my name wasn't stored as part of my membership application.
OK, now I can look at the photo that Fritz posted:
|
|||||||||||||||||||||
That's a classical example of requiring “click to enlarge” to see anything. But it doesn't work! Only my copy shows it. after 3 clicks. Here in original size:
|
|||||||||||||||||||||
|
|||||||||||||||||||||
The photo is clearly a photo of a photo, and I have my doubts as to its accuracy. I'm pretty sure that Monika Klein left some time before 1987, and Richard Carr never worked from that location. In addition, Harald Sammer and his HPRC people must have left for Friedrichsdorf before then. But it's an interesting document.
|
Walking the dogs in the Eagle's Nest
|
Topic: general, opinion | Link here |
When we lived in Kleins Road we used to walk the dogs to the south of the house, starting from a bend where Kleins Road became Swamp Road. There we once found what we think was an eagle's nest blown out of a tree:
|
|||||||||||||||||||||||
Down there again today. Things look different in the forest:
|
|||||||||||||||||||||||
That's the result of a burnoff a couple of weeks ago. But the overall appearance isn't nearly as pleasant as we recalled.
|
Still more system strangenesses
|
Topic: technology, opinion | Link here |
Seen in my daily security check emails today:
eureka.lemis.com setuid diffs:
--- /var/log/setuid.today 2021-05-23 04:09:21.202725000 +1000
+++ /tmp/security.mc8GJjo2 2021-05-24 02:18:33.930816000 +1000
@@ -1,74 +1,74 @@
- 1524878 -r-sr-xr-x 1 root wheel 19952 Nov 25 11:45:59 2015 /bin/rcp
- 86133 -r-sr-xr-x 1 root wheel 21016 Nov 15 12:19:38 2017 /destdir/bin/rcp
- 86504 -r-sr-xr-- 1 root operator 10592 Nov 15 12:20:19 2017 /destdir/sbin/mksnap_ffs
- 86524 -r-sr-xr-x 1 root wheel 32480 Nov 15 12:20:20 2017 /destdir/sbin/ping
...
+ 3370766 -r-sr-xr-x 1 root wheel 19952 Nov 25 11:45:59 2015 /bin/rcp
+ 323774 -r-sr-xr-x 1 root wheel 21120 May 6 14:55:20 2018 /destdir/bin/rcp
+ 3933429 -r-sr-xr-- 1 root operator 15376 Mar 22 08:43:02 2019 /destdir/sbin/mksnap_ffs
+ 3933442 -r-sr-xr-x 1 root wheel 40080 Mar 22 08:43:03 2019 /destdir/sbin/ping
There were plenty more, about 80 of them. What's that? No program that I use, especially not the ones in /destdir (the old, mouldy root partition waiting to be overwritten). But how could it happen? Some of the files are different in size, many have different modification timestamps, and all of them have different inode numbers. It took some time to realize why: it's the other side of the coin from yesterday's problem with the mail log: eureso and eureka are logging to the same place.
This asks a more basic question: how do you share log files across systems? Yesterday I thought that one maillog file would be enough, and in principle it is. But I have two copies of syslogd running on different systems, both logging to the same file. Clearly that won't work.
In general, what will work? I'm coming more and more to the conclusion that local file systems should be shared between all systems on the LAN. But clearly some directories should be local, and /var/log is one of them.
| Tuesday, 25 May 2021 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another bloody migraine
|
Topic: health | Link here |
I've had migraine on and off for over 30 years. Since it's not associated with headache, it took me a long time to understand what the aura was. And I thought that since I started taking regular aspirin (originally intended for heart problem prophylaxis) it would go away. But no, from time to time it still hits me. Today it was mild, and it was gone within 30 minutes.
|
Tortilla mix, again
|
Topic: food and drink, opinion | Link here |
Huevos rancheros for breakfast today, and once again head-scratching about the tortillas. The masa harina that I bought in January has not proven to be the best I have ever had. I soon established that it needs much less water than other kinds, 1.45 parts water to 1 part masa.
But that was the last package. I bought 5, and I'm now on the second package. And for some reason it needs more water. I've gradually been increasing the quantity, and now I'm up to 1.6:1. Why? Different moisture content? The difference between 1.45:1 and 1.6:1 would mean about 125 g of water in the (neither metric nor US Customary) 816 g package. That's not plausible.
|
Yet another “new” camera
|
Topic: photography, opinion | Link here |
Yvonne went shopping today and came back with my latest “new” camera, really the oldest I have: an Exakta II body, serial number 651435, once again the oldest camera I have. According to Andrzej Wrotniak's listing, it's subtype 1, made in 1949. Interestingly the serial numbers overlap with the original Kine Exakta.
So now I have three cameras made before 1960. In order of age:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's the “new” Exakta II (1949), Asahiflex 1a (1951) and the Edixa Reflex (1959?). I now have enough to recognize some trends:
The two oldest cameras don't just come without a pentaprism, they can't be fitted with one. That came with the Exakta Varex and the original Asahi Pentax, though later models of the Exakta II than mine could be adapted. The Edixa, like the Nikon F, has interchangeable viewfinders.
The German cameras have their serial numbers on the body, inside the film compartment (so they can't be inspected when there's a film in the body). All Japanese cameras that I have seen have the numbers on the top plate. But my Leica copy (Fed) also had the number on top.
The Exakta, the first ever 35 mm film SLR, looks remarkably primitive by comparison with the other cameras. That's not just the age: things like the overall appearance didn't change that much for as long as they were produced. Here the inside and the film pressure plate:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
By comparison, the Asahiflex looks almost the same as the Pentax KM, nearly 25 years later:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The 1950s lenses were particularly primitive. On both I have difficulty adjusting the aperture without changing the focus. Here the Takumar 50 mm f/3.5 and the Steinheil Cassar 50 mm f/2.8:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
New phone, dammit!
|
Topic: technology, general, opinion | Link here |
I've been thinking, researching and discussing a new mobile phone for several days now. That's wasting more time than I can afford. Made a decision: a Xiaomi Redmi 9T with 6 GB of main memory and 128 GB of flash storage.
Why that?
| Wednesday, 26 May 2021 | Dereel | Images for 26 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Whose fault is SARS-CoV-2?
|
Topic: health, opinion, technology | Link here |
During the bad Donald Trump era, He Himself proclaimed that SARS-CoV-2 “Coronavirus” was “The Chinese Virus”. Thus it was that I was happy
to find this article, referring to this
one, which shows in some technical detail that the discovery of the virus in
Wuhan was by no means the beginning
of the story. I discussed it two weeks ago.
But now people in the USA are getting up on their hind legs again and suggesting that it did escape from a laboratory in Wuhan, like this article in the Washington Post. The Washington Post isn't exactly trash press, but it is US American. And it seems that President Biden is getting in on the act. I read a couple of articles presenting circumstantial evidence (“they're the only lab in China doing research into coronaviruses, so there must be a connection”), but nothing nearly as nitty-gritty as the articles I saw 2 weeks ago.
Could they both be correct? The genetic analysis analyses mutations, not intentions. But it also seems that the initial infections in Italy were from an ancestor of the virus first detected in Wuhan. It doesn't say where the common ancestors come from, though it assumes “China”. Doubtless more investigation can't do any harm, if only to confirm that it wasn't a lab accident.
In the process, I came across this URL, sadly with very broken markup, which I think is intended to convey the relationships between individual mutations of the virus. In its current form it appears to be almost useless.
|
How much rice?
|
Topic: food and drink, opinion | Link here |
I had asked Yvonne to buy some Basmati rice today. She brought it back. How much rice?
|
|||||||||||||||||||||||
Where's the weight? Surely it must be there somewhere. After much searching, I found:
|
|
|||||||||
When you consider that photos usually make things more legible, you can imagine what the original looks like. Now surely that must be difficult for people with better eyes than mine. Why does ALDI do this?
|
Whence these security warnings?
|
Topic: technology, opinion | Link here |
On Monday I was puzzled by the false positive security messages I had been receiving. And I found out what the problem was: sharing /var/log between eureka and eureso.
Problem: I'm not. eureso stores its log files locally. So where is it coming from. Looking at just one file, /bin/rcp, I had on Monday:
--- /var/log/setuid.today 2021-05-23 04:09:21.202725000 +1000
+++ /tmp/security.mc8GJjo2 2021-05-24 02:18:33.930816000 +1000
- 1524878 -r-sr-xr-x 1 root wheel 19952 Nov 25 11:45:59 2015 /bin/rcp
+ 3370766 -r-sr-xr-x 1 root wheel 19952 Nov 25 11:45:59 2015 /bin/rcp
The only difference is the inode number. Which does eureka have?
=== root@eureka (/dev/pts/22) ~ 81 -> l -iT /bin/rcp
3370766 -r-sr-xr-x 1 root wheel 19,952 25 Nov 11:45:59 2015 /bin/rcp
And eureso?
=== root@eureso (/dev/pts/22) ~ 299 -> ls -liT /bin/rcp
1524878 -r-sr-xr-x 1 root wheel 19,952 25 Nov 11:45:59 2015 /bin/rcp
So yes, it's something to do with eureso. But what? The key would be the first file, /var/log/setuid.today. But eureso symlinks /var/log to /home/var/log. Where's the problem there?
/home, of course! That's mounted on /eureka!
=== root@eureso (/dev/pts/4) ~ 300 -> df /home
Filesystem 1048576-blocks Used Avail Capacity Mounted on
eureka:/home 3,705,520 2,190,480 1,218,598 64% /home
Somehow this is all more complicated than it should be.
| Thursday, 27 May 2021 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Lockdown!
|
Topic: health, animals, food and drink, technology, general, gardening, opinion | Link here |
Victoria has been without any new locally transmitted COVID-19 cases for 45 days now. Or at least, that was the case a couple of days ago. Now we have 35 of them! Time for a fourth lockdown. It'll last a week or so.
What superb timing! Over the next 3 days we had planned:
|
Coffee machine failure
|
Topic: food and drink, opinion | Link here |
Turning the coffee machine on today proved more of a problem than I had expected. Nothing happened. I've seen that before: the power strip that feeds it seems to have less than reliable contacts. Wobble the plug, and the lights went on.
Oh. Red LED lights up. Water container not inserted correctly? No, it's a different one, an red triangle that is clearly intuitive. Off to find the instruction manual. Possible cause? “The inside of the appliance is very dirty”.
Isn't that specific? OK, where could it accumulate dirt? It tends to spread a bit of coffee powder, but not much. A bit more searching found a door and a device that is apparently called an infuser, but which clearly includes the grinder. Yes, a bit of encrustation on it. OK, clean it, replace it... still the red triangle.
Dammit, this device is only 18 months old. On the one hand, that means that it's still under warranty, but I would still have to wait for its repair. The DeLonghi agents are John Thomas, who have already made themselves unpopular. I can't expect them to get it fixed soon: last time it took much negotiation and a month of waiting to get things done.
OK, buy a replacement one and have the old one repaired at John Thomas' leisure. What do The Good Guys (about the only game in town) have to offer? The cheapest was the DeLonghi Magnifica S, exactly the model I have and didn't like much even before it died on me—only $749, $100 more than I paid for this one. Never mind, there are others, up to the Jura Giga 6, for exactly $6,490, 10 times what I paid for mine. I don't even want to investigate what could justify that kind of price.
So what do I do? A little detail occurred to me: after putting the infuser back in the coffee machine and turning it on, the red triangle lit immediately. That doesn't look like mechanical problems: it would first need to try to do something. It could, of course, mean electrical issues. But how about disconnecting the thing completely (pull the plug), reconnect and turn it on again? Did that. Worked!
So here, it would seem, we have another case of poor electronic design. Was it necessary to clean the infuser? I'll never know, but clearly it wasn't the wrong thing to do.
|
Pedro: still not house-trained
|
Topic: animals, opinion | Link here |
We've had Pedro for nearly 2 months now, and Yvonne can get him to urinate and defecate on command. But he still doesn't go to too much trouble to go outside when nature calls, and today we once again found a pool of urine in the hallway. When will he ever learn?
|
Updating GPS navigator
|
Topic: technology, opinion | Link here |
Since wanting to return the new GPS navigator six weeks ago I haven't really done anything with it, though I got both a refund and the device. As a result, I gave the old one to Petra Gietz' granddaughter Kaylin (or similar). And I have newer maps than the ones in the device, so how about installing them?
I've been through these directory trees before, and there wasn't anything particularly difficult this time. But what happens if the software doesn't like the maps? The map directory is (relative to the device storage) AZ\ 2020/CONTENT/Map/. But the new image has AZ\ 2021/CONTENT/map/. Does the device care about capitalization?
And then there were more files in the new directory. After cleaning suffixes and things, we have:
AZ\ 2020:
Australia_HERE_Easy_2019.hnr
Australia_HERE_Fast_2019.hnr
Australia_HERE_Short_2019.hnr
Australia_R3_HERE_2019.fbl
Australia_R3_HERE_2019.fsp
AZ\ 2021: As above, also
Australia_HERE_2020.fda
Australia_HERE_2020.fjw
Australia_HERE_2020.fpa
Australia_HERE_2020.ftr
Australia_HERE_Economical_2020.hnr
Australia_HERE_Green_2020.hnr
Installed them anyway, and sure enough it worked. But what are the other files? Presumably the newer version of the software can handle them. Why not install everything? The device has an 8 GB memory, and the software only takes up a little over 2 GB, so there's space for both. Renamed the AZ\ 2020 so that the software wouldn't choose it, and installed AZ\ 2021. Crash! Clearly it was looking for something it couldn't find. And that's easy to guess: there's a second top level directory in the system, System\ Volume\ Information:
=== root@dereel (/dev/pts/1) /mnt 5 -> l System\ Volume\ Information/
total 1
-rwxr-xr-x 1 root wheel 76 5 Apr 02:34 IndexerVolumeGuid
-rwxr-xr-x 1 root wheel 12 5 Apr 02:34 WPSettings.dat
They're binary, of course, and there's nothing obvious in them:
=== root@dereel (/dev/pts/1) /mnt/System Volume Information 7 -> hd IndexerVolumeGuid
00000000 7b 00 36 00 41 00 34 00 42 00 36 00 34 00 30 00 |{.6.A.4.B.6.4.0.|
00000010 30 00 2d 00 45 00 42 00 45 00 34 00 2d 00 34 00 |0.-.E.B.E.4.-.4.|
00000020 37 00 44 00 41 00 2d 00 41 00 39 00 38 00 34 00 |7.D.A.-.A.9.8.4.|
00000030 2d 00 31 00 42 00 46 00 46 00 46 00 45 00 38 00 |-.1.B.F.F.F.E.8.|
00000040 30 00 41 00 34 00 32 00 39 00 7d 00 |0.A.4.2.9.}.|
0000004c=== root@dereel (/dev/pts/1) /mnt/System Volume Information 8 -> hd WPSettings.dat
00000000 0c 00 00 00 3c 44 a5 f8 98 8e a9 ed |....<D.ø...í|
0000000c
Presumably some non-existent update instructions tell me how to update them, but for now it's simpler to rename AZ\ 2020 to AZ\ 2021. And that did the trick.
Now, where are my favourite locations? Added a new one and found recently modified files with find: /AZ\ 2020/SAVE/PROFILES/01/user.upoi. OK, .upoi it is. The newest file is:
-r-xr-xr-x 1 root lemis 45770 25 Sep 2013 /home/src/GPS/ScanFast-7020-saved/iGo8/save/user.upoi
Oh. It seems that I forgot to back up the old GPS navigator before I gave it away. I wonder if I can get it back. Or maybe I should just find a way to build one with sane tools.
In passing, it's interesting to note the ballast that these distributions carry with them. In both versions the top level directory contains a file CRASH.TXT or crash.txt. An overview of the (identical) contents:
Current System time: "2017-07-17 07:40:49"
Build version: 9.6.13.405512
Build configuration: Product--ARCHFAM
OS version: 6.0
Uptime: 0 h 0 m 6 sec 0 msec
Memory Stats:
FreeMemory: 80 MB, 224 KB, 0 B
TotalUseableMemory: 89 MB, 284 KB, 0 B
MaxMemory: 129 MB, 0 KB, 0 B
ReservedMemory: 1 MB, 768 KB, 0 B
ActualCacheSize: 3 MB, 20 KB, 959 B
GarbageSize: 3 MB, 20 KB, 959 B
There are a total of 5 crash entries spread over half an hour in early morning four years ago. I wonder what other corpses are floating around in there.
| Friday, 28 May 2021 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Chicken gone bad
|
Topic: food and drink, opinion | Link here |
A few days ago Yvonne had some fresh chicken breast left over from a dish. Can I use it? Of course, it'll fit in my breakfast somewhere.
But there was quite a bit, so I only used half of it on Wednesday. Today I prepared the other half. And it had gone bad!
That's not that surprising, given that she had bought the chicken 9 days ago, and that she had used the rest 6 days ago. But on the other hand, it is surprising, since that happens to us so seldom.
|
Creating POIs for GPS navigator
|
Topic: technology, opinion | Link here |
So it seems that I have lost the POI (Point of Interest) file that I had in my old GPS navigator. That's not so bad: it was quite a mess, and it would be interesting to find out how to create one with sane tools. What does it look like?
=== root@dereel (/dev/pts/4) /mnt/AZ 2020/save 14 -> less profiles/01/user.upoi-orig
"profiles/01/user.upoi-orig" may be a binary file. See it anyway?
ÿþ2^@|^@@^@F^@a^@v^@o^@u^@r^@i^@t^@e^@s^@|^@H^@o^@m^@e^@|^@|^@-^@3^@7^@.^@8^@0^@0^@2^@0^@1^@|^@1^@4=== root@dereel (/dev/pts/4) /mnt/AZ 2020/save 15 -> hd profiles/01/user.upoi-orig
00000000 ff fe 32 00 7c 00 40 00 46 00 61 00 76 00 6f 00 |ÿþ2.|.@.F.a.v.o.|
00000010 75 00 72 00 69 00 74 00 65 00 73 00 7c 00 48 00 |u.r.i.t.e.s.|.H.|
00000020 6f 00 6d 00 65 00 7c 00 7c 00 2d 00 33 00 37 00 |o.m.e.|.|.-.3.7.|
00000030 2e 00 38 00 30 00 30 00 32 00 30 00 31 00 7c 00 |..8.0.0.2.0.1.|.|
00
What a mess! What kind of file is that?
=== root@dereel (/dev/pts/4) /mnt/AZ 2020/save 15 -> file profiles/01/user.upoi-orig
profiles/01/user.upoi-orig: Little-endian UTF-16 Unicode text, with CRLF line terminators
Oh. Can Emacs handle it? Yes:
2|@Favourites|Home||-37.800201|143.751655|_a**||_avi||3352|Dereel|Stones Rd|29||||143.7516804,-37.8003838
@My POI
OK, much of that looks straightforward. How about copying some of my old POIs? My last saved one (8 years old) looks like this:
1|Aall mine.Foodstuffs|Casa iberica||-37.798047|144.975922|_a**||_avi||3065|Fitzroy, VIC|Johnston Street|25||+61-3-|
2|Aall mine.Foodstuffs|Mediterranean Wholesalers||-37.766798|144.962170|_a**||_avi||3056|Brunswick, VIC|Sydney Road|482||+61-3-9380-4777|
Clearly the same format. Put it in instead? The navigator doesn't want to know. OK, the second field in the new navigator is @Favourites, while in the old one it's Aall mine.Foodstuffs. And the last line (@My POI) is missing. We can fix that. But it still doesn't want to know.
I'm sure that this can be fixed, but it will take playing around. Why should I bother? At the very least I would need to find a good way to output UTF-16 text.
|
Where's my phone?
|
Topic: technology, general, opinion | Link here |
The phone that I bought on Tuesday stood a fighting chance of arriving today. But it didn't, so it'll be at least another 3 days.
And then in the evening I got a message: the phone has been sent, tracking number attached. Tracking number shows that Australia Post issued the tracking number at 16:06 today. And that's all. No evidence that it has been sent, though Australia Post is very lax with that kind of detail. Still, no thanks to oz_trade for their handling speed, despite their promise of fast delivery. Like this they run the danger of not delivering within eBay's generous delivery time frame.
| Saturday, 29 May 2021 | Dereel | Images for 29 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Corymbia ficifolia in flower
|
Topic: gardening, opinion | Link here |
It was only a week ago that I took my late autumn garden photos. But since then a surprise has happened:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The Corymbia ficifolia that suffered so badly from last winter has come back into flower, long after its flowering season. Hopefully it will weather this winter better and flower normally from now on.
|
Bibimbap again?
|
Topic: food and drink, opinion | Link here |
My interpretation of bibimbap for breakfast today:
|
|||||||||||||||||||||||
I wonder why. It's a ridiculous amount of work, it's not very authentic, and it doesn't even taste very exciting.
|
GPS POIs: success
|
Topic: technology, opinion | Link here |
Why did my GPS navigator not want to know my manipulated user.poi yesterday? Is it something basic, or some trick it has up its sleeve to refuse files it didn't create? Try one step at a time. The file it created, and which it recognizes, has only the one line:
1|@Favourites|Casa iberica||-37.798047|144.975922|_a**||_avi||3065|Fitzroy, VIC|Johnston Street|25||+61-3-|
2|@Favourites|Home||-37.800201|143.751655|_a**||_avi||3352|Dereel|Stones Rd|29||+61-3-|
@My POI
It's worth parsing this line. Clearly the delimiter is a |, and some of the fields are always empty (and have been on all the other versions of these files that I have seen over the years). Others are unidentifiable, like the _a** in every single entry. Still, let's try:
At the end, this version has a line
@My POI
I haven't seen anything like that on any of the others. Also, invisible in Emacs, the file starts with the bytes ff fe:
00000000: fffe 3100 7c00 4000 4600 6100 7600 6f00 ..1.|.@.F.a.v.o.
OK, fake that with another entry:
1|@Favourites|Casa iberica||-37.798047|144.975922|_a**||_avi||3065|Fitzroy, VIC|Johnston Street|25||+61-3-|
And how about that, it worked. Now to think how to convert a relatively simple address list into an entry that I can put in this file.
After going through all that, I discovered that I had been there before ten years ago. And yes, the long description doesn't work, because the navigator I had at the time truncated it almost completely.
|
Pedro and Piccola
|
Topic: animals, opinion | Link here |
Piccola is gradually coming to terms with Pedro. It would be much easier if he wasn't so rough with her, but she's getting there:
|
|||||||||||||||||||||||
One thing that's still a problem is house training. In the evening Yvonne took him outside to lighten himself, with the usual exhortations „Mach Pipi“ and „Mach Kacki“. No Kacki. Well, not until about 10 minutes later, when he was back inside.
It's been nearly 2 months since we got him. When will he ever learn?
|
“Walking the dogs” photos: success
|
Topic: photography, opinion | Link here |
I bought the Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3 super zoom lens mainly to take photos of the dogs, notably Pedro at a distance. But my first photos were less than satisfactory.
That has changed:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But I don't really know why. I had thought that it might be slow shutter speed or depth of field, but The second one was taken at 1/160 s at f/6.1, while this one (last month) was taken at 1/250 at f/6.3:
|
|
||||||||||
Admittedly that's a crop, but the original is also clearly unsharp.
Nearly 60 years taking photos, and I still keep being surprised.
| Sunday, 30 May 2021 | Dereel | Images for 30 May 2021 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More bread experiments
|
Topic: food and drink, opinion | Link here |
At the beginning of the month I experimented with using more rye in my bread. On that occasion I ran out of rye and didn't quite reach my intended quantities. Now I have started a new 25 kg sack of rye, so I have plenty, and I made it the way that I had originally intended last time.
But the dough was ridiculously stiff. Like last time, I had taken a total of 935 ml of water instead of the normal 975 ml to make up for the higher proportion of rye. Could it be due to the new sack of rye? Hardly, since if anything it would be moister. Added another 30 ml and it was still stiff, but usable.
And once again the bread rose very quickly, in less than 3 hours:
|
|||||||||||||||||||||||
Why the fast rising? Today it could have been because my water was hotter, 55° instead of the more typical 45°. But we'll see about that next time.
It wasn't until later that I checked. Fool! Rye needs more water, not less. But the bread seems OK. I'll aim for 1005 ml total next time.
|
Lens correction parameters
|
Topic: photography, opinion | Link here |
Question on Quora today: How does camera firmware obtain lens distortion correction parameters?.
Well, the answer is straightforward enough: the maker measures the distortion characteristics and provides them to the firmware in coded form. But how? You don't want to have to upgrade your firmware every time a lens comes out. So logically it should be in the lens.
Out to try with my oldest Micro Four Thirds system camera, the Olympus E-PM1 and my new Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3 super zoom. Here one image processed in three different ways. First raw, no correction. Then in-camera JPEG, then a JPEG created from the raw image with DxO PhotoLab (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
So clearly the camera has found out how to correct some distortion. It's still not perfect. In passing it's interesting to note the difference in angle of view between the in-camera JPEG on the one hand and the raw image and the DxO result on the other. At first I thought that the JPEG had been cropped, but the original image is the same in all cases. That must imply a considerable remapping. I wonder if other cameras do the same thing.
|
Still more Bratwurst experiments
|
Topic: food and drink, opinion | Link here |
I still haven't found a reliable way to cook Bratwurst in collagen casings without having the casings split. My last attempt was to first cook them in a microwave oven and then fry them. It wasn't successful. But why fry them? What's an air fryer for? That's a good question, but maybe this could be a good application.
This time round we had sausages with gut casings, so they wouldn't split. But I first need to know the cooking times. OK, get the hair dryer and put in the raw sausages. RTFM. It wasn't encouraging:
Food Temp (°C) Time (min) Recommended accessories Sausages 180—200 6—8 Pan and high rack
And that was all. No mention of the type or size of sausages. Both temperature and in particular time seem to be on the low side. OK, select 180° and watch the time. Here they are at the start and after 12 minutes:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Still not browned! OK, increase the heat, first to 200°, then to the maximum of 230°. After 23 minutes they looked like this:
|
|||||||||||||||||||||||
Still not brown enough, but they were starting to look dry, so I took them out. And yes, they were a little dry. Next time I'll start at 230° and see what happens.
| Monday, 31 May 2021 | Dereel | Images for 31 May 2021 |
| Top of page | ||
| previous day |
|
More breakfast experiments
|
Topic: food and drink, opinion | Link here |
Fake Phat Thai for breakfast again today. In principle, Phat Thai should have egg in it, but I haven't had much success in doing a flat fried egg the way most recipes want it. But how about doing it in a microwave oven? Say 2 minutes at 400 W? Here before and after:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What's that uncooked hole in the middle? This was the oven that bored a hole in my butter last year:
|
|||||||||||||||||||||||
So I'd expect it to have cooked there first. OK, offset the plate and try again. Bang! Something popped. Why now? If it happens, I'd expect it to happen at the beginning. Never mind, put it in the Panasonic oven for another minute. Much better, so I could cut it into strips:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I deliberately didn't add any oil, so it stuck in places, but not badly enough to be a problem. The final dish:
|
|||||||||||||||||||||||
In summary, yes, the microwave oven method works. 3 minutes at 400 W next time.
|
Another PV system battery calibration
|
Topic: technology, Stones Road house, general, opinion | Link here |
Another PV battery calibration today, once again with minor impact. It started at 10:13 with 35% battery charge, and was done by 11:17. And this time it did use the PV power, even to the point of feeding in to the grid during that time. Looking at the graphs, I'd guess that it only cost me 1 or 2 kWh from the grid:
|
|||||||||||||||||||||||
While I was at it, discovered that the graph generator does show output power usage, obfuscated at “Inverter Active Power” (the baby vomit coloured trace). Combined with the grid power usage, that's quite interesting.
|
First narcissus of autumn
|
Topic: gardening | Link here |
Seen in front of the house today:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That must be some kind of narcissus. Not pretty, but very early for a spring-flowering bulb.
|
JPEG image size again
|
Topic: photography, opinion | Link here |
While doing something else, tried the combination of Olympus OM-D E-M1 Mark II and Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3 to confirm my expectations that the JPEGs have a considerably narrower angle of view than the raw images. Yes, of course.
Now why? Surely this isn't a general issue. Could it be related to the lens, or even the focal length. I should really do some more precise measurements, but I couldn't be bothered.
|
Shaving Pedro
|
Topic: animals | Link here |
Pedro is the first non-purebred dog that we have had. When we got him, he looked like a whippet, but his mother is an Australian staghound:
|
||||||||||||||||||||||||
And now some of her hairiness is manifesting itself in Pedro:
|
|||||||||||||||||||||||
Yvonne wanted to shave off the excess hair, using my hair clippers, but I wasn't able to get any video of the event.
|
Upgrading my systems, yet again
|
Topic: technology, opinion | Link here |
The computers I was expecting last weekend didn't quite make it to the state border before the lockdown was declared. Now it looks as if they won't make it here until September.
What should I do? My current thinking is to make dereel (or whatever I'll end up calling it) my personal workstation, and leave eureka as the house server. The server software is not as time-critical as the workstation software (such as firefox, Google Chrome and other obscenities).
OK, first bring dereel up to FreeBSD 13. For that, I need to use git, a program that I find profoundly irritating. But it gave me a chance to follow the instructions in the committer's guide.
OK, that worked, at about half the speed of subversion. And when I was done, I found, for example:
=== grog@dereel (/dev/pts/5) /usr/src 6 -> ls -l bin/ls
-rw-r--r-- 1 grog wheel 4,913 31 May 16:25 cmp.c
-rw-r--r-- 1 grog wheel 2,970 31 May 16:25 extern.h
-rw-r--r-- 1 grog wheel 21,220 31 May 16:25 ls.1
-rw-r--r-- 1 grog wheel 25,498 31 May 16:25 ls.c
All the file modification timestamps have been set to “now”! Why did they do that? Apart from the fact that make uses timestamps to decide what to build, I use them to see what has changed.
And what about when you're editing the file? Once source files contained informative strings like
__FBSDID("$FreeBSD: stable/10/bin/ls/cmp.c 242807 2012-11-08 23:45:19Z grog $");
That includes the revision number (monotonically increasing) and the date and time of the modification, as well as who did it. But now it's just
__FBSDID("$FreeBSD$");
Unless there's some magic that I don't see, that's now completely superfluous. The only version number and date in there is
static char sccsid[] = "@(#)cmp.c 8.1 (Berkeley) 5/31/93";
A lot of good that will do. Unless there's some magic that I don't see, that's now completely superfluous.
There is a tag or ID associated with the version, but it's completely opaque. It seems that the “correct” way to find the details what were previously easily accessible is:
ESC[33mcommit 8a16b7a18f5d0b031f09832fd7752fba717e2a97ESC[m`
Author: Pedro F. Giffuni
Date: Mon Nov 20 19:49:47 2017 +0000
General further adoption of SPDX licensing ID tags.
...
bin/ls/cmp.c | 2 ESC[32m++ESC[m
1 file changed, 2 insertions(+)
ESC[33mcommit fbbd9655e5107c68e4e0146ff22b73d7350475bcESC[m
Author: Warner Losh
Date: Tue Feb 28 23:42:47 2017 +0000
...
bin/ls/cmp.c | 2 ESC[32m+ESC[mESC[31m-ESC[m
1 file changed, 1 insertion(+), 1 deletion(-)
ESC[33mcommit dfd91f79a34a3e1b51fa5ec7233fb9c99efd69bfESC[m
Author: Greg Lehey
Date: Thu Nov 8 23:45:19 2012 +0000
The tags are far too long to recall easily, and they're in apparently random order. In addition, git appears to be too stupid to know when to output control sequences, making the thing even messier than necessary.
I have never liked git, but the more I see of it, the less I like. Why did they call it git? As a successor to Subversion, “Aversion” would be more appropriate.
|
Pedro still not house-trained
|
Topic: animals, opinion | Link here |
Found another puddle in the hallway this evening. When will Pedro finally be house-trained? It's seriously getting on my nerves.
Do you have a comment about something I have written? This is a diary, not a “blog”, and there is deliberately no provision for directly adding comments. It's also not a vehicle for third-party content. But I welcome feedback and try to reply to all messages I receive. See the diary overview for more details. If you do send me a message relating to something I have written, please indicate whether you'd prefer me not to mention your name. Otherwise I'll assume that it's OK to do so.
| Top of page | Previous month | Greg's home page | Today's diary entry | Next month | Greg's photos | Copyright information |
