|
|
|
Saturday, 1 August 2020 | Dereel | Images for 1 August 2020 |
Top of page | ||
next day | ||
last day |
Strange RCS message
|
Topic: technology | Link here |
Found a typo in my diary entry for 51 years ago today. OK, not a problem, check it out and fix it.
=== grog@eureka (/dev/pts/16) ~/public_html 198 -> co -l diary-aug1969.php
co: RCS file RCS/diary-aug1969.php,v is in use
Huh? What's that? I've been using RCS for 25 years now, but this is the first time that I've seen that message. Off to the web to look for answers, nothing really good, but it seems that there should be a lock file name enclosed in commas, or maybe starting with an underscore, somewhere, and that this file should be removed.
=== grog@eureka (/dev/pts/16) ~/public_html 199 -> locate diary-aug1969.php
/home/grog/public_html/diary-aug1969.php
/home/src/Sysconfig/eureka/MasterRCS/home/grog/public_html/RCS/diary-aug1969.php,v
Nothing there. Of course, locate is historical. By default it's rebuilt once a week, but I do it every night. So the missing file name could mean that it happened today. OK, look in likely places:
=== grog@eureka (/dev/pts/16) ~/public_html 200 -> l _*
-rw------- 1 grog lemis 98,304 1 Aug 15:36 _1cKgnvW=== grog@eureka (/dev/pts/16) ~/public_html 201 -> l RCS/,*
-r--r--r-- 1 grog lemis 163,840 1 Aug 15:36 RCS/,diary-aug1969.php,=== grog@eureka (/dev/pts/16) ~/public_html 202 ->
I didn't check the first file, though in retrospect it's interesting that it had roughly the same time stamp as the one I was looking for. But it seems that something went wrong with checkout, and that it only half got done. Removed the file and all was well.
A day for photography
|
Topic: photography, general, opinion | Link here |
Yvonne is using her PIXIO “Robot Cameraman” more and more. Today that was particularly obvious. As usually on a Saturday, I had also taken my house photos, and since the sun was shining, I did it round solar noon, the same time as I took my analemma photos.
The result was three different tripods in front of the house:
|
From left to right, the tripod for the house photos, with the Olympus OM-D E-M1 Mark II and the M.Zuiko Digital ED 8mm f/1.8 Fisheye PRO, the tripod for the horse videos with the PIXIO, OM-D E-M5 Mark III and M.Zuiko Digital 14-42 mm f/3.5-5.6 II R, and the E-30 with the Zuiko Digital ED 9-18 mm f/4.0-5.6.
Sony HDR-CX405 results
|
Topic: multimedia, technology, opinion | Link here |
Finally found time to look at the video clips that I took on Chris Bahlo's Sony HDR-CX405 yesterday. First, though, how do I get the videos off the camera? As on the Olympus cameras, all images are stored in a single directory (MP_ROOT\100ANV01). And that needs to start this silly PlayMemories program—not to use it; it's useless—but to set the camera USB connection into mass storage mode.
Did that: no clips! Searched the whole file system and found nothing obvious. But they had been displayed in “Windows Explorer” before I started PlayMemories. OK, stop it again, disconnect and reconnect the camera, and take a look:
|
This time they're displayed in “subdirectories” with the date, a completely different organization from in the real file system. Why do they not store them that way in the first place?
OK, how do I get this onto a sane machine? “Copy” is what I need, but that's an old, worn-out magic word. I can't use CMD.EXE, because this thing isn't a file system. Off to search the web, and came up with this page.
TL;DR: there are four methods: drag and drop, cut and paste, use context menu or (the easiest) CMD.EXE. But somehow the first three didn't work for me: when I selected the files and tried to drag them, they were automatically deselected. But even if it had worked, what a way to work!
In the end, since I wanted the whole directory, went to the parent directory and copied that. Only 30 minutes for something that in the old days would have taken 10 seconds.
And what did they look like? Terrible! Yes, now the resolution is better, but it's still interlaced, and it shoves the fact into your face:
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
These are slightly cropped, but immediately obvious when looking at the video. And yes, of course I can deinterlace them, but that's associated with a significant loss of quality, not what we need.
But why? I was sure I had set 50p. Still more head-scratching needed.
Other items of interest: the image gradation somehow seemed worse than the Olympus, though it's difficult to pin it down. But the image stabilization was better than I expected, probably on a par with the Olympus. At the very least I'd have to do some more rigorous comparisons.
VoIP recording software?
|
Topic: technology, multimedia, language, opinion | Link here |
Yesterday's fracas with Telstra left me wishing that I had had an audio transcript of the call. Surely that can't be difficult: in principle I could have sniffed the traffic with Wireshark, but that's a bit clunky. Aren't there any recording programs out there?
Yes, but it seems—paradoxically—mainly for Microsoft “Windows”. And there's an Open Source Audio Recorder offering a free 30-day trial. rms would love that and point out that “open source” is far from as obvious as “free software”.
Interpreting COVID-19 lockdown rules
|
Topic: health, general, opinion | Link here |
Not surprisingly, there were many reports of people disobeying the current lockdown rules. Or did they just not understand them?
To quote two such cases, as reported by Victoria's Police Minister Lisa Neville:
Can I be really clear, just in case there is any doubt at all, that there is absolutely no reason or need to drive from Melbourne to Wodonga to have a Big Mac.That is one of the fines that was issued yesterday.
Well, yes, but do the regulations prohibit it? Probably: my understanding is that it is prohibited, because Wodonga is (a long way) outside the Melbourne exclusion zone. And you really can't say that the BigMac isn't available within the zone.
To be fair, I'm not sure that there is an exclusion zone. I certainly can't point to any regulation mentioning it. I've gone and checked again, but the more this thing progresses, the less sure I am of anything about these restrictions. They're getting more vague and confusing by the day.
But she goes on:
There is absolutely no reason to drive from Werribee out to Springvale to buy groceries.Why would people think that is OK? They don't care, it can't be because people aren't aware of the rules.
I disagree. There is a good reason to go to Springvale to buy groceries: you can get things there that you can't get elsewhere, notably East Asian foodstuffs. I have even driven there from here (170 km, coincidentally past Werribee) for exactly that reason.
And why not? Both Werribee and Springvale are inside the exclusion zone. Yes, they're nearly 60 km apart, but I don't see anything in the rules that says that you have to buy within a certain distance of your home.
So, people: don't just complain. Write down sensible rules and make them clearly available. If you don't want people going long distances for groceries, set a limit. I can't see one at the moment.
Changing hand bandage
|
Topic: health, opinion | Link here |
I've had the same bandage on my left hand since Wednesday. It hasn't been causing any trouble, and I've had less pain from the hand than from the tetanus injection, but it's clearly time to change the bandage. It's not looking overly pretty:
![]() |
|
![]() |
|
Somehow it doesn't cover the wound overly well. But there's more:
![]() |
|
![]() |
|
The second photo shows how off-centre the bandage was. But how did the sutures bend like that? They should have been straight. And they no longer cover the wound properly. In fact, it looks as if one of them has disappeared altogether, though I thought that they were thinner. Something to do with the position of the wound?
Put a new bandage on, hopefully better centred, and despite the position it seems to stick. But we're not completely over with the bleeding:
![]() |
|
Cordon bleu revisited
|
Topic: food and drink, opinion | Link here |
We've cooked Cordon bleu a number of times recently, most recently in an “air fryer”. I decided that that was a mistake: fry in butter.
But how long? Tonight I gave it about 12 minutes, which seems reasonable. It wasn't until later that I discovered I had written 4-5 minutes per side in the recipe, so clearly it depends on the temperature. But being crumbed, it swallows a lot of butter. Next time I should weigh it.
Sunday, 2 August 2020 | Dereel | Images for 2 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Still more Sony HDR-CX405 strangeness
|
Topic: multimedia, technology, opinion | Link here |
Why did the the Sony HDR-CX405 switch to interlaced mode after I set it for 1080p the other day? Off to check, revealing even more strangeness.
Here's the Image Quality / Size menu:
![]() |
|
Of these, only Frame Rate and Image Size have obvious meanings. Well, maybe. In fact, Image Size refers only to still images. So how do you select the image size?
I still don't know. Last week I selected it via REC Mode, but I discovered that the format I had selected was only interlaced. Yes, you can select 50p from the Frame Rate menu:
![]() |
|
It accepts it, but when you exit the menu it sets it back to interlaced without any further indication.
Spent some time trying to understand the File Format menu:
![]() |
|
More buzzwords. Yes, I can look them up in Wikipedia, but apart from that there's only guesswork and experimentation. So that's what I did, and I discovered:
What a mess! OK, I don't really care about the other REC Modes, which are even worse quality, and I hope that Chris doesn't either.
Off later to try out these settings with the horses. And the results? No time to look at them. Mañana.
Disaster!
|
Topic: health, general, opinion | Link here |
Despite the measures taken, the number of COVID-19 cases in Victoria continues to rise. On Thursday the government imposed “lockdown”, whatever that may mean, and I grumbled about the lack of precision in their statements and the lack of information on the web. My comments on the web site went without answer, as I had expected. And yesterday I questioned the illegality of going shopping in Springvale from Werribee: where are the distance limits?
But today we have something new. “Stage 4” (of course also not clearly defined) restrictions, including grocery shopping no more than 5 km away. That's reasonable, at least for Melbourne; round here the nearest supermarket is 28 km away. But what was it before? My guess is that they hadn't defined it, but expected people to adhere to some concept of locality anyway.
Round here, we're back to having “no visitors in the house”. Ah, but that's the way it was before, sort of. Changes to restrictions that apply to Greater Geelong and surrounds (not a title that would normally attract my attention; Geelong is 75 km away, compared to 30 to Ballarat) applies to me anyway, and it states:
Apart from the strange idea that going to cafes is safer than staying at home, does the exception include allowing the cleaners in? My interpretation is “yes”. And that page is still there, as of 3 August 2020, 13:20 (time zone defaults to UTC+10). But then there's this page, published on 2 August at 14:38, titled “Melbourne placed under stage 4 coronavirus lockdown, stage 3 for rest of Victoria as state of disaster declared”, which contains:
Regional Victoria will move to stage three restrictions from Thursday, meaning restaurants, cafes, bars and gyms will close from 11:59pm on Wednesday.
And what else does that mean? They don't say.
Off to the health.vic.gov.au site to see if they had improved anything. No, but I noticed the ability to submit feedback. So, at 15:55, I sent:
Please, PLEASE finally publish exact details of the lockdown restrictions. I've just read the latest premier's statement, which refers to a previous statement, which refers to some nebulous "phase 2" and "phase 3" lockdowns, which I can't find anywhere.
How do you expect people to adhere to the restrictions if they don't know what they are? Even now, it seems, you're allowed to go grocery shopping. But yesterday somebody was fined for doing just that, simply because he/she drove 60 km to Springvale to do so. I've been looking for some time to find whether there is a distance limit on where to shop for food. There's a good reason to choose Springvale: foodstuffs (notably East Asian) are available there that aren't available elsewhere. So if you don't want people to go that far, make it clear.
My suggestions:
1. Write a clear, concise web page describing what people may and may not do, and KEEP IT UP TO DATE.
2. Link to this page from http://health.vic.gov.au and other appropriate pages.
3. Link to this page from ANY statement published about lockdown restrictions.
This will not only make life easier for the general population. It will give a better legal framework for prosecuting offenders.
And I got a reply fairly quickly, at 07:28:33 +0000 (their choice of time zone). I had to fill out a form, of course, and so they got information that they quoted, like this:
Submitted on Sun, 08/02/2020 - 15:55
Submitted by: Anonymous
Submitted values are:
Email (required)
groggyhimself@lemis.com<mailto:groggyhimself@lemis.com>
Name (optional)
Greg Lehey
Postcode (optional)
3352
At least they had an (unspecified) Australian time zone there. But there's my name, so why “anonymous”?
The reply was not helpful:
We are working as quickly as possible to get the relevant information up - please see the new restrictions page to see what is now allowed - https://www.dhhs.vic.gov.au/updates/coronavirus-covid-19/premiers-statement-changes-melbournes-restrictions
Now isn't that helpful! As the URL indicates, it's not what I was asking for. In fact, it's what I was complaining about. And then they say:
You shouldn't be travelling more than 5km from your home unless you have certain needs.
Which needs? Where's that in the link they sent? Does it even apply to me? No, according to what information I have found: it applies only to Melbourne, and they have my post code, so they should have known that.
Discussed the matter on IRC, with a surprising lack of understanding on the part of the others. They pointed to further URLs with snippets of information: https://www.abc.net.au/news/story-streams/coronavirus/, https://www.abc.net.au/news/2020-08-02/victorias-latest-coronavirus-restrictions-explained/12516182, https://www.dhhs.vic.gov.au/victorias-restriction-levels-covid-19, https://www.vic.gov.au/coronavirus-covid-19-restrictions-victoria...
Why is it so hard just to keep an up-to-date summary and point to it from everywhere that mentions the issue? Disaster indeed, but one that could be easily fixed.
I won't even start on formulation issues like “You can not” instead of “You may not”, or “Coronavirus” instead of COVID-19.
Spraying weeds
|
Topic: gardening | Link here |
Nice weather today, so out to finally spray some of the weeds that are popping up. Managed the bed to the south of house and the “garden” to the north.
Monday, 3 August 2020 | Dereel | Images for 3 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Present from Piccola
|
Topic: general, animals | Link here |
Piccola has been tending to put on weight lately, so we've put her on a diet. That has done her good, less so the mice that have infested the garage. But today, after her frugal breakfast and subsequent walk outside, Yvonne found:
![]() |
|
![]() |
|
![]() |
|
A present? A thrown gauntlet? Not even Nikolai wanted to eat it.
Too many tropaeolums
|
Topic: general, animals, opinion | Link here |
Spring is just around the corner, and new plants are coming up. In front of the music room we have an issue: the whole bed is covered in tropaeolums (tropaeola?). But other plants are struggling to get out:
![]() |
|
![]() |
|
![]() |
|
Tore some of them out, stepped into the bed and... fell into the hole for the water extraction pump, up to my knee. Fool! I knew that it was there, and that I had left the lid off to observe it better. That's it in the front of the first photo. Time to replace it and nurse my bruises.
Sony HDR-CX405, the last
|
Topic: multimedia, technology, opinion | Link here |
Finally got round to looking at the video clips that I took with the Sony HDR-CX405 yesterday.
First: where are the files? Well, this time there was only one. Copied it from the fake directory 2020-08-02 to eureka, where it showed itself as:
-r-xr--r-- 1 grog wheel 736,886,784 3 Aug 12:33 20200802134529.MTS
That file name looks like a date YYYYMMDDHHMMSS. But it doesn't match the timestamp of the file. Presumably Microsoft's “copy” function is too polite to maintain the original timestamp. Further checking with PlayMemories showed that yes, indeed, it's the timestamp relating to the recording time (presumably the end of the recording), which could be useful in some bizarre circumstances.
While searching, also found the name that it uses in USB mode. With the fake MTP file system, it's 20200802134529.MTS, as above. But internally, as the USB mode reveals, it's \PRIVATE\AVCHD\BDMV\STREAM\00013.MTS! Why PRIVATE? If there's one thing you want to be able to find on the device, it's the videos you've just taken!
While I was searching, accidentally clicked on the image, started a program that was too polite to state its name, and took forever to display. Isn't Microsoft fun?
So now I had the clip that I was looking for. As might be expected, it was better than the previous ones. But somehow it lacked the crispness of the Olympus OM-D E-M5 Mark III videos taken at the same time. It also showed pronounced flare: Yvonne had placed the camera so that it pointed towards the sun. Yes, it was in the shade of the trees, but enough sunlight got through. And it made itself more unpleasantly visible than with the Olympus (last image):
![]() |
|
![]() |
|
![]() |
|
And yes, this particular Olympus shot doesn't look good. But that's not typical.
And what do these buzzwords mean? Found this page, really about the Sony A7, which confuses me all the more:
AVCHD looks similar to XAVC S, and if you just want to upload the video straight away it might be okay. There are three disadvantages to AVCHD:
Poor audio codec Harder to edit Doesn’t grade very wellThere are two ‘advantages’:
It’s half the data rate, therefore half the storage. You don’t need an SDXC card for it, you can use SDHC cards (32GB and lower).
Clearly this doesn't relate to the HDR-CX405, but it's not clear why the data rates are the other way round on this camera.
Olympus edge conditions
|
Topic: photography, technology, opinion | Link here |
I've had my “new” Olympus E-30 for just over a year (in fact, the beginning of August has been a time for buying cameras over the decades). I only use it for my analemma photography series, and for the fun of it I didn't delete the images from the CF card, just to see how long it would last. It's a 4 GB card, and the pair of images I take amount to about 25 MB per pair, so I'd expect it to last for about 160 pairs.
Today I was nearly there: only 30 MB left on the card. Took the first image. 10 MB. “Card full”! Why?
...
58021542 ORF 13142830 2020-08-02 12:25
58031543 ORF 10993858 2020-08-03 12:25
337 files 3 973 199 392 bytes
Total files listed:
341 files 3 973 199 392 bytes
19 595 264 bytes free
My best guesses are that either the camera first writes the image to the card and then compresses it, or that it checks free space based on a worst-case scenario. In any case, my fault, missed the image pair. And my guess of 160 pairs (320 files) was pretty close to the mark.
Time to finish dereel
|
Topic: technology, opinion | Link here |
Somehow I've had lots of things to do lately, and one of the multitude of things I have left unfinished was the installation of dereel, my second desktop machine.
But time has made it more urgent: for reasons I don't understand, a number of sites no longer want to talk to the firefox version on eureka. Time to continue with installing dereel.
In principle, the next step is simple: install multi-monitor X on dereel and join it to eureka with x2x. I've been using X on the desktop for over 30 years, and I published my first article on how to configure it over 27 years ago. What's the problem?
Two issues: hardware and software. Where do I connect the outputs from the card (an old Nvidia GeForce GT 710 (or GK208; for some reason they all have two names)?) It has three outputs D-Sub, DVI and HDMI, but I only have one spare monitor. Never mind, I have monitors with multiple inputs, and coincidentally the two leftmost of the existing 4 have a spare DVI and HDMI input. The one I've been using for testing can thus have the D-Sub connector. The only issue was that that's the output I had connected to distress; I'll have to add a card with DVI output to distress, just because of that.
The other issue was the cables. Do I have the right kind and the right length? Once again, somewhat to my surprise, yes. Connected them up, just for the fun typed startx and... it worked! Out of the box. And it ran on all three monitors.
Well, almost. I ended up with a single display across all three screens, and of course the chances of them being in the right order were 5:1 against, but it worked. And when I inserted a dongle for a wireless mouse, that worked too. That's so different from the pain I've been through in recent years.
So now I have 5 monitors again, for the first time in 13 years:
|
From left to right:
Monitor | Resolution | Input | System | |||
Acer G246HL | 1920x1080 | D-Sub | dereel:0.0 | |||
DVI | distress (to be completed) | |||||
HDMI | not connected | |||||
BenQ GW2265 | 1920x1080 | D-Sub | eureka:0.0 | |||
DVI | dereel:0.1 | |||||
BenQ E2200HD | 1920x1080 | DVI | eureka:0.1 | |||
HDMI | dereel:0.2 | |||||
D-Sub | not connected | |||||
FRT DIGITAL | 2560x1440 | DVI | eureka:0.2 | |||
BenQ G2400W | 1920x1200 | DVI | dischord | |||
HDMI | eureka:0.3 | |||||
D-Sub | not connected | |||||
So not only do I have four machines connected, I have three inputs free.
But things aren't as good as they might seem. eureka has 4 display outputs and dereel has 3. That means that I could connect another two monitors and use them all—if I could find a use for them. In the meantime, I suppose I should start a new server on eureka talking just to the rightmost two monitors. And switching the inputs is a different pain for each monitor. The BenQ GW2265 in particular reverts to DVI when the D-Sub input goes away (screen saving in particular). It doesn't revert when only the D-Sub input returns.
Still, more progress than I had expected. At the very least it shows that the card can really drive three monitors. Now to tidy up the configurations for both systems.
Covered riding arena: progress?
|
Topic: Stones Road house, general | Link here |
It's been several days since I called the Golden Plains Council and left a message with Rob[iy]n about updating our planning permit. No reply. Tried again today and was connected to Sandra, who suspects that it's too late to update it, but will see what she can do, and will call back by Wednesday at the latest.
Tuesday, 4 August 2020 | Dereel | Images for 4 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Snow!
|
Topic: general, photography, opinion | Link here |
Winter may be on its way out, but it's not gone yet. When I got up this morning it was snowing! That's a very seldom occurrence. But how do I capture it as an image (or “on film”, as we used to say)? This photo shows it clearly:
![]() |
|
... except that it's hardly visible. Only when I enlarge it can anything be seen:
![]() |
|
![]() |
|
Clearly I can't discern individual flakes at that distance. But it's not immediately obvious that it's snow. How about a different shutter speed? That one was at 1/100 s. Faster (1/200 s, first image)? Slower (1/20 s)?
![]() |
|
![]() |
|
Neither looks convincing. And I don't get much chance to repeat: it stopped snowing almost as soon as I took those photos.
Spring under way
|
Topic: gardening, general | Link here |
Despite the snow, the spring flowers are on their way. Last week we only had snowdrops, but now the daffodils are out in force:
![]() |
|
![]() |
|
Where's my measuring jug?
|
Topic: gardening, general | Link here |
Filling up the dish washer, I missed my 500 ml measuring jug (the one I bought on 4 October 1967 as part of my darkroom equipment). Where was it? Ah, I had used it a couple of days ago to measure weed killer. Did I forget it outside? Yes! In fact, it was in one of the photos of the daffodils above:
![]() |
|
![]() |
|
And it was still full! I had carefully washed out the empty container of weed killer and forgotten to put the measured quantity into the sprayer! All that work for nothing, and now I have to wait for another sunny, windless day!
Entrance pupil locations revisited
|
Topic: photography, opinion | Link here |
I've been taking panoramic photos for nearly 12 years now, and it's been some time since I have learnt all the details, including the positioning requirements, which mainly relate to keeping the entrance pupil in the same place. Eight years ago I wrote a page with tables of positions for my lenses, based mainly on information available on the web.
So I consulted it when I wanted to take this panorama of my desktop:
|
Oh horror! Not a single Micro Four Thirds system lens in the table! How times change! Yes, I've measured the values for my Olympus M.Zuiko Digital ED 8mm f/1.8 Fisheye PRO, though I hadn't added them to the table, but that was the only one.
What to do? I was going to use the M.Zuiko Digital ED 12-100 mm f/4.0 IS PRO, but it was probably too long anyway. In the end, I fell back to the Zuiko Digital ED 9-18 mm f/4.0-5.6.
Why is no entrance pupil information available for newer lenses? Do people no longer care?
What's that noise?
|
Topic: general, opinion | Link here |
In the evening, we had just finished watching a film on TV when we heard a noise reminiscent of a washing machine spinning. But there was none. Off to find out where the noise came from. In front of the house! The extraction pump was running continuously, sucking in water, not having enough to fill the chamber, and stopping again. Turned it off. Tomorrow is early enough to find out what happened, presumably as a result of me falling into the hole yesterday. Hopefully it's not damaged.
Wednesday, 5 August 2020 | Dereel | Images for 5 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Hand wound
|
Topic: health, general | Link here |
It's been a week since I cut my hand trying to open inappropriate packaging. Time to change the bandage again:
![]() |
|
![]() |
|
Took a shower before putting a new cover bandage on it, after which the suture showed its advertised peeling away at the edges:
![]() |
|
Things look much better now, but I'll leave the bandage on until the suture detaches by itself.
Pump issues
|
Topic: Stones Road house, general | Link here |
Out to take a look at the extraction pump this morning:
![]() |
|
![]() |
|
The float has lodged on the frame and was thus not able to fall low enough to turn the pump off. I'll have to tidy the whole thing out and test it, once we have more water. At any rate enough water had gathered in the sump overnight that I could pump it out, apparently showing that the pump has not been (seriously) damaged.
COVID-19: fines for non-compliance
|
Topic: health, general, opinion | Link here |
The state government has imposed stricter rules for non-compliance with COVID-19 restrictions. If you're positive and don't abide by the self-quarantine requirements, you can be fined $1,652. And if you're caught a second time, you can be fined $4,957. This is what alerted the news media, who report it typically as “about $3,500”, politely omitting their reference to US currency or to the second time. To quote the New York Times:
Nearly everyone traveling for a job deemed essential must carry a signed work permit. And after a campaign to check in on people with Covid-19 found that one in four was not at home, a fine equivalent to about $3,500 was established for future violators. Some of the new rules have led to widespread confusion, with unclear guidance for everything from dog-walking to seeing romantic partners.
Good! It should be higher. In fact, it can be. Those are on-the-spot fines (you do have that much cash on you, right?). The $1,652 can be extended to $10,000 in court for repeat offenders, and the $4,957 can be extended to $20,000 for repeat offenders (isn't that what the $4,957 is for)? Still not enough considering the damage it can do; $50,000 wouldn't be too high.
But why these strange sums for the on-the-spot fines? They're in penalty units, of (currently) $165.22, so the first one is 10 units, and the second one is 30 units. Does that make sense? I can't decide, but at least I have an explanation.
Thursday, 6 August 2020 | Dereel | Images for 6 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Accident in Grassy Gully Road
|
Topic: general, opinion | Link here |
While walking the dogs today, noted several skid marks. First, heading west:
![]() |
|
![]() |
|
![]() |
|
![]() |
|
What could cause that? There was nothing else on the road, but clearly the driver had left the road and hit a tree.
But there was another skid mark in the opposite direction (going east):
![]() |
|
This one was mainly of interest because of all the litter left behind. But it would seem that two cars came along too fast, and to avoid collision they left the road and collided with vegetation instead.
Only: if that's what happened, they left to the right. Why? We drive on the left round here.
Measuring salt
|
Topic: food and drink, general, opinion | Link here |
The US Americans have a mania for measuring foodstuffs by the volume of some not clearly defined household device that wasn't designed for this purpose, notably spoons and cups. I've ranted about this at length. There are three basic issues:
Today I found a question on Quora: How many ounces are in a teaspoon of salt?. OK, why not measure it? The result: between 0.19 and 0.56 (avoirdupois) ounce.
In detail. Three different kinds of salt:
![]() |
|
Take two different teaspoons, measure 10 “spoonfuls” of each kind of salt, and weigh them.
![]() |
|
Here the average weight per spoonful:
Salt kind | Spoon 1 | Spoon 2 | ||
Rock salt | 7.55 g | 11.36 g | ||
Kitchen salt | 7.51 g | 10.87 g | ||
Table salt | 7.84 g | 9.98 g | ||
All these were done with naturally heaped teaspoons. Interestingly, and contrary to my expectations, the kind of salt didn't make a big difference to the weight: it's within experimental error (table salt was the heaviest from spoon 1, the lightest from spoon 2).
I also tried as level as possible a spoon (spoon 1) of table salt and got 5.35 g.
In the past I have tried to measure the volume of teaspoon 2 (level) and come up with a volume of about 3.5 ml. In the USA, a "teaspoon" is defined as 4.93 ml (customary) or 5 ml (legal, § 101.9 paragraph (b)(5)(viii)), in other countries only 5 ml, all larger than any teaspoon I have measured.
The difference between definitely too little salt and definitely too much salt is probably about 30%. Here we have a difference of 300%. That's a perfect way to ruin a dish. And why?
Steak and kidney pie again
|
Topic: food and drink, opinion | Link here |
We're still not home with correct preparation of steak and kidney pie. Last time I baked at the bottom of the oven for 20 minutes, followed by 5 minutes with the top element. The main issue there was to cook the dough under the filling, but I hadn't wanted to cook in two different ways.
Today I tried a different tack: cook the dough underneath by itself before filling, this time for 5 minutes at 180° in the “coffee machine” “air fryer”:
![]() |
|
Is that enough? It wasn't too much, anyway.
Then fill the pies and cook for 20 minutes at 210° with the top element in the middle of the oven. After 10 minutes they looked like this:
|
And after 20 minutes:
|
![]() |
|
![]() |
|
OK? Somehow the browning seemed too uneven. But we're getting there.
Friday, 7 August 2020 | Dereel | Images for 7 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
COVID-19: Too close for comfort
|
Topic: health, general, opinion | Link here |
We take protection against COVID-19 seriously:
![]() |
|
![]() |
|
We've been pretty well spared any proximity to COVID-19 so far, but a couple of reports in the Ballarat Courier today were unsettling. One referred (without links, as seems to be their wont) to an infected staff member at Woolworths in Sebastopol yesterday. Yvonne had been shopping there on Wednesday, the day before. With a bit of searching found this page, which links to content to which I can't refer directly, and which will probably soon disappear. But today I read:
We have been notified that a team member at our Sebastopol store has tested positive for the Coronavirus (COVID-19).
We’re making contact with our Sebastopol team members and will provide our full support to those required to self-isolate in line with advice from the health authorities.
As a food retailer, we already have very high standards of cleaning and hygiene in place. As an extra precautionary measure, the store will closed overnight for an additional deep clean.
While the risk of transmission to customers and team members is low, the safety and wellbeing of the local community is our priority.
Customers and team members should be assured they can continue to safely shop and work at our Sebastopol supermarket.
Any customers who shopped in our Sebastopol store on Thursday 23 July and feel unwell in the next two weeks should make contact with the Department of Health and Human Services (DHHS).
What's the danger? Pretty low. Yvonne was wearing a KN95 mask, she is unlikely to have come near the person in question, and she had been there the day before. But somehow the suggestion to do nothing until you feel sick is superfluous: if you develop symptoms, you need to be tested whether or not you were in Woolworths on Thursday. Somehow it's not overly reassuring.
And then there was a second incident that we heard only indirectly: an employee of Sinclair Meats in Delacombe tested positive this morning. Unlike Woolworths, they have closed their operation completely until further notice.
Yvonne was there too on Wednesday:
![]() |
|
Again, no particular cause for alarm, apart from the fact that things are coming so close.
Spring on its way
|
Topic: gardening | Link here |
Another sign of the coming spring:
![]() |
|
There's nothing that surprising about an Arum blooming at this time of year, except that it's one of the row that we transplanted in April.
A new gardener?
|
Topic: gardening, opinion | Link here |
While walking the dogs today, found a gardener working on Lorraine Carranza's property. Anton's (or Tony's?) mowing, phone 0411 029 736. He seems to specialize in lawn mowing and whipper snipping, but he might make a good alternative if Mick continues to be unable to come.
More corroded spoons
|
Topic: food and drink | Link here |
Last month I noted some strange corrosion on an old spoon. It's not the only one. Here another case:
![]() |
|
![]() |
|
That looks less like corrosion than mechanical damage. But how?
X configuration on dereel
|
Topic: technology, opinion | Link here |
It's been several days since I've looked at the X configuration on dereel. There are at least two reasons: firstly, I've been busy, and secondly I expect pain.
Back to look at it again today. I had started editing a configuration file, but it looked strange. Here an example:
Section "Screen"
Identifier "Screen2"
Device "Device1"
Monitor "Monitor1"
DefaultDepth 24
Option "TwinView" "0"
Option "TwinViewXineramaInfoOrder" "DFP-0"
Option "metamodes" "nvidia-auto-select +0+0"
SubSection "Display"
Depth 24
EndSubSection
EndSection
Why the TwinView entries? They're for displays spanning multiple screens. In this case I want three individual displays, one per monitor. OK, I've been there before, so why not check what I did last time?
It proved that I didn't record the thought processes that led to these decisions. And Nvidia have considerately removed some of the pages to which I referred. But in particular I didn't explain anywhere why I had included the "TwinView" "0" entry.
OK, start from scratch, do a Google search, which took me to this page (is this even from Nvidia?), which told me part of what I needed to do: simply duplicate the Screen entries, which all look like this:
Section "Screen"
Identifier "Screen1"
Device "Device1"
Monitor "Monitor1"
DefaultDepth 24
SubSection "Display"
Depth 24
EndSubSection
EndSection
The only difference for the three entries was the number at the end of the first three lines.
And how about that, it worked! And I was able to connect to it from eureka with:
DISPLAY=:0.3 x2x -east -to dereel:0
Well, almost. Of course the configuration couldn't know anything about the physical layout of the three monitors, and I arrived in the middle one. Going “east” from eureka:0.3 took me first to the middle one, then to the westmost one, then to the eastmost one.
Not a problem, we just tell the config what they really look like. That's what the Monitor entries are for. Well, that's actually what the Monitor section is for. The first entry looks like this:
Section "Monitor"
# HorizSync source: edid, VertRefresh source: edid
Identifier "Monitor0"
VendorName "Unknown"
ModelName "Acer G246HL"
HorizSync 30.0 - 94.0
VertRefresh 50.0 - 76.0
Option "DPMS"
EndSection
And that, of course, is the leftmost (westmost) monitor. But it seems that ModelName is merely for documentation purposes, and X assigns this number (Monitor0) to the second monitor and Monitor1 to the first. OK, there are various ways of fixing it. The obvious one is to swap the numbers in the Screen entries:
Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor1"
DefaultDepth 24
SubSection "Display"
Depth 24
EndSubSection
EndSection
Section "Screen"
Identifier "Screen1"
Device "Device1"
Monitor "Monitor0"
DefaultDepth 24
SubSection "Display"
Depth 24
EndSubSection
EndSection
OK, try that, restart X. No difference! That must be a bug.
But there's another option: right at the beginning of the config file is a description of the server layout:
Section "ServerLayout"
Identifier "Layout0"
Screen 0 "Screen0" 0 0
Screen 1 "Screen1" RightOf "Screen0"
Screen 2 "Screen2" RightOf "Screen1"
InputDevice "Keyboard0" "CoreKeyboard"
InputDevice "Mouse0" "CorePointer"
Option "Xinerama" "0"
EndSection
Just a little change to the config should do it:
Screen 0 "Screen1" 0 0
Screen 1 "Screen0" RightOf "Screen1"
Screen 2 "Screen2" RightOf "Screen0"
But again it made no difference! And just to be sure, yes, the layout is in the log file:
[351814.689] (==) ServerLayout "Layout0"
[351814.689] (**) |-->Screen "Screen1" (0)
[351814.689] (**) | |-->Monitor "Monitor1"
[351814.689] (**) | |-->Device "Device1"
[351814.689] (**) |-->Screen "Screen0" (1)
[351814.689] (**) | |-->Monitor "Monitor0"
[351814.690] (**) | |-->Device "Device0"
[351814.690] (**) |-->Screen "Screen2" (2)
[351814.690] (**) | |-->Monitor "Monitor2"
[351814.690] (**) | |-->Device "Device2"
[351814.690] (**) |-->Input Device "Keyboard0"
[351814.690] (**) |-->Input Device "Mouse0"
Clearly something's severely broken here. I found a way to work around it last time, but that was 7½ years ago. Maybe that's the background for the TwinView entries. I'm not in much hurry right now, so it seems a good idea to analyse what's really wrong before working around it Yet Again.
Saturday, 8 August 2020 | Dereel | Images for 8 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
X installation continued
|
Topic: technology, photography, opinion | Link here |
Into the office this morning to find that my BenQ E2200 HD monitor (variously eureka:0.1 via DVI and dereel:0.2 via HDMI) was not responding. dereel's X was on screen blank, and though eureka was delivering a signal, there was no way to get the monitor to display a menu so that I could switch to DVI. I had to wake dereel's X server to do the switch. It's time to wonder whether I should connect one or two more monitors now that I can.
OK, now that I have an X server, I can run Hugin on dereel (Hugin is too polite to run across the network). First, though, I needed to install PHP, after which my pto generation worked.
Oh. Somehow it doesn't know that the lens I was using was a fisheye. What a mess! Another thing to postpone until I find where this information is stored.
Bloody Microsoft!
|
Topic: technology, opinion | Link here |
When I fired up distress, I was distressed to find this display on the screen:
![]() |
|
I don't want to meet no steenking Edge! Don't you tell me what to do, you foul Microsoft! And what's this insult?
|
My security? My privacy? You have violated my machine and rebooted it! And now you have only given me one option: “Get started”. Get stuffed! Fired up the Task Manager and shot it down in flames. I wonder if I can remove it completely.
Porting revisited
|
Topic: technology, opinion | Link here |
It's been nearly 25 years since O'Reilly published my first book, “Porting UNIX Software”
Time have changed, reflected by the fact that O'Reilly now seem to have taken the book out of their collection without telling me. That might also be a sign of the times, though it looks more like web site breakage; I'm still there, and there are broken links to my books. But I can't even find their book catalogue from their home page. Have they gone out of that business?
Today I discovered that, while my x2x worked, it was maxing out a CPU core. I know this problem: it checked all file descriptors on return from select(), and since the number of file descriptors has increased dramatically, the time taken has done so too.
OK, the problem has been fixed in a new version. Just install that.
Sounds easy, doesn't it? But the system on eureka is 5 years out of date, and the Ports Collection doesn't want to know about it. OK, it's a tiny program. Unpack the tarball on another system, then compile it the old-fashioned way on eureka.
Oh. No Makefile, just a Makefile.am. That rings a bell, but how does it work? Something to do with autotools? The file INSTALL in the top level directory is helpful: “Just build/install x2x as any other autotools-based program.”. OK, there's a file bootstrap.sh containing the single line autoreconf -isv. That worked, and I was able to run configure, but trying to build things had it hang on a strange command:
echo 'char *lawyerese =' > lawyerese.c.tmp && sed -e 's|.*|"\0\\n"|g' >> lawyerese.c.tmp && echo ";" >> lawyerese.c.tmp && mv lawyerese.c.tmp lawyerese.c || rm -f lawyerese.c.tmp
It's (ultimately) clear why it hangs: there's no input for the sed command, so it reads from stdin.
More messing around. Built it on dereel, which gave me the commands and the singularly useful file lawyerese.c. Here the entirety apart from multiple repetitive lines:
char *lawyerese =
"0\n"
... 28 repeats omitted
"0\n"
;
And then I could build, almost:
cc -DHAVE_CONFIG_H -I. -I/usr/local/include -D_THREAD_SAFE -pthread -O2 -pipe -fstack-protector-strong -fno-strict-aliasing -MT x2x.o -MD -MP -MF .deps/x2x.Tpo -c -o x2x.o x2x.c
cc: error: unknown argument: '-fstack-protector-strong'
*** Error code 1
-fstack-protector-strong? What's that? It must be some new-fangled optimization option. Get rid of it! And that worked, and finally I had my x2x, and it no longer loops. But it brings back a taste of the Bad Old Days. I wonder which is worse, that or the pain with the Ports Collection.
Flying ducks and pollen
|
Topic: animals, gardening | Link here |
Seen while walking the dogs:
![]() |
|
Those are Australian Wood Ducks. The one at top right must be 10 m off the ground. I know that they fly, but I've never seen one in a tree like that before.
And then there's this stuff:
![]() |
|
The yellowish tinge round the puddle appears to be some kind of pollen, but what? For some reason I think of Wattle, but though it's pretty widespread, it's not clear where it could have come from.
Sunday, 9 August 2020 | Dereel | Images for 9 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
A day in the kitchen
|
Topic: food and drink, opinion | Link here |
Somehow a whole lot of things have ganged up on me in the kitchen: I had planned four different dishes for the week, and we had three different kinds of noodles left over that would need cooking before they went mouldy.
For one of the packages of noodles, though, the Tak On Singapore Chow Mein, it came too late: they were already mouldy, after only three weeks. Damn! I cooked and froze large quantities of “Wife's noodles”, but didn't get round to the Shanxi noodles.
Kimchi revisited
|
Topic: food and drink | Link here |
I used to make large quantities of kimchi, which I ate at lunch on most days. But then we stopped eating lunch, so I stopped making kimchi.
But now I've found a few Korean recipes for brunch, so it's on the menu again. Time to revisit the recipe. In the past I had had difficulty to find daikon, the long radish used for kimchi, so I used spring onions instead. Now, though, it's easy enough to find daikon. And I had found that I was running a little low on the marinade, which I had planned at 400 ml for 2.4 kg. The result, I suspect, is that the lack of coverage on top of the cake allowed it to go mouldy. So more: today the total (cabbage and radish) was 3 kg, so I tried 600 ml.
And it still wasn't enough! At the end I still didn't have enough left over, so I mixed 7.5 g of gochugaru with 100 ml of water and spread it on top. Still not enough! Next time I'll plan 600 ml for 2.4 kg, fully 50% more than previously.
Just for the fun of it, I substituted garlic shoots for garlic: I bought some recently, and I don't know what to do with them. They're nearly as strong as garlic, so I put 75 g in instead of the normal 50 g of garlic. We'll see how it comes out.
Dollee rendang
|
Topic: food and drink, opinion | Link here |
I have had half a package of rendang paste from Dollee in the fridge for some time now. Time to use it.
The instructions are simple: mix with beef or chicken, fry for 10 minutes, add water and cook for 10 minutes, add coconut milk, bring back to the boil and serve—with rice or bread!
Rendang is a slow-cooked dish, and 20 minutes is barely enough to cook beef! So I added the coconut milk early and simmered slowly for about 2 hours. At least it looks right now; we'll see how it comes out.
Pollo en nogada?
|
Topic: food and drink, opinion | Link here |
We have lots of nuts in the fridge. A whole shelf is full of them! Time for some dishes with nuts. How about pollo en nogada? I have recipes in two different books, so close that you'd suspect some close relationship: chile ancho, pecan nuts, boiled chicken (done after preparing the sauce). There must be others.
Off to look. The most authentic are probably in Spanish: receta pollo en nogada.
And sure enough, there were plenty. But very different (Spanish rather than Mexican?). Milk! No chilis! But peanuts instead of almonds; that last could be plausible, since almonds don't come from America.
The most important recognition, though, was that this is a dish for leftovers: thus the chicken cooked separately. I can do better than that. Put the chicken in a freezer for another day.
Jägerschnitzel again
|
Topic: food and drink | Link here |
Another dish, eaten tonight, was Jägerschnitzel, mainly to get rid of some of the ridiculous amount of dried mushrooms that we have in the pantry. This time we ate only the dried mushrooms (Boletus edulis, labelled “porcini”, corresponding to German Steinpilz. And at the last minute Yvonne decided that we had some capsicum left over, so what we finally ended up with was somewhere between a „Jägerschnitzel“ and a „Zigeunerschnitzel“:
![]() |
|
![]() |
|
It's OK, but somehow nothing special.
Monday, 10 August 2020 | Dereel | Images for 10 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Finishing the cooking
|
Topic: food and drink | Link here |
Yesterday was spent in the kitchen, but I didn't quite finish: cook the Shanxi “planed” noodles. They come out very sloppy:
![]() |
|
Somehow they remind me of my fresh noodles. Maybe I can try that some time; these were not cheap.
Analemma accuracy
|
Topic: photography, general, opinion | Link here |
I've been taking photos of the sun at mean solar midday (12:25:00 AEST) for nearly a year now. Time to think more about assembling them all.
My method: place the Olympus E-30 with the Zuiko Digital ED 9-18 mm f/4.0-5.6 and a 10 EV neutral filter in front of a post in the garden. Take one photo at 1/4000 s, f/5.6 exactly at 12:25:00 (13:25:00 in the summer), and then another at 2 s at f/5.6 for the background:
![]() |
|
![]() |
|
The purpose of the second photo is to locate the sun accurately: with programs like Hugin I can first align the series of second photos and then swap the image to have correct alignment of the sun images. I haven't decided yet what I do after that, but it's gradually becoming time.
But how accurate are the images? How accurate must the time be? Today I tried a comparison: take photos exactly at 12:25:00, 12:26:00 and 12:28:00, and compare them. Since they're on a tripod, I can just take a crop of the same part of each image (run the cursor over an image to compare it with its neighbour):
![]() |
|
![]() |
|
![]() |
|
So there is a difference: it takes about 2 minutes for the sun to travel its own diameter. Probably a second or two inaccuracy wouldn't be that bad. And yes, the times reported by the camera are inaccurate. It clearly doesn't believe in sub-minute accuracy. It must have been programmed by a mobile phone developer.
Dead kangaroo
|
Topic: animals | Link here |
On our daily dog walk today, just outside our driveway:
![]() |
|
![]() |
|
![]() |
|
Another dead kangaroo! This one appears to have been hit by a truck, breaking its legs. But who cut off an ear? I'm reminded of Tom Lehrer.
![]() |
|
![]() |
|
Weed spraying for real
|
Topic: gardening | Link here |
Another sunny day today without too much wind. Back to finish the spraying I had attempted last week. This time I noted the foaming when filling the sprayer, as well as when spraying. Hopefully I won't forget again.
Cockatoos
|
Topic: animals | Link here |
We frequently have large swarms of sulphur-crested cockatoos flying over the house, but it's relatively seldom that they hang around. Today was such a day:
![]() |
|
![]() |
|
![]() |
|
New masks for Yvonne
|
Topic: health, general | Link here |
Yvonne has a new, reusable face mask, sent by Yana:
![]() |
|
![]() |
|
I'm still trying to work out what it means.
Tuesday, 11 August 2020 | Dereel | |
Top of page | ||
previous day | ||
next day | ||
last day |
More animals in the household
|
Topic: animals, general | Link here |
Living in the country, mice are inevitable. We had our first mouse in Stones Road just after we moved in, and it's clear that Piccola enjoys being in certain places in the hope of catching one.
But lately we've found more and more in the garage. Here's what it looks like two nights after last sweeping the floor:
|
![]() |
|
![]() |
|
Where do they come from? What are they eating? It seems to be a combination of dog food and the old pink sponge cleaners that we have had for decades, and which I thought were of no use:
![]() |
|
![]() |
|
I wonder why they like them. And how do they get at the dog food? It's sealed in the blue and purple containers to the left of the middle of the first photo, and there's no evidence of the mice getting in.
Where do they hide? An obvious place would be under the bottom boards in the shelves that I installed recently:
![]() |
|
The bottom of the other shelf unit (in the middle of the first photo) has nothing on it, so I lifted the board to see. Lots of droppings, but no mice. Where are they? They must be dozens of them. Do they come in at night and leave in the day? Time for Piccola to spend some nights there, and of course also for some mousetraps.
Sarawak laksa!
|
Topic: food and drink, opinion | Link here |
I've taken the hint from the current COVID-19 pandemic and planned to do more shopping online. An obvious choice is East Asian food, so today I went looking for likely suppliers and wrote a suppliers page for them.
Some look quite good, and I've found nearly all the ingredients I normally buy in Belmont, Geelong, and also some that I haven't seen. But one thing surprised me was a recipe for (apparently) Sarawak laksa. Clearly they're talking about Kuching laksa; I doubt they would have tried it in Sibu or Miri. It's surprising for a number of reasons:
The image in the recipe looks nothing like my recollection of the dish.
Wikipedia agrees. Of Sarawak laksa, it currently writes:
It has a soup-base of sambal belacan, which contributes to its crimson colouring, adding both coconut milk and sour tamarind,...
It seems to have been a favourite of Anthony Bourdain, apparently a famous chef de cuisine who died a couple of years ago. There are many hits for his choice of Sarawak laksa.
My own opinion was less positive. As I wrote 51 years ago,
... after having a bowl of laksa in Carpenter St. - this stuff is completely different from what one gets in KL, or West Malaysia in general, and I think I prefer the West Malaysian stuff.
Still, it's something to investigate.
Wednesday, 12 August 2020 | Dereel | Images for 12 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Fighting dereel X configuration
|
Topic: technology, opinion | Link here |
I've left another couple of days before continuing with the X configuration on dereel. I had feared that it would be painful, and my fears proved to be underestimates.
In principle it works, but the three monitors are configured in the wrong sequence. The one in the middle (BenQ GW2265) comes up as screen 0 and the one on the left (Acer G246HL) comes up as screen 1. By some coincidence, the one on the right (BenQ E200HD) comes up as screen 2:
![]() |
|
In passing, there's another issue that I'm going to have to deal with. This photo was taken to show the contents of the monitors, but from my sitting position what I see is:
![]() |
|
How am I going to work around those reflections?
Back to the problem at hand: nothing I can do in the configuration file can change the sequence! I've already discussed the obvious way, the ServerLayout at the top of the screen. But there are other ways too.
It's worth stepping back and thinking about the structure of the tree. There are three objects: Device (display card), Screen and Monitor. Screen is abstract: it's an addressable part of the overall display area controlled by the server, and it can span one or more monitors. In my case, I want one screen per monitor, though on the server eureka:1 there's only one screen spanning all four monitors. The hierarchy is as mentioned: device at the top controlling the screens, screens controlling the monitors. Traditionally the objects are named after they type with a trailing number, starting at 0: Device0, Device1, Screen0, Screen1, Monitor0 etc.
In this case, we have one device, three monitors and three screens. How do you map? It seems that you have three device entries, all the same entry except for the device number. Here's one of them:
Section "Device"
Identifier "Device0"
Driver "nvidia"
VendorName "NVIDIA Corporation"
BoardName "GeForce GT 710"
BusID "PCI:4:0:0"
Screen 0
EndSection
It's worth noting that there's no way to tell the server which port of the display card to use. You might expect an entry like:
Port "HDMI"
But that doesn't exist. There may be other ways, part of today's investigations.
The Screen entries are similar to the Device entries:
Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor0"
DefaultDepth 24
SubSection "Display"
Depth 24
EndSubSection
EndSection
The monitor entries contain, inter alia:
Section "Monitor"
Identifier "Monitor0"
VendorName "Unknown"
ModelName "Acer G246HL"
EndSection
So, everything's clear, right? Device 0 controls screen 0 and monitor 0, and monitor 0 is the Acer G246HL. The ServerLayout makes it clear that it's on the left. But when I start the server, monitor 1 (the BenQ GW2265) comes up as screen 0!
Clearly something's broken. Am I using the wrong config file? No, /var/log/Xorg.0.log contains:
[190255.485] (++) Using config file: "/etc/X11/xorg-0.conf"
But I got it working years ago with eureka. Took a look at eureka's config file and came to the conclusion that I had messed around until, by some coincidence, I had got it to work. I'm in no hurry this time (one of the things I found very important about setting things up this way).
So: it's clear that this is a bug. X is not honouring ServerLayout. If I change the layout specification, nothing happens, so it's not just a case of mistaken identity. What other ways do I have to influence the mapping?
The first I tried was to change the mapping of the devices to screens, here device 0 to screen 1:
Section "Device"
Identifier "Device0"
Driver "nvidia"
VendorName "NVIDIA Corporation"
BoardName "GeForce GT 710"
BusID "PCI:4:0:0"
Screen 1
EndSection
But that made no difference. OK, screen to device?
Section "Screen"
Identifier "Screen0"
Device "Device1"
Monitor "Monitor1"
DefaultDepth 24
SubSection "Display"
Also no change. It didn't occur to me until much later that I should have made both these changes at once to remain consistent.
OK, devices? Change the first Device entry to contain:
Section "Device"
Identifier "Device1"
Screen 1
EndSection
Still no change.
Change the mapping between screen and monitor?
Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor1"
EndSection
Still no change.
Did the TwinView entries in eureka's config contain the clue? They refer to names like CRT-0 and DFP-1. What do they mean? Off googling for TwinViewXineramaInfoOrder, which at least has no false positives, and came up this page, which describes the option in some detail. The interesting part is:
The option string is a comma-separated list of display device names. The display device names can either be general (e.g, "CRT", which identifies all CRTs), or specific (e.g., "CRT-1", which identifies a particular CRT). Not all display devices need to be identified in the option string; display devices that are not listed will be implicitly appended to the end of the list, in their default order. Examples:
"DFP"
"TV, DFP"
"DFP-1, DFP-0, TV, CRT"
In the first example, any enabled DFPs would be reported first (any enabled CRTs or TVs would be reported afterwards). In the second example, any enabled TVs would be reported first, then any enabled DFPs (any enabled CRTs would be reported last). In the last example, if DFP-1 were enabled, it would be reported first, then DFP-0, then any enabled TVs, and then any enabled CRTs; finally, any other enabled DFPs would be reported.
Default: "CRT, DFP, TV"
Now isn't that helpful? I still don't know what a DFP is, but clearly a CRT is an old, worn-out magic word. Still, worth trying the entries I had in eureka:
--- xorg-0.conf 2020/08/07 23:52:19 1.1
+++ xorg-0.conf 2020/08/12 06:34:08
@@ -102,7 +102,9 @@
Section "Screen"
Identifier "Screen0"
Device "Device0"
- Monitor "Monitor1"
+ Monitor "Monitor0"
+ Option "TwinView" "0"
+ Option "TwinViewXineramaInfoOrder" "CRT-0"
DefaultDepth 24
SubSection "Display"
Depth 24
@@ -112,7 +114,9 @@
Section "Screen"
Identifier "Screen1"
Device "Device1"
- Monitor "Monitor0"
+ Monitor "Monitor1"
+ Option "TwinView" "0"
+ Option "TwinViewXineramaInfoOrder" "CRT-1"
DefaultDepth 24
SubSection "Display"
Depth 24
@@ -123,6 +127,8 @@
Identifier "Screen2"
Device "Device2"
Monitor "Monitor2"
+ Option "TwinView" "0"
+ Option "TwinViewXineramaInfoOrder" "CRT-2"
DefaultDepth 24
SubSection "Display"
Depth 24
No change, of course. But clearly everything here refers to TwinView, a proprietary way to join two monitors into a single screen, so there is no reason to believe that it makes any difference, especially since on eureka some screens have it and others don't.
And that was about as far as I got with frobbing the config file. But maybe that's an old way to do things. Who reads config files any more? Not even X, it seems, or it looks on them as comments. Doesn't Nvidia have a program to do this? Yes, nvidia-settings, which offers this view of the server layout (click to enlarge):
![]() |
|
![]() |
|
At least that's a correct representation of the server layout. Where did it get the monitor names? From the config file or the EDID? My guess is that it was the config file. But the config file says that this is screen 1, not screen 0. OK, try dragging. Yes, it works, but what a mess! It doesn't snap into place, and some of the comments suggest that I'll leave unaddressable space between monitors, or have overlapping space.
![]() |
|
OK, I can just select “X screen 0” at the bottom:
![]() |
|
![]() |
|
But that led to still more surprises. In particular, it forgot the name of the 2200HD, it moved the GW2265 to the extreme right, leaving a space, and also forgot about screen 2:
![]() |
|
![]() |
|
OK, create a new screen 2 in the right place:
![]() |
|
Save the config? Sure, but don't overwrite the old one. In the past I have found that config files generated by this utility were severely broken: they omitted all comments, changed the indentation, and maybe more. But that was enough for today. I'll look at them in the morning.
In passing, it's interesting to note this detail in nvidia-settings:
|
There they really are, CRT, flat screen and TV. Could it be that DFP is an abbreviation of
“Digital Flat Panel”? That sounds particularly archaic. I get the feeling that the whole
Nvidia software offering contains a lot of dead wood and ignored worked-around
bugs.
And apart from Nvidia, a couple of other things I hadn't expected. I had to install RCS as a port! And somehow I still can't start X from an xterm, despite following all the instructions to work around this “security” measure. But it seems that I had discovered the real problem months ago. Another thing to check on tomorrow.
COVID-19: Seen before?
|
Topic: health, opinion | Link here |
There have been some musings going on lately that COVID-19 may not have originated in or near Wuhan, as currently supposed, but that it was a mutation of an older virus; it seems that many people seem to have some immunity in their T cells.
It'll be interesting to see if anything comes of this idea, but it had occurred to me already: last year I got a severe cold or chest infection, the like of which I have never experienced. That's not typical for me: I never get things like that, and I wondered at the time what had caused it.
Thursday, 13 August 2020 | Dereel | Images for 13 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
A better mousetrap
|
Topic: animals, general, opinion | Link here |
Build a better mousetrap, and the world will beat a path to your door. What an archaic statement! It certainly doesn't apply to computing equipment any more, particularly phones.
But I was reminded of it when Yvonne brought home some mousetraps for the garage yesterday. I'm used to primitive ones made of wood. These are made of plastic and have a dip into which to put the bait:
![]() |
|
That makes sense. The mouse has to go to more trouble to extricate it, making it more likely to trigger the trap. And sure enough,
![]() |
|
But it's not all as positive as it seems. The first photo is of the other mousetrap after a mouse had extricated the bait (a piece of bacon fat) and eaten it.
OK, tonight I'll fill the well with lard. The mice can't pull that out.
Next, of course, offer it to Piccola:
![]() |
|
![]() |
|
It took a while for her to get the idea:
![]() |
|
![]() |
|
Too long for my liking. I put them both in the laundry and left them to fight it out. When I returned, the mouse was gone.
X on dereel again
|
Topic: technology, opinion | Link here |
Took a look at the X configuration files that I generated with nvidia-settings yesterday. To my surprise, the formatting wasn't that bad, but they had still lost the comments. But there were a number of surprises:
The file did have a different ServerLayout section:
Section "ServerLayout"
Identifier "Layout0"
Screen 0 "Screen0" 0 0
Screen 1 "Screen1" 3840 0
...
EndSection
I've never seen a Screen entry with other than 0 0 after the name. Clearly these are absolute coordinates. But look at them! Screen 1 is 3840 pixels to the right of screen 0, and there's no screen 2. How can that work? The monitors are only 1920 pixels wide.
The Screen sections contain things I've never seen before:
Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor0"
DefaultDepth 24
Option "Stereo" "0"
Option "metamodes" "VGA-0: nvidia-auto-select +0+0, DVI-D-0: nvidia-auto-select +1920+0"
Option "SLI" "Off"
Option "MultiGPU" "Off"
Option "BaseMosaic" "off"
SubSection "Display"
Depth 24
EndSubSection
EndSection
Section "Screen"
Identifier "Screen1"
Device "Device1"
Monitor "Monitor1"
DefaultDepth 24
Option "Stereo" "0"
Option "metamodes" "HDMI-0: nvidia-auto-select +0+0 {AllowGSYNC=Off}"
Option "SLI" "Off"
Option "MultiGPU" "Off"
Option "BaseMosaic" "off"
SubSection "Display"
Depth 24
EndSubSection
EndSection
One thing is very obvious: the metamodes option. It looks to be quite close to what I was looking for yesterday, a Port entry. But there's other stuff in there as well, and I should investigate the option carefully.
And clearly the layout isn't what I asked for. nvidia-settings displayed this layout when I saved the file:
![]() |
|
It appears to have merged the first two monitors into a single screen 0, not what this display claims.
Still, it has given me more insight. Now I can either read on or just adjust the config entries. But I think I should first understand what the other options mean.
More dereel upgrade fun
|
Topic: technology, opinion | Link here |
While doing my X config experiments, checked in the config file with RCS. Or at least I tried. Not there! People have taken RCS out of the base system and made a port of it. OK, that's “progress”, I suppose.
Then I tried to install a couple of graphics programs. A new version of darktable has just come out. The port has already been updated, but the package hasn't. OK, I've had enough pain building ports: this sort of thing tends to have hundreds of dependencies. Install the old package to satisfy the dependencies, then build a new one.
Ha, ha, there are build depends! And you have to answer the config dialogue for Every Single One! But at least it installed. Some day I must investigate what it can do.
Ports pain: so nice, so nice, we do it twice
|
Topic: technology, multimedia, opinion | Link here |
While I was installing ports, there's handbrake, a video transcoder that Julie Lannen had recommended to Yvonne. I've heard good things about it, not for the reasons that Julie stated, but to transcode some ridiculously large video files. Typical German videos are either 720p (a 45 minute episode takes up about 1.4 GB) or 1080p (2.4 GB). But lately we've been finding episodes of Die Firma Hesselbach, also 45 minutes. The series was first aired in 1960, in black and white of course. The sizes speak for themselves:
=== grog@teevee (/dev/pts/11) /spool/Already/Series 590 -> mpid Hesselbach/01/01-11-Der-Betriebsausflug-20*4
Hesselbach/01/01-11-Der-Betriebsausflug-20170319-113000.mp4 VIDEO: [H264] 512x288 24bpp 25.000 fps 510.3 kbps (62.3 kbyte/s)
Hesselbach/01/01-11-Der-Betriebsausflug-20200805-001000.mp4 VIDEO: [H264] 1920x1080 24bpp 50.000 fps 7487.0 kbps (913.9 kbyte/s)=== grog@teevee (/dev/pts/11) /spool/Already/Series 592 -> ls -l, Hesselbach/01/01-11-Der-Betriebsausflug-20*4
-rw-r--r-- 1 grog wheel 260,868,776 14 Jul 2019 Hesselbach/01/01-11-Der-Betriebsausflug-20170319-113000.mp4
-rw-r--r-- 2 grog wheel 3,275,599,658 7 Aug 11:00 Hesselbach/01/01-11-Der-Betriebsausflug-20200805-001000.mp4
OK, they're 1080p, but they're still black and white. And why 50 fps? My guess is that this was all recorded on film at 24 or 25 fps. Yes, the 1080p version is better than the 288p version, not surprisingly, but the images are not of modern quality. It must be possible to recode them down to a quarter of that size. Thus handbrake.
OK, install the handbrake package. It installed without too much trouble. Then:
=== grog@teevee (/dev/pts/11) /spool/Already/Series 593 -> handbrake
bash: handbrake: command not found=== grog@teevee (/dev/pts/11) /spool/Already/Series 594 -> cat /usr/ports/multimedia/handbrake/pkg-plist
...
%%X11%%bin/HandBrake
...
Ah, we're studly, too studly to tell anybody that we want to be started as HandBrake.
=== grog@dereel (/dev/pts/8) /eureka/home/grog 3 -> HandBrake
ld-elf.so.1: Shared object "libmp3lame.so.0" not found, required by "HandBrake"
OK, why wasn't that dependency installed? That must be lame, at least something I know.
=== root@dereel (/dev/pts/1) /usr/ports/graphics/darktable 34 -> pkg search lame
blame-1.3.1 Display last modification for each line in an RCS file
flamegraph-g20170824_1 Stack trace visualizer
...
No lame! But what's this blame? That's part of RCS, which I just installed.
=== root@dereel (/dev/pts/1) /usr/ports/graphics/darktable 35 -> blame
su: blame: command not found
Oh. Maybe it's not part of RCS. All this messing around to install dribs and drabs.
And lame? Installed it from the port without difficulty. And discovered in the Makefile:
LICENSE= LGPL20+ PATENTS
LICENSE_COMB= multi
LICENSE_NAME_PATENTS= Intellectual Property Precaution
LICENSE_FILE_LGPL20+ = ${WRKSRC}/COPYING
LICENSE_FILE_PATENTS= ${FILESDIR}/lame_patents.txt
LICENSE_PERMS_PATENTS= dist-mirror dist-sell no-pkg-mirror no-pkg-sell auto-accept
# The reason for "no-pkg" was discussed with core:
# Core feels that the risk versus any potential benefit is not in the
# project's best interest and we should continue to not package
# audio/lame.
Is that really such an issue?
Friday, 14 August 2020 | Dereel → Ballarat → Dereel | Images for 14 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Blood tests and vaccinations
|
Topic: health, general | Link here |
Into Ballarat today to hear the results of last month's blood test. It seems that Paul's not happy with my cholesterol values and is considering putting me on statins. That's interesting considering that I first came to him four years ago to get a second opinion about my cholesterol levels, for which my doctor at the time had prescribed statins, and which at the time he found unnecessary.
Why the change of heart? Now I'm a diabetic. Now that's a good explanation. I've had diabetes for 13 years, but it has taken a while to convince Paul. OK, maybe the cholesterol values are higher, in which case it might be justified. I'd suggest a second test in 3 months and then decide. But he's the doctor, let him decide. And he decided to decide in 3 months. OK, fine by me.
Also discovered that my Vitamin D levels are too low, so ended up being prescribed large quantities to last me for the next 3 months. Also a couple of vaccinations: shingles and a new vaccination for pneumonia and related illnesses. If I recall, the last one was called Pneumovax 23, but that's an old, worn out magic potion, and this one is better.
Leave the cities!
|
Topic: health, general, opinion | Link here |
Yvonne told me that she had heard that our old house in Kleins Road is on the market again—for $779,500! We listed it 5 years ago for $420,000 and ultimately got $360,000 for it. Why the great increase in asking price? Sure, if you can get it, go for it, but is it realistic?
Looking at the other houses on offer in Dereel, yes, it might be realistic. There are two others on offer, one for $395,000, the other for $547,000, and both not very interesting.
Has he made great improvements to the property? We've seen it from the outside in passing, of course, but there are photos that show parts that we can't see. So I looked with interest at the photos.
The result: no, he has done very little with the house. Even the old antennas at the south end of the house are still there. It's also much emptier than when we lived there. About the only addition was a slow combustion heater in the kitchen:
Why that? Doesn't he know how to operate the (reverse cycle) air conditioner, one of the heads of which is also in the photo? I had considered the air conditioning in every room to be one of the big selling points for the house. Yes, it's mentioned in the exposé, but in the same breath as the dishwasher. It brings home to me how primitive house temperature control is in this country.
The other photo is the east side of the house:
He has removed nearly all the vegetation. Maybe it was a little bit over the top, but now it looks completely bare.
Still, that won't make a big difference in price. The real question is: why have the prices increased like that? An obvious reason is that it's still much cheaper than in Melbourne, and out here you're much less likely to get infected, be it with COVID-19 or something else. That probably means that our property has also appreciated considerably in value, which can't be a bad thing.
Nvidia X investigations
|
Topic: technology, opinion | Link here |
Yesterday I had put off my investigation of the X configuration in the expectation that it could take some time. I didn't have much time today, but I made a start. First, check the entry for metamodes in Nvidia's X config options page. Sure enough,
Option "MetaModes" "string"This option describes the combination of modes to use on each monitor when using TwinView. See Chapter 13, Configuring TwinView for details. Default: string is NULL.
This is going round in circles! My config file was generated by Nvidia without any reference to TwinView. But it's the only reference, so off to look at that page. The entry is quite long, but refers almost only to various resolutions for the screen. About the only exception is at the end of the entry, where I read:
Because it is often unclear which mode within a MetaMode will be used on each display device, mode descriptions within a MetaMode can be prepended with a display device name. For example:
"CRT-0: 1600x1200, DFP-0: 1024x768"
These bloody CRT-0 and DFP-0 again! But that's still nothing like what nvidia-settings generated:
Option "metamodes" "VGA-0: nvidia-auto-select +0+0, DVI-D-0: nvidia-auto-select +1920+0"Option "metamodes" "HDMI-0: nvidia-auto-select +0+0 {AllowGSYNC=Off}"
In passing, it seems that they have partially changed their terminology from CRT, DFP and TV to the more readily understandable VGA, DVI and HDMI. After all, how can you know that a CRT will be connected at the other end of the VGA connector, or a TV at the other end of the HDMI connector? In my case, all three are “DFP”, assuming I have deciphered the TLA correctly.
So what do I do now? What I had been trying to avoid: suck it and see. Mañana.
Saturday, 15 August 2020 | Dereel | Images for 15 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
dereel in earnest
|
Topic: technology, opinion | Link here |
So I've given up on finding relevant documentation for the Nvidia X driver. Take what I have and frob it with values that seem likely, even if I can't find any documentation for them. For the three screens, add:
Option "metamodes" "VGA-0: nvidia-auto-select +0+0"
Option "SLI" "Off"
Option "MultiGPU" "Off"
Option "BaseMosaic" "off"
...
Option "metamodes" "DVI-D-0: nvidia-auto-select +0+0"
...
Option "metamodes" "HDMI-0: nvidia-auto-select +0+0 {AllowGSYNC=Off}"
The other options (SLI, MultiGPU, BaseMosaic) are repeated in each Screen section.
And how about that, it worked! Score 0 for RTFM. So now I have 5 monitors with 7 screens:
|
From left to right: dereel:0.0, dereel:0.1/eureka:0.0, dereel:0.2/eureka:0.1, eureka:0.2, eureka:0.3.
The first problem is obvious: two of the monitors have two screens each. Which do I use? The intention is to use them for the dereel screens: the eureka screens are no longer needed, at least until proof of the contrary.
But that brings the second problem: how do I navigate from eureka to dereel? I can't go “west” from eureka:0.2, because that would take me to the hidden eureka:0.1. So I have to go “east” from eureka:0.3 to dereel:0.0 at the other side of the row. And that proved to be irritating. Now to get from the two screens dereel:0.2 and eureka:0.2 that are straight ahead of me and next to each other, I need to wrap through all the other screens. The next step should be to start a third server on eureka (I already have a second) with only the two rightmost screens.
But that wasn't the only thing. It looks like just a small modification—after all, I'm still sitting in the same place, looking at the same monitors—but even the slight difference is quite irritating. I need to rearrange things altogether. And somehow things still aren't completely right. Normally I can mark a text in any random window and copy it to an Emacs buffer with c-y, but now I have to paste it with the “middle” mouse button. And though I can move right through the dereel monitors, moving left jumps over one: from dereel:0.2 I go to dereel:0.1, and from dereel:0.2 I go to eureka:0.3 (thanks to x2x). What is causing that?
And then, of course, there's working out how to share files across the machines. I had originally symlinked /eureka/home/grog to /home, but that meant that I had the same directory ~/.mozilla for firefox, which it didn't like much. So: create a separate directory and symlink myriad files. Maybe it would really be just easier to create separate profiles for dereel. But I've done it now.
And then, of course, there's the usual pain with newer versions of standard software. I've now lost my Emacs key bindings for firefox, with little of obvious advantage in return—why is it so difficult to add a bindings table to this bloated behemoth? The size difference wouldn't be measurable. But even Emacs had surprises for me. I write this diary with a number of PHP macros, so the start of every day looks like this:
|
But now I end up with different highlight colours and indentation:
|
Arguably I could live with the indentation changes, but I have 423 diary files that would need to be converted, and the indentation colours clearly relate to punctuation that hadn't bothered it before. There are reasons not to upgrade to the latest and gratest.
Preparing for spring
|
Topic: gardening | Link here |
As usual, I've left things late in the garden. I started pruning the roses over 6 weeks ago and then put it off to read about the fine details. But I can't wait much longer, so today addressed a few more bushes. Somehow this is frustrating. It's not difficult, but after a couple of bushes I can't take it any more.
Also did a bit of weeding in the herb garden, which was also greatly in need thereof.
Sunday, 16 August 2020 | Dereel | Images for 16 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
More X pain
|
Topic: technology, opinion | Link here |
Why is my screen navigation on dereel so strange? I can move right through the monitors, but moving left jumps over one: from dereel:0.2 I go to dereel:0.1, and from dereel:0.2 I go to eureka:0.3. The answer, it seems, is x2x. Today, for some reason, I picked up the mouse attached to dereel, and there things work normally. What's causing that?
But at the moment that's not my issue. First I need to castrate the X configuration on on eureka. Currently it has 4 screens, two hidden behind dereel:0.1 and dereel:0.2. That's more of a pain than it appears: apart from them occasionally popping up when dereel's X server screen blanks, it means that I can't move left from eureka to dereel; instead I have to go the long way round.
OK, I'm more up to speed on Nvidia X configuration now after my fun over the last few weeks. It was relatively simple to create a config file with two monitors, each on one of the two graphics cards. And for the second, it's clear that it should be the monitor with the HDMI connection:
Option "metamodes" "HDMI: nvidia-auto-select +0+0"
Start X. Yes, it comes up—with only the first monitor! What does the log file say?
[3631312.274] (II) NVIDIA(GPU-1): NVIDIA GPU GeForce GT 640 (GK107) at PCI:5:0:0 (GPU-1)
[3631312.275] (--) NVIDIA(GPU-1): Valid display device(s) on GeForce GT 640 at PCI:5:0:0
[3631312.275] (--) NVIDIA(GPU-1): BenQ GW2265 (CRT-0) (connected)
[3631312.275] (--) NVIDIA(GPU-1): BenQ G2400W (DFP-0) (connected)
[3631312.275] (--) NVIDIA(GPU-1): BenQ E2200HD (DFP-1) (connected)
[3631312.275] (--) NVIDIA(GPU-1): BenQ GW2265 (CRT-0): 480.0 MHz maximum pixel clock
[3631312.275] (--) NVIDIA(GPU-1): BenQ G2400W (DFP-0): 165.0 MHz maximum pixel clock
[3631312.275] (--) NVIDIA(GPU-1): BenQ G2400W (DFP-0): Internal Single Link TMDS
[3631312.275] (--) NVIDIA(GPU-1): BenQ E2200HD (DFP-1): 330.0 MHz maximum pixel clock
[3631312.275] (--) NVIDIA(GPU-1): BenQ E2200HD (DFP-1): Internal Dual Link TMDS
Oh. We're back to CRT and DFP. OK, which one is it? I think it's the E2200HD. Change HDMI to DFP-1. No change, still no screen 1. Did I get the monitor name right? Where is the monitor name? Modern monitors are far too polite to display it on the frame, but there's a tiny label on the back of the monitor I want to use, written in 4 point script:
![]() |
|
I barely managed to make out the model number: ET-0009-B. That doesn't match any of the names I know. It wasn't until I took the photo that I saw that it also has a product name.
But no, nothing. There were no error messages in the log file; it didn't complain that it couldn't fulfil the second screen section. It didn't mention GPU-1 further.
So what's wrong now? Despite the verbosity of the X log, it doesn't tell me why it's not giving me what I ask for. Missing entry in the config file? How can I tell? It looks like more experimentation.
Seed germination revisited
|
Topic: gardening | Link here |
On 9 July I planted four Thai basil seeds. On 13 July I planted “sweet” basil. And on 28 July I planted chili and tomato seeds. How are they doing?
One of the Thai basil seeds (right side) germinated a week or so ago. Another may be trying. But it's been over 5 weeks, so I'll have to assume that the germination rate is less than 50%. I don't have exact figures for the “sweet” basil, but there seem to be enough plants.
|
And the tomatoes and chilis?
![]() |
|
All of the tomatoes (right side) seem to be doing well, with germination rates of about 80%. The second row of chilis from the right, the serranos, are now coming out, and there's an indication that the árbol are also on the way.
And the others? It seemed that I should have soaked them before planting. But according to the instructions, they can take up to 3 weeks to germinate anyway, so I'll wait for a while.
House maintenance issues
|
Topic: Stones Road house, general, opinion | Link here |
We've been here for over 5 years, and gradually it's starting to show. The drain from the kitchen is blocked up, and I can't find where the blockage is. We've seen this before in Kleins Road, but there there was an inspection opening outside the house. That's clearly far too old-fashioned: I couldn't find anything here. So: buy a drain cleaner? Call in a plumber?
And then there were issues with Even More light globes burning out. These LED globes are really remarkably unreliable, and this is the second one that has died on us in the dining room, where it's almost impossible to get at the ceiling. The big dining table is in the way, and I don't think it's strong enough to stand on.
Monday, 17 August 2020 | Dereel | Images for 17 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Riding arena, next step
|
Topic: Stones Road house, general, opinion | Link here |
Two weeks ago I spoke to Sandra of the Golden Plains Shire planning department about the planning permit for the covered arena. She said she would check and get back to me in two days.
She didn't. OK, finally get round to calling her again. As I feared, we will need a new planning permit. $640, if we can convince her that it's for private use only. She asked for my email address and said that she would send URLs for the required forms. Nothing arrived.
OK, off looking, and with Google's help found this page, with links to many PDF documents, including an application checklist. There I read:
A COPY OF THE TITLE
Provide a full copy of the current certificate of title (no more than 60 days old)., including any registered restrictive covenant/s or Section 173 agreements that may apply to the land.
Include a copy of the Plan of Subdivision / Title Plan.
A copy of title, covenants and plans can be purchased from Land Victoria www.landata.vic.gov.au or see Councils website My Home/Planning/Planning Permits for more information
OK, which do I choose? http://www.landata.vic.gov.au is a home page, of course, requiring searching. OK, try https://www.goldenplains.vic.gov.au/My%20Home/Planning/Planning%20Permits. Ha ha, only joking, that's a 404. And there was nothing else obvious.
OK, back to http://www.landata.vic.gov.au, where, after searching, I found this page, which wanted me to specify what I was looking for in some detail before it would divulge anything at all. Finally I got to a page with the URL https://www.landata.vic.gov.au/tpc_confirm_property.aspx, which doesn't work directly. There I read:
![]() |
|
How does that relate to “a copy of title, covenants and plans”? The best correlation I can find is “Register Search Statement (Title)” for “copy of title”, and “Copy of Plan” for “plans”, though the latter seems too generic. And which of those items correspond to “covenant”?
To get any further, I had to select something. OK, select the Register Search Statement (Title). And how about that, I got two more for free:
![]() |
|
For free? Or do I have to pay for them too? Of course! Not to (finally) mention a previously unmentioned service charge:
|
OK, pressed “Next”:
|
The message is wrong, of course. My guess is that it was a timeout. Pressing “Retry” took me back to the start of the mess.
So I'm going to have to call up the council again to make up for their inadequate documentation.
And then there's the question of other things I need to supply:
PROVIDE A COPY OF A SITE PLAN
The site plan must be A4 or A3 in size and must show;
Orientation of land (direction of north). A scale. Location of proposed and existing buildings on land and show setbacks (ie: metre distance from title boundaries). Topographic features on land, ie: dams, watercourses, native vegetation. Current or proposed driveway entry. Site plan with levels and a feature survey for subdivisions.
Dammit, all the reasons I wanted to have nothing to do with the project. Native vegetation! That caused me untold pain six years ago. At least this time it's clear that we won't be removing any vegetation apart from a couple of weeds. Maybe we should spray them before an inspector comes to take a look.
And then there's a question of boundary distances, which so far nobody has addressed. Wouldn't it be fun if they decided that we couldn't build there at all, or even that the current arena contravenes some regulation? Somehow it's all more pain than it's worth. I put up with it for the house, but for a riding arena cover?
Minor garden stuff
|
Topic: gardening | Link here |
I should be pruning roses, of course, but somehow my inner pig-dog (to translate a German saying) got the better of me. Planted a few more Thai Basil seeds, and outside set to removing the Pelargonium bushes that have passed their use-by date:
![]() |
|
But they're now quite substantial. The trunk must be 8 cm across:
![]() |
|
So I'll need something to cut them with.
Tuesday, 18 August 2020 | Dereel → Ballarat → Dereel | Images for 18 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Time for more shopping
|
Topic: general, opinion | Link here |
A number of things have ganged up on me: blocked drains, mousetraps, the wine glasses that I had planned to buy last Friday, the coffee that I had forgotten to put on the shopping list...
So off to town, first to Bunnings, where I looked for mousetraps and drain cleaners. The mousetraps were many and varied, but this one caught my eye:
![]() |
|
Not the packaging, but the “100% Catch & Kill GUARANTEED”. That's only on the boxes, not on the individual packages. But clearly a good start.
Drain cleaners were the same old thing. Chemical or mechanical? There were many and varied chemicals, including some that were once-only (put in the whole package), apparently containing enzymes and sodium hydroxide. $8.50 a go! Or just plain sodium hydroxide, no enzymes, $3.50, enough for about 8 attempts. So the NaOH it was.
And the mechanical ones? Not much choice. A “gun type drain cleaner”, a 5 m spiral that goes down the drain, or a “high pressure unblocker”, something like a plunger on steroids (or advertising hype). The “gun type” is simply the handle. But 5 m would probably not do it; my guess is that the blockage is outside the house, more than 5 m away.
While I was there, also looked for sharpeners for my secateurs. What do they even look like? Could this be one?
![]() |
|
![]() |
|
The maker was too polite to put a description on the packaging, but it looks as if it could be. But I was sent down to the tool sales area, where a surprisingly helpful sales assistant confirmed that yes, it's a sharpener, but showed me another one, well documented, and with his personal seal of approval, for about $8 less. I think I would have chosen it even at the same price: why should sloppy makers get away with bad documentation?
Then on to Harris Scarfe to look at and measure the glasses that I had seen (and not reported on) on Friday. 21 cm high! WHY are modern wine glasses so big? I want a glass, not a bucket, and I had set myself a limit of 18 cm. Even that's big by the standards of the last millennium. So still no glasses.
Catching mice
|
Topic: animals, general, opinion | Link here |
Our experience with catching mice in the garage has been a bit mixed. Last week Yvonne bought a couple of mousetraps, and we caught a mouse on Thursday. Then I baited it again, and the mice stole the bait. On day 3 (Saturday) I put some lard in the trap in the hope that the mice couldn't steal it. But they didn't, though they managed once again to trigger the traps without being caught. So after two days I changed bait to... cheese. But not just any cheese, Saint Albray. That's not as wasteful as it seems: Saint Albray is wasteful, and we used the rind that we would have otherwise thrown away.
Bingo! Finally both traps caught a mouse. Piccola was delighted and brought one to the door to show me:
![]() |
|
![]() |
|
Well, most of it, anyway. Later she brought the other one:
![]() |
|
After returning from shopping, took a look at the “electronic mouse trap”. Yes, I suppose it's marginally electronic. It's also clearly modern: the documentation is appalling. It's not even in their interest:
![]() |
|
It claims
|
And then? Throw it away? There's a * there, which suggests a footnote somewhere, but no, only joking. Further reading shows that it takes 4 AA batteries (not supplied), and that's how many mice you can expect to kill per set of batteries.
OK, cut through this horrible stiff plastic, risking another wound like I inflicted on myself last month and took a look at the thing:
![]() |
|
Where are the instructions? On the strip at the right of the packaging, it seems. They refer to the “enclosed instruction manual”, and after a little searching (it was hidden in a fold of the backing carton) I found it: a black-and-white version of the same instructions.
The real question was: where is the “tray at the front of the unit” mentioned in step 2 of the instructions?
![]() |
|
That's quite legible here, but the original is 2.8×3.4 cm, and it didn't really match what I saw. It proved that the tray had become dislodged and was lying on its side. After putting it in the correct place, it looked like this:
![]() |
|
Finally got it together and discovered that the batteries must be non-rechargeable (in other words, 1.5 V). Tried it with NiMH batteries anyway, and it didn't even turn on. I had no non-rechargeables, so I put in NiZn batteries, which worked. And how about that, within 90 minutes we had our first mouse! If it carries on like this, it will be well worthwhile.
Unblocking drains
|
Topic: general, opinion | Link here |
Also took the first steps in unblocking the drains. The instructions want 55 g of sodium hydroxide in a litre of water. Pour down the drain, wait 15 minutes, and flush.
How can that work? Somehow I get the feeling that this is designed for a different kind of blockage than I have. My suspicion is that the drain is blocked with congealed fat. Yes, sodium hydroxide works against that, but a 5.5% solution or weaker? OK, did it anyway. Filled the sink with water, let it out after 15 minutes. Yes, maybe it's draining faster. Help with the plunger? No, it didn't help. It seemed to get worse. It finally drained, but we're anything but done. I'll try another dose tomorrow.
Goodbye NiZn?
|
Topic: technology, general, opinion | Link here |
While taking the photos of Piccola and the mice, my flash gun gave up on me: low battery. That can happen, of course, but these were freshly charged. Checking, I found that three of the batteries were showing round 1.72 V, which is normal for a fully charged battery, but the fourth was showing only 0.7 V.
I've seen similar things with NiMH batteries. I'm coming to the conclusion that rechargeable batteries are not as uniform in their characteristics as non-rechargeable batteries. They may work well individually, but in series with others it seems that they tend to fail like this.
I've been using NiZn for power applications for nearly 9 years now. Time to move on?
Wednesday, 19 August 2020 | Dereel | Images for 19 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Mobile phones: too modern?
|
Topic: technology, opinion | Link here |
Seen on IRC today:
<mattik> older brother complaining about no seconds field on his samsung android hehe
<mattik> there again have no idea how to do that with iOS on iphone either
<afark> there is a new icon on my phone and I don't know what it means..
<afark> ..a quarter filled rectangles with wobbly lines on either side
<afark> also running out of green... might have to plug her in again
<pj> chuzz: If you swipe down, it might associate it with some text.
<bb0> gc: you get seconds if you go into one of the clock apps
<bb0> gc: I can't help with the wobbly lines icon
<bb0> gc: take a screenshot of it?
<bb0> gc: then crop and do a reverse image search in google
How about that, two of the things that I love to grumble about: lack of seconds display on mobile phones, and inscrutable icons. In principle that's not surprising, except that these are the people (well, some of them, anyway) who otherwise don't understand my grumbling. Behind the times, indeed! But it seems that people are catching up with me.
Microsoft surprises
|
Topic: technology, opinion | Link here |
I have migrated most of my Microsoft activity from dischord (Microsoft “Windows” 7) to distress (“Windows” 10), but the scanner is still connected to dischord, and it will stay there until I move dischord to the other side of my monitors.
Today I wanted to scan something. Turn on the scanner, wake dischord, start the scanner program. “Can't start scanner” or some silly popup. Do you want to open the Troubleshooting Assistant? assistant? No, not really. I don't think I could stomach what I found inside. But the message persisted. Check the “Control Panel”. Device is working normally.
Dammit, the Microsoft solution is simple: reboot. Wait. This machine normally doesn't have a display, so it's hard to know what it's thinking. But after a considerable time I gave up and switched the rightmost monitor to dischord, not helped by the fact that it has Yet Another button layout and almost completely illegible button markings:
![]() |
|
The inscriptions are there, but it takes some photo magic to see them:
|
|
Finally selected the right input. No display! OK, try rebooting again. No reaction. Long press on power button, and the computer came back normally. Boot. Start Windows normally or Launch startup repair (recommended). OK, launch repair. “Your computer was unable to start. Attempting repairs”. For a very long time, half an hour or so, with no disk access. What is it doing? OK, press the Cancel button. “The current repair operation cannot be cancelled”:
![]() |
|
OK, what are my options? There's nothing really important on this box, and one option would be just to remove the disk and put in one of the myriad “Windows” 10 disks that I have lying around. But that would take time, and I'd have to go looking for the scanner software. How about just rebooting and starting “Windows” normally?
Oh. How do I restart? I can't interrupt. Power button? “I said can't interrupt!”. OK, long power button. And up it came. After 15 seconds displaying the startup screen, it went blank. Hmm, problems after all? Let's see if it repents. And how about that, after 7 minutes of complete darkness, just as I was thinking about looking for replacement disks, it came back online as if nothing had happened. And everything worked, including the scanner.
I'll never understand Microsoft.
The better mousetrap
|
Topic: animals, general, opinion | Link here |
Into the garage this morning to check the mousetraps. The “electronic” mousetrap had another occupant, but the el-cheapo mousetraps were empty. One of them didn't even have any bait in it any more. Clearly a mousetrap that's not as clever as mice.
Changed the bait, including in the electronic one. An hour later Yvonne came with the interesting information that it had caught another one! That's four mice in 24 hours, as much as the two el-cheapo ones managed in 5 days. And it has the advantage that I can let Piccola near it, since there's no way she can hurt herself on it.
Thursday, 20 August 2020 | Dereel | Images for 20 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Mouse mutilation
|
Topic: animals, opinion | Link here |
Into the garage this morning to find a mouse in the “electronic” mousetrap and one in a conventional mousetraps. It was mutilated beyond recognition:
![]() |
|
![]() |
|
![]() |
|
How did that happen? I've seen something like this before with a rat:
![]() |
|
![]() |
|
But on that occasion the rat was recognizable, and I suspected that it had done the damage to itself. Today I could barely recognize any features of the mouse, and it appears to have been chewed on by something else. Another mouse? Are they cannibals? In any case, one thing is clear: the “electronic” trap is much more humane.
More drain deblocking
|
Topic: Stones Road house, general, opinion | Link here |
The sodium hydroxide that I put down the drain didn't have any immediate effect. But what if I repeat it a couple of times? Today, for the third day in a row, I put my 55 g in a litre of water down the drain and flushed it. And this time it looked as if things were improving. There's plenty left.
But then in the evening I found the toilets not flushing properly. Clearly it's not enough. Took a look at the photos I took during construction, notably this one:
![]() |
|
The long pipe in the trench is part of the rainwater collection system, but there are three risers: one on either side of the slab for the hot water system, and one on the other side of the rainwater pipe, with a connection to one of the risers near the house. Out again to examine what had become of them. I couldn't find the rear riser, and though I found one on the other side of the rainwater pipe, it's not clear that it's the one in the photo. But both of them showed no blockage, and the water level in them was far lower.
So: it looks like we will need professional help.
Bloody “air fryers”!
|
Topic: food and drink, opinion | Link here |
We had planned a West Lake Fish for dinner tonight. Where do we get the fish? From the freezer, of course, but it was a Tilapia mossambica, not exactly fresh water. And far too thick. The recipe I had wanted it boiled for two minutes. No way that would work.
OK, what can we do with it? Fry it? Bake it? In the end, decided on the “air fryer” (specifically, the “hair dryer”, which was the only one big enough). How long? 10-15 minutes at 230° according to the instructions.
230° seems too hot for this kind of fish. Gave it 20 minutes at 200°, after which it looked fine—but was almost raw inside! What a waste of a fish! Next time I'll measure temperatures before serving.
Friday, 21 August 2020 | Dereel | Images for 21 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
More mice
|
Topic: animals, general, opinion | Link here |
The morning check of the garage showed two more mice in the conventional traps, but none in the electronic trap. OK, get rid of them, and leave the electronic trap to do its thing. Nothing, all day long.
Unblocking the drains
|
Topic: Stones Road house, general, opinion | Link here |
First thing this morning set to looking for a plumber to unblock our drains. First tried Brad (phone 0409 323 878), a plumber with whom Petra Gietz had had dealings, but he wasn't able to come today. How about calling Mumpy? You can get his phone number from the Internet.
Too much trouble. Called Paul Kinna from Ballarat Plumbing Services (0418 527 576), who had installed the hot water service last year. No, he doesn't do that sort of thing, but the people to call are Black Hill Plumbing and Blocked Drains, phone 0413 611 225, who specialize in that sort of thing. Called up and got a confirmation that he could come out today. Just “text” me your address. OK, agreed on email, and with only a couple of confirmation calls and an emergency shoved into the queue, he finally arrived at 16:00. Took a look around, saw the Chromagen hot water system that I still haven't disposed of, and said “They're crap, aren't they?”. Clearly they have a reputation—that's pretty much what Paul Kinna said last year.
He looked at exactly the risers I had examined yesterday, and got me to flush toilets and let water out of the kitchen sink. What he didn't find was any problem. Indeed there was no blockage at the moment, but it seems just a matter of time before it would happen again. No, he thought, it's probably gone by now. Poke around there (in the riser) if you have any more problems. That'll be $100, please.
And how about that, he had a credit card reader connected to his mobile phone, with which he was able to cash the sum immediately and send me a poorly formatted email message as receipt. At any rate I suppose I have a feeling of security, and if it does happen again, I know whom to call.
Another grid power outage
|
Topic: Stones Road house, general | Link here |
Another very brief grid power failure today at 2:22:19.
Garden flowers in late winter
|
Topic: gardening, general, opinion | Link here |
Winter is nearly over, at least according to the calendar. It has been raining heavily, and when the plumber arrived, he reported having seen snow on the ground in Enfield, something unusual at any time.
Some of my problem plants have given up, including, sadly, the original Hibiscus rosa-sinensis “Uncle Max”, presumably not helped by being put outside:
![]() |
|
I wonder if I can still plant the stems.
On the other hand, the Buddlejas that I grew from cuttings are doing well, though they're not exactly tree-like:
![]() |
|
![]() |
|
I should finally tie them to a stake.
And Yvonne tells me that our Ginkgo biloba is also looking dead:
![]() |
|
That's not unusual for this time of year, but there should be some sign of new growth. It's so wet out there at the moment that I didn't want to come closer to inspect it.
Spring is on its way, and the one kind of plant which seems to do well here is bulbs and tubers. In particular, daffodils are everywhere:
![]() |
|
![]() |
|
![]() |
|
Apart from bulbs, the Alyogyne huegelii is also flowering profusely:
![]() |
|
The Camellia japonica is starting to flower:
![]() |
|
![]() |
|
The leaves still look as if it's missing something. But what? By comparison, the Marriotts have what I think is a very similar bush, and it looks fine:
![]() |
|
The Clematis “Edo Murasaki” that was torn apart three months ago isn't completely dead:
![]() |
|
The stems are about 5 cm long, and at the top it looks like buds. Is that going to work? It almost looks like it will make it.
The Paulownia kawakamii is producing new shoots, but there's only one bud, far fewer than last year:
![]() |
|
![]() |
|
And the Schinus molle that seemed to be doing better after transplantation has now lost a lot of its leaves:
![]() |
|
Hopefully it will improve in the spring.
The Strelitzia nicolai is also coming out with its first flower of spring:
![]() |
|
![]() |
|
The Tropaeolum had taken over the “front” bed in front of the “library”, so I pulled most it out, with the result that there's now only a single flower:
![]() |
|
On the other hand, that has given light to the Hellebores and Cyclamen. The former is flowering, and hopefully the latter will soon too:
![]() |
|
![]() |
|
And as I suspected, the Epazote that I planted 5 months ago is growing:
![]() |
|
Saturday, 22 August 2020 | Dereel | Images for 22 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Where have all the mousies gone?
|
Topic: animals, general, opinion | Link here |
Into the garage early this morning, as on every morning recently, to look at the mousetraps. No mice! And the bait was still there, so they hadn't just stolen it. What had happened?
Yvonne's first question was “how many droppings are there?”. None! Have we eradicated the mice already? Could only 15 or so mice leave that many droppings?
A bridge over calm waters
|
Topic: general, opinion | Link here |
It has been really wet recently, and there's water lying on the ground everywhere. For once, one of Yvonne's ornaments came in handy: an ornamental bridge that we put across the “river” of gravel that runs parallel to the north side of the house:
![]() |
|
Sunday, 23 August 2020 | Dereel | Images for 23 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Return of the mice
|
Topic: animals, general | Link here |
Yesterday we seemed to have run out of mice, but one can hope, so I left the traps set. And today I had two more mice, but no more droppings. Do they maybe come in from outside during the night? Given the current wet weather, it's not surprising that they might have stayed away. In any case, the great advantage of the “electronic” trap is that I can just leave it without worrying about any collateral damage. When I'm confident that the garage is clear, I could move it to the hay shed.
Another PV battery calibration
|
Topic: Stones Road house, general, opinion | Link here |
Given the current wet weather, I decided to alter the charging limits for the PV battery, from 20%/30% to 60%/80%. Kept an eye on things, but it didn't do what I expected: charged both from the PV array and the mains, and way over the 80% limit.
Huh? It took me a while to realize (and check) that the system had entered a battery recalibration cycle, on the dot 30 days after the last one. So once again it's clear that things were greatly improved by the firmware upgrade in March.
But this time things were different. It charged the battery to 100%, like previously, and then discharged—but only to 10%, not the 0% I had seen on previous occasions. And it used the PV input during this time:
|
That's all very positive, but why now? Has Fyodor silently installed new firmware? Checked the current settings and found:
Device ID | 0AM172B16A22 | |
FW Version | AAX1055AP | |
HW Type | AAX0075 | |
IS Monitor | connected | |
VPN | connected | |
Uptime | 0/4/8 22:18:6 | |
What does that “Uptime” value mean? It would make sense to think that it means 4 months and nearly 9 days. But the hours and minutes don't change when I reload. Still, this would match probably the last time the inverter was power cycled, as the result of Daryl Buchanan's incompetence. Arguably that means that the firmware is the same, but in fact installing new firmware doesn't require restarting the inverter, or at least it hasn't done so so far.
Anyway, I now have the current firmware version, AAX1055AP; the last time I recorded it was in January, when I installed version AAX1055AL. And then there was the version in March, which I didn't record; has it changed since?
Mobile phone clocks again
|
Topic: technology, opinion | Link here |
Mail from Michael James today, informing me that iPhones can display seconds on the idle screen—but only on the analogue display, unless you jailbreak them. That's even more confusing. Why limit it to analogue displays? Are digital users too leet to want to display seconds?
Another possibility comes to mind: the things are inaccurate! Or at least, mine is. I have an app for time display. What's it called? Anonymous, it seems. At least nothing wants to display its name. And it is something like 10 seconds slow!
In the Good Old Days that would have been not only normal, but even good. But these things are networked, and they use NTP. How can they drift that far?
More water
|
Topic: general, opinion | Link here |
The rain continues, even if the Bureau of Meteorology doesn't want to know: it claims that no rain at all has fallen in Ballarat since 20 August 2020, and then only 0.8 mm. By contrast, here we have had 60 mm in that time, and it shows:
![]() |
|
![]() |
|
![]() |
|
Somehow the water likes collecting in places where I want to go.
Potatoes revisited
|
Topic: food and drink, opinion | Link here |
Yesterday Yvonne made Zürcher Geschnetzeltes and real Rösti. But not much Rösti. How much? 4 potatoes, which reminds me of the “clove of garlic”:
![]() |
|
![]() |
|
![]() |
|
Today I made some deep-fried potatoes, so I checked the weight of the remaining potatoes; about 60 g each. I would have thought that 240 g of potatoes would have been ample for the Rösti, but to be on the safe side I made 350 g today. Boil for 15 minutes, cool down, deep fry until brown.
What's missing here? Feedback. I didn't check if the 15 minutes were enough, and they weren't. Why didn't I check before discarding the water?
Monday, 24 August 2020 | Dereel | Images for 24 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Still no mice
|
Topic: general, animals | Link here |
Once more the mousetraps in the garage had caught no mice, and the bait was still there. Have we beaten them already?
Fixing rain gauges
|
Topic: general, technology, opinion | Link here |
The Bureau of Meteorology has still not
recorded any rain for Ballarat since Thursday. Clearly something's wrong. Ring them
up? No, nowadays you use email web forms for communication. The BoM's one is
particularly emetic: fill out one of hundreds of reasons, browser type and version, enter
your name once but your email twice, the second into a field that doesn't accept pasting,
and at the end you had to answer a trick question. People, you should be open to
communication, not trying to annoy people!
Filled out the form and got the confirmation:
REF2020-237-13537.
Your message was:
Your rain gauge in Ballarat http://www.bom.gov.au/climate/dwo/IDCJDW3005.latest.shtml shows 0.8 mm rainfall since Thursday 20 August. In this time we have measured 60 mm. I cant believe that this is due to local differences.
I find your requirement of a feedback question offensive.
Browser/OS Details:
Not browser related
I got a reasonably quick reply:
We are aware that there is an issue with the rain gauge at the Ballarat Aerodrome weather station and technicians have been sent to the site today to hopefully resolve the issue.
I suppose that's reasonable, though “we are aware” sounds very much like a truncation of “we are now aware”. But it somewhat diminishes my trust in the data that they publish.
CJ hacked?
|
Topic: technology, general, opinion | Link here |
(VoIP) phone call from CJ Ellis today, with the information that his Internet is broken.
That sounded unlikely, but it seems that some scammer has been annoying him, and now he has lost Internet connectivity from his computer—not the new one, because sound doesn't work on that, but the old one (which I think is running Ubuntu). Does he really have connectivity issues, or is there something else? Got him to start TeamViwer, but he couldn't get a user ID or password, which appears to be TeamViewer's polite way of saying “sorry, can't contact our server”.
So I got him to connect the other computer, which connected fine, and I was able to browse through it with TeamViewer. Why no sound? This thing runs Microsoft “Windows” 10, and I have little experience with that. But it told me that there was no hardware support, and would I like to run the troubleshooter? This time it told me that I should reboot, because it might have installed drivers, but the status showed nothing. This is painful at the best of times, but remotely with a non-technical person at the other end doesn't work well. He'll have to bring his other computer here some time, so we can look at it then.
Another reason for CJ to bring his computer here is because he wants to connect
a webcam camera to it. That'll be interesting, but it's something that I should
investigate myself.
Tuesday, 25 August 2020 | Dereel | |
Top of page | ||
previous day | ||
next day | ||
last day |
Procrastination
|
Topic: general, opinion | Link here |
So what do I have to do in the immediate future?
What did I do today? Nothing. Well, I investigated backup disks, but the cheapest offer seemed to want me to pay 1.5% more than advertised (for PayPal premia, but with no alternative). I should call them up, but I didn't.
Why are some days like this?
NBN revisited
|
Topic: technology, opinion | Link here |
Discussing Internet traffic charges with Callum Gibson on IRC. I'm currently paying $65 per month with Aussie Broadband for 600 GB at 25 Mb/s down, 5 Mb/s up. That's more than I currently need, since they have generously cut charges during the daytime, so I could go down a few 100 GB if I wanted. Or get an unlimited tariff for $4 more per month.
Off to confirm. Damn, this web site is confusing. To start with, it doesn't want to know about the old firefox on eureka. OK, fire up the latest and gratest firefox on dereel. “Please answer this CAPTCHA”. Damn you! Answered the CAPTCHA. “Please answer this CAPTCHA”. What's an airplane, anyway? At least they didn't want me to identify crosswalks, whatever they may be. But I couldn't get past the bloody CAPTCHA. Also had (different) problems with chrome. Time for a nastygram to their support. Finally accessed it with chrome on eureka. And there's none of what I saw a few months back, just a total of three tariffs with silly descriptions: $70 for 100 GB (“Casual”), $75 for 500 GB (“Everyday”), $79 for unlimited (“Family”)!
Huh? Where's my tariff? What speed is this, anyway? They say “65Mbps [sic] Typical evening speed”, but no way to change, and no nominal speeds. This morning I loaded the German News at 26.5 Mb/s, so at least for my “plan” they stick to the speeds. Was invited to “chat” with somebody on the web page, where they explained that there's only one tariff, 75/10, and it's been like that for ever. Further questioning suggests that “for ever” means “since I've been working here”. She also claimed that the speed variations are due to interference in the fixed wireless link. That's not right: it's congestion. And of course she couldn't explain why the nominal speeds (or any upload speed) aren't stated on the web page.
But it seems that Aussie have changed their tariffs radically, Yet Again, and haven't informed me. That's quite frustrating: I don't think I'd really get much benefit from the 75/10 speeds, and if I had known they were going to do it, I could have changed to the unlimited 25/5 tariff for $69. Now if I want to change to more data, I'll be up for $14 more.
But it's an interesting development. Where did they get 75/10 from anyway? Is it even correct? And why are they taking their customers for fools?
Their key facts sheet (PDF) only specifies the download speed, and claims that “typical” evening speeds are 43 Mb/s, not the 65 Mb/s mentioned on their overview page (which says nothing useful until you you specify your address). That page also states “$79 per month [for unlimited traffic] with no connection fee”, and further down states
Is there any cost to connect?
We do not have any standard connection fees for our nbnTM services. A new housing development charge of $300 for the first nbnTM installation of premises in a newly developed area i.e. a suburb with no pre-existing telecommunications infrastructure.
“This sentence no verb”. Also no else. But the other documentation claims that things will be installed for free, as before, and the Critical Information Summary (another PDF document) states the speed as
This service can perform to a maximum of 75Mbps download and 10Mbps upload, but these speeds can be affected by a range of factors including line of sight to the tower, NBN cell congestion and NBN backhaul congestion.
I suppose that's adequate CYA, but somehow things seem to be going downhill. When I signed up for the NBN, the cheapest tariffs were round $30 per month, though my charge was $54.50 per month for only 50 GB! So it's likely that the cheap tariffs were low-volume loss leaders. But it also seems that bandwidth and traffic is no longer the issue. I should investigate what other providers are doing.
Wednesday, 26 August 2020 | Dereel | Images for 26 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Huevos sobrantes
|
Topic: food and drink, opinion | Link here |
I've made huevos a la tigre a few times now, but somehow it doesn't seem to quite work. But this morning I was faced with a number of leftover potatoes and carrots, along with a raw egg with a broken yolk. Breakfast?
![]() |
|
That's with some peas, chorizo, onions and a fried chili. After further processing, I ended up with:
![]() |
|
And the result? OK. Maybe that's the best way to approach this kind of breakfast. It's a far cry from the original huevos a la flamenca.
Chicken kebabs again
|
Topic: food and drink, opinion | Link here |
Sinclair Meats in Delacombe are open again after closing down to eliminate COVID-19 infections, so we had chicken kebabs (for want of a better word; “spits” was the traditional word, but is too easy to misunderstand) for dinner tonight.
Somehow things didn't work well. It's always fiddly to put the spits on the rotisserie of the “air fryer”, but today one of the spits helped by losing half its load. It seems that the slices were too thin, and they just slid town the spit. And when we were finished, the opposite happened: I couldn't get one off the plate at the end of the rotisserie. It broke in the middle, and I had to cut it off at the table:
![]() |
|
There must be an easier way. As it is, it's also a problem cleaning things when it's done.
Thursday, 27 August 2020 | Dereel | Images for 27 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Inopportune zipper problem
|
Topic: general, animals, opinion | Link here |
Yvonne came in in early afternoon and asked me to take a photo. OK, no problem:
![]() |
|
![]() |
|
![]() |
|
But why that? It's this horribly expensive pneumatic riding vest that she bought last year: when you fall off a horse (as you do), it inflates from a small cylinder of nitrogen and prevents serious injury.
Now wouldn't you think that that kind of clothing would be made with a really robust zipper? It seems that it has to be sent to the manufacturer for repair.
A new duck
|
Topic: animals | Link here |
Seen while walking the dogs today:
|
They prove to be Australian shelducks. The one with the white rings round the eyes is the female. I don't think I've seen them before. The ones I remember are the common Australian wood duck and some “New Holland” duck, though I can't find any reference for the latter.
Of course, part of the issue is my memory. It seems that I have, indeed, seen shelducks here before, about three years ago and in exactly the same place:
![]() |
|
What focal length lens?
|
Topic: photography, opinion | Link here |
In the course of the years I've tended to use longer and longer zooms. My first Olympus camera, the E-510, came with a 14-42 mm lens, which I quickly replaced with a 12-60 mm lens. And still later I got a 12-100 mm lens, which is what I use as a “standard” lens nowadays.
But that's insufficient for things like the shelduck photos. I took the photo at 100 mm focal length, but I still had to crop it by 73%:
![]() |
|
|
To take that photo without cropping would have required a 450 mm lens (corresponding to 900 mm on the old 35 mm cameras). I don't even have one that long, though the Leica DG Vario-Elmar 100-400 mm f/4.0-6.3 would come close. But I can't walk around with that all the time.
For the time being, it would make sense to take the Olympus M.Zuiko Digital ED 14-150 mm f/4.0-5.6. But maybe it's an argument for the new 12-200 mm lens.
unix: old, worn-out magic word
|
Topic: technology, opinion | Link here |
Today I discovered that all ten xearth processes running on eureka's root screens had crashed. Why?
Tried to restart them.
=== grog@eureka (/dev/pts/42) ~ 2 -> xe
Error: Can't open display: unix:0.1
Error: Can't open display: unix:0.0
Error: Can't open display: unix:0.3
Error: Can't open display: unix:0.2
Huh? They're all working. In fact, I ran that script on unix:0.2. What has happened?
It proved to be the display name: unix:0.0, for example. Change that to :0.0 and all is well. In principle, it's the same thing: in front of the display identifier you specify either a system name (for TCP domain) or unix (for Unix domain). It seems that the use of the name unix is deprecated, but it worked fine until last week.
What changed? It wasn't the script. I haven't updated it for years:
RCS file: RCS/xe,v
Working file: xe
...
revision 1.9
date: 2015/11/25 04:13:00; author: grog; state: Exp; lines: +16 -16
Use xearth, not xearth-old.
And it wasn't the X servers. I'd like to say “I haven't started a new one for years”, but in fact they haven't been running for quite a year:
=== grog@eureka (/dev/pts/42) ~ 25 -> ps aux | grep Xorg
root 4570 1.1 0.1 3484960 49588 v0 S 11Sep19 11064:24.36 /usr/local/bin/X :0 -config xorg-0.conf -logverbose 6 -listen tcp -auth /home/grog/.serverauth.4548 (Xorg)
root 39256 0.0 2.3 4131444 761248 v4 S 13Sep19 2490:13.76 /usr/local/bin/X :1 -config xorg-1.conf -listen tcp -auth /home/grog/.serverauth.39236 (Xorg)
But on 16 August I did some experimentation with a third X server. I didn't record it, but I think I tripped over some parameter and may have confused server :0. Could that have been the reason? I don't know how to find out now. In any case, it makes sense to remove the unix in the display identifiers, after which all is working well.
Friday, 28 August 2020 | Dereel | Images for 28 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Chili seeds, a month later
|
Topic: gardening, opinion | Link here |
A month ago I planted 7 different kinds of tomato and chili seeds, one seed per pot:
![]() |
|
How did they do? It's interesting. The tomatoes are going fine. Two of the chilis (serrano and arbol) have come up reasonably well, though not as well as the tomatoes. But the other two (poblano and pasillo bajio) have shown no signs of life:
![]() |
|
OK, as I noted at the time, I should have soaked those two seed types before planting. But does that make the difference between success and complete failure? Off to soak 5 remaining seeds of each type. Oh. I only got 7 seeds of the poblanos, the ones I was particularly looking for. Soaked them and planted them, but I'm not very impressed. It will be up to 3 weeks before I can give up on them; maybe I should try a different supplier.
Saturday, 29 August 2020 | Dereel | Images for 29 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
Bibimbap out of character
|
Topic: food and drink | Link here |
How do you serve bibimbap? It seems that there's a tradition of hot stone vessels, but it's also served in normal bowls, where it's difficult to separate things. Yvonne had an idea: banana leaves. Well, the fake banana leaves that people use in Singapore, a world away from Korea:
![]() |
|
I can't make up my mind whether I like it like that or not. I'm sure the Koreans would be horrified. And yes, I forgot the kimchi.
Goodbye, Uncle Max
|
Topic: gardening | Link here |
My uncle Max died over a year ago, but the Hibiscus rosa-sinensis that he gave me for my 60th birthday lived on. But it, too, was sick, and I repotted it earlier in the year. But since then it hasn't really done well, and last month I put it outside to spray against scale insects. That didn't improve it: it looked dead.
So today I cut off four cuttings from the main stems and planted them in the hope that they might take:
![]() |
|
They're about 1.5 cm in diameter, and they're still green, so there's hope. What's left is a shadow of its former self, but it's not beyond the bounds of possibility that it might recover:
![]() |
|
I should bring it back inside.
Sunday, 30 August 2020 | Dereel | Images for 30 August 2020 |
Top of page | ||
previous day | ||
next day | ||
last day |
More approach of spring
|
Topic: gardening, opinion | Link here |
Walking the dogs today, there are more signs of spring. Here another Acacia (myrtifolia?) in flower:
![]() |
|
And what's this?
![]() |
|
I had thought it was gorse, an exotic invader, but clearly it isn't: the leaves are succulent rather than prickly.
And this? From the flowers, it's some kind of Fabaceae, but what?
![]() |
|
Now that's gorse. I've never really looked at the flowers closely before.
Face masks: not created equal
|
Topic: health, general, opinion | Link here |
Just about everybody in the civilized world, with the possible exception of some supporters of Donald Trump, now have experience with face masks. There are two basic kinds: the disposable surgical masks that have been in use for ever, and which people know from TV shows about operating theatres. They're designed to protect the patient, not the doctor.
And then there are the KN95 masks that are designed to protect the wearer. They're considerably more expensive. Like the cheap versions, they're designed for single use, whatever that may mean.
For reasons I don't understand, people in the West confused the intended use of surgical masks with the level of effectiveness they would have in the current COVID-19 pandemic. Protect others, not the wearer. That never made sense to me; if they're a barrier to transmission, it should work in both directions. People have come to see that now, but it took a long time, and it's anybody's guess how many people gave their lives for that short-sightedness.
Which one should I use? The KN95 masks cost about $4 at the cheapest, while the surgical masks sell for under $1. And I'm supposed to put one on Every Time I leave the house. Which do I wear?
The real question, of course, is the risk. When I leave the house to walk the dogs, I don't come closer than 30 m to any other human (with the exception of Yvonne). And I'm outdoors, where I can't run into virus aerosols. So to a certain extent, wearing a mask in those circumstances is more symbolic than effective.
So I haven't really run into any danger when I come back from half an hour in the open. Why discard the mask? It's not infected. So I hang them on the giraffe by the front door and use them again:
![]() |
|
Well, there's one reason to discard them. They're only designed for single use, so they're not overly robust. The straps tend to become detached:
![]() |
|
That's fair enough. So on Wednesday Yvonne picked up some masks from ALDI. But though they look the same, there's one significant difference:
![]() |
|
They're noticeably narrower, and it makes a difference. When worn they're effectively open at the sides.
![]() |
|
I wonder if there's a standard for them; I know that KN95 is a standard.
Another grid outage
|
Topic: Stones Road house, general | Link here |
Another one-second grid power failure this evening at 18:20:10.
Monday, 31 August 2020 | Dereel | Images for 31 August 2020 |
Top of page | ||
previous day |
Another mouse
|
Topic: general, animals | Link here |
Another mouse in the “electronic” mouse trap this morning, the first in over a week. We really seem to have beaten them, and the mouse trap is doing its job with the occasional straggler. I hadn't expected it to be that easy.
Ports pain
|
Topic: technology, opinion | Link here |
What's the purpose of porting software? I wrote the book on the subject, so I went looking for a statement.
None. Clearly it was too obvious: take some sources you have received and convert them to binaries that run in your environment, typically operating system and display environment.
But lately we've been in a discussion on the FreeBSD lists which makes me wonder if everybody sees that as the purpose. People wanted to remove ports because they no longer compiled under the set of options chosen for the Ports Collection, because there's no upstream maintainer, or just because they're too old. The bikeshed continues, but in the process I wrote:
A case in point: x11/xtset. Maintenance stopped in 1993, 11 days after the FreeBSD project came into existence. It works fine, and I find it very useful. If at some time in the future it should no longer work with the latest and greatest iteration of the C programming language or ports structure, that shouldn't be a reason to discard it.
And of course I received a bug report against xtset: it compiled, but with warnings. Why? Here part of the patch:
--- xtset.c.orig 2002-03-29 22:15:00.000000000 +1100
+++ xtset.c 2020-08-31 15:29:45.583899000 +1000
@@ -135,21 +136,21 @@
- if (pwent = getpwuid (geteuid ()))
+ if ((pwent = getpwuid (geteuid ())) != 0)
In other words, the compiler flags were set to complain about automatic casting of an int return value to a Boolean value. This isn't my code, but I see this as a clean feature. But no, nowadays it's untidy, so it must be made explicit.
OK, it needs to fit the FreeBSD framework. But I'm damned if I'm going to change the base package for it. You want a patch, you get a patch.
OK, add patch, try to build.
![]() |
|
|
“There is no license for this port. Will you accept it?”. AAAARCGH! What kind of bureaucracy can lead to that? Yet Another thing to interrupt automatic builds.
OK, accept license, build. Patch doesn't apply cleanly. Why not? I cut and pasted it from the bug report, and it seems that I managed to change the quality of the white space. After applying the patch manually and recreating it from the differences, it looked identical to the original. But it applied.
Moving on to the build:
=== root@dereel (/dev/pts/0) /usr/ports/x11/xtset 66 -> make all
===> Patching for xtset-1.1
===> Configuring for xtset-1.1
===> Building for xtset-1.1
make[1]: "/eureka/home/src/FreeBSD/svn/ports/x11/xtset/work/xtset-1.1/Makefile" line 9: Need an operator
make[1]: Fatal errors encountered -- cannot continue
make[1]: stopped in /eureka/home/src/FreeBSD/svn/ports/x11/xtset/work/xtset-1.1
What's that? The last target of of the Makefile is:
install: all
install -C xtset ${PREFIX}/bin/
install -C xtset.1 ${PREFIX}/man/man1/
gzip -c xtset.1 > ${PREFIX}/man/man1/xtset.1.gz
Line 9 is the gzip line. Looks fine, right? But it seems that the last line was indented with spaces rather than a tab. Whether that works or not depends on the version of make, and it seems that make, too, has become more fussy. The latest versions no longer work. It seems that it did work for the person (“freebsd”) who submitted the patch.
OK, fix that. Where's my base version? I had far too many floating around. Finally found a version that looked right, and pondered: how do I create a distribution archive? The Makefile target is simple:
tarball:
tar czvf xtset-1.1.tar.gz Makefile xtset.c xtset.1
But do I want that in the distribution? I think not. But it's in Makefile. In the end I gave up and left it there. Now upload to http://www.lemis.com/pub/xtset-1.1.tar.gz, remove my cached version of the old tarball, and rebuild:
=> Attempting to fetch http://www.lemis.com/pub/xtset-1.1.tar.gz
fetch: http://www.lemis.com/pub/xtset-1.1.tar.gz: size mismatch: expected 12046, actual 4310
=> Attempting to fetch http://distcache.FreeBSD.org/ports-distfiles/xtset-1.1.tar.gz
fetch: http://distcache.FreeBSD.org/ports-distfiles/xtset-1.1.tar.gz: size mismatch: expected 12046, actual 4310
=> Couldn't fetch it - please try to retrieve this
Of course: the SHA256 checksum has changed, and I need make makesum. Do that, build the program. Builds and installs cleanly, as you'd expect. OK, check in the new files:
=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 68 -> svn add files
A files
A files/patch-xtset.c=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 69 -> svn ci
Waiting for Emacs...
Sending Makefile
Sending distinfo
Adding files
Adding files/patch-xtset.c
Transmitting file data ...done
Committing transaction...
svn: E165001: Commit failed (details follow):
svn: E165001: Commit blocked by pre-commit hook (exit code 1) with output:
Path "head/x11/xtset/files/patch-xtset.c" has fbsd:nokeywords AND svn:keywords. Remove one.
== Additional errors may compound and may not be accurate ==
Path "head/x11/xtset/files/patch-xtset.c" does not have a valid $FreeBSD$ string (keywords not disabled here)
== Pre-commit problem count: 2
svn: E165001: Your commit message was left in a temporary file:
svn: E165001: '/home/src/FreeBSD/svn/ports/x11/svn-commit.tmp'
What's that nonsense? I've just added the file, and now it has a conflict. That looks like a bug to me. More to the point, though: how do I fix it? I've seen that before. svn propget is maybe not my friend, but it's related.
=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 70 -> svn propget svn:keywords files/patch-xtset.c
FreeBSD=%H=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 71 -> svn propget svn:keywords Makefile
FreeBSD=%H
That looks straightforward enough, but clearly it doesn't like it. OK, remove the property. But how?
=== grog@eureka (/dev/pts/32) ~/Downloads 61 -> man svn
NAME
svn - Subversion command line client tool
...
Documentation for Subversion and its tools, including detailed usage
explanations of the svn, svnadmin, svnserve and svnlook programs, his-
torical background, philosophical approaches and reasonings, etc., can
be found at http://svnbook.red-bean.com/.
Run `svn help' to access the built-in tool documentation.
No real man page! Clearly the book is over the top, but svn help doesn't give me the detail I need. Off to ask Google, which took me to this page, where I indirectly discovered the propdel subcommand:
=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 72 -> svn propdel svn:keywords files/patch-xtset.c
property 'svn:keywords' deleted from 'files/patch-xtset.c'.
You have new mail in /var/mail/grog=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 73 -> svn propget svn:keywords files/patch-xtset.c
svn: warning: W200017: Property 'svn:keywords' not found on 'files/patch-xtset.c'
svn: E200000: A problem occurred; see other errors for details=== grog@eureka (/dev/pts/42) /usr/ports/x11/xtset 74 -> svn ci
Waiting for Emacs...
Sending Makefile
Sending distinfo
Adding files
Adding files/patch-xtset.c
Transmitting file data ...done
Committing transaction...
Committed revision 547150.
Done! Only about 2 hours for a “fix” that wasn't necessary in the first place. Oh brave new world, that has such rules in't.
Recovering hibiscus
|
Topic: gardening | Link here |
I've made cuttings of the Hibiscus rosa-sinensis “Uncle Max” in the hope that they might take, but that still leaves a rump in the big pot. Brought it into the house again to see if there's any life left in it:
![]() |
|
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 |