|
| Greg's diary |
| March 2022 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Tuesday, 1 March 2022 | Dereel | Images for 1 March 2022 |
| Top of page | ||
| next day | ||
| last day |
|
Photos through windows
|
Topic: photography, opinion | Link here |
Looking at yesterday's photos showed a problem. This photo isn't as sharp as it should be:
|
|
|||||||||
There were two, both with the same problem. It looked for all the world like camera shake, but it was taken at 1/80 s, which should be no issue at all for the OM-D E-M1. Looking more carefully (here Graeme's right leg), the problem is only in the verticals, and it's almost 2 distinct images, not smeared. My best guess is that it's related to the (double glazed) window, though that doesn't explain why both images are about equally strong.
|
Earthworks, day 2
|
Topic: Stones Road house, general | Link here |
Warrick Pitcher back today with still more gravel and sand. Helen Miller was there and admiring his work when she suddenly said “he's knocked over a fence post”.
What, our expert Warrick? But yes, he had:
|
|||||||||||||||||||||||
I didn't mention it to him; I suspect that he was pissed off enough as it was. In the evening he returned with a replacement post, which still needs fitting:
|
|||||||||||||||||||||||
Then I got him to dig some holes for trees in the front garden, with moderate success:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And remove the stumps outside the driveway? Sure, no problems. Where does the telephone cable run? Ah, yes, we've been there before when building the driveway:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Back into the office to find the photos. There were only two! The images above are the same photo, the second magnified. And the other didn't show the location relative to the fence. But it did show a stump, one of the two that I wanted to remove. Here 7 years ago and today:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The cable must go straight past the stump, probably the reason that we didn't remove it at the time. And on the other side of the driveway the other two stumps also seem to directly touch where the cable would go. The images above look straight parallel to the cable, and so does this one on the other side:
|
|||||||||||||||||||||||
So for the time being, at any rate, no stump removal.
In passing, we had a lot of trouble with idiots at Telstra while handling liability for the cable. Today I confirmed with Warrick: neither of us paid for it, although we admitted liability. Telstra was just too stupid to handle the situation.
|
Google Pay in Russia
|
Topic: politics, technology, opinion | Link here |
Russia's invasion of Ukraine is the topic of much talk, of course. Today Jamie Fraser wondered on IRC:
<jfraser> i guess russians are pretty dependent on things like google/apple pay?
Google Pay? Who uses that? I don't, and I have no intention of doing so. My guess was that they don't use it much in Russia either. But off to check. And without so much as a by-your-leave I got a message that I had been signed up for the service!
You added your account, googgled@gmail.com, to the new Google Pay app. Go to support.google.com/googlepay for help with the new app.
Damn you, Google, I don't want you! Why are you signing me up? Off through a maze of little twisty menus, all different, looking for ways to cancel the service. FInally ended up with this display:
|
|
|||||||||
OK, try that. No “Data & personalization”, just “Data & privacy and “privacy & personalization”. You'd think that they could at least match their help texts to what the confused user finds.
Finally got to the list of services. No Google Pay, of course. The first message must have been yet another incorrect claim.
Finally found what I was looking for. It seems that Russia has one of the highest penetrations world-wide, over 50%, compared to 39% in the USA, 35% in the UK and 22% in Australia. So yes, it will be interesting to see what effect that has on the Russian man in the street.
In passing, it seems that Statista has decided to charge for their information. That's a bad idea: their information isn't worth paying for.
|
Ukraine conflict: the BBC viewpoint
|
Topic: politics, opinion | Link here |
Seen today on BBC news, after the European Union closed its airspace to Russian planes:
Russian planes have also been banned from UK airspace.
Russia's biggest airline, Aeroflot, said it would cancel all flights to European destinations until further notice in a retaliatory move on Sunday.
So that's retaliation!
| Wednesday, 2 March 2022 | Dereel | Images for 2 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
5 Years E-M1 Mark II
|
Topic: photography, opinion | Link here |
It's been 5 years since I got my current camera, the Olympus OM-D E-M1 Mark II. Since then, Olympus has introduced two new cameras, one, the Olympus OM-D E-M1 Mark III, a replacement for my camera, then sold its camera business to a company calling itself OM System (most of the time, anyway), which has just brought out a replacement for that, with the silly name OM-1 and a web site so hard to navigate that I gave up trying to find the description of that camera. On the other hand, it's on the Olympus web site.
The camera is certainly interesting. Looking at the timeline, the E-M1 Mark I was released in September 2013, the Mark II in December 2016 (3¼ years later), the Mark III almost exactly 2 years ago on 28 February 2020 (another 3¼ years later), and the digital OM-1 some time last month (less than 2 years later). And the OM-1 is a bigger step than any of the previous models, with some really interesting features.
Time to buy the new camera? No, not yet, I think. Five years ago I wondered whether it was worth the trouble to buy the E-M1 Mark II, and I held on to my E-M1 Mark I, now over 8 years old. And I'm finding it convenient to still use the Mark I for “normal” photos and keep the Mark II for more complicated shots.
The real issue is clear: digital cameras have come of age. There's still room for improvement, but when I compare the E-M1 Mark I with the 8½ year younger OM-1, there's a big similarity. Same general appearance, same basic features. 8½ years before the Mark I Olympus was only just bringing out its first real DSLR, the E-1. That was a very different camera: they were still finding out what a DSLR was, and the difference between each of my 4 Olympus DSLRs, introduced between 2003 and 2008, is bigger than the difference between the E-M1 Mark I and the OM-1.
The E-1 was a 5 MP camera. The E-M1 Mark I has 16 MP, and the very latest OM-1 has 20 MP, though like most of its predecessors it'll produce 80 MP images if asked. Resolution is now sufficient for normal use. The fact that I'm still using the Mark I for a lot of my photos is an indication of that.
|
Next garden work
|
Topic: gardening | Link here |
Paul Donaghy over today to fix up the front garden irrigation after Warrick's work yesterday. It took him over 2 hours! No wonder I never want to do this work myself. Now we have the Ginkgo planted:
|
|||||||||||||||||||||||
Hopefully it will finally grow. We have some hope: the oak near it has shown considerable improvement in the last few months:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
The first photo was taken only 5 months ago.
|
eBay: Resolution!
|
Topic: technology, opinion | Link here |
Finally my issue with the Xioami Note 9T has been resolved. Clearly the seller, mobileciti_estore, had attempted to close the case without fulfilling his part of the bargain, and eBay had closed the case without consulting me. But after “appealing” they checked and fixed things. And I got a surprisingly friendly, almost bending-over-backwards message from them:
Good News! I have closed this case in your favour with partial refund.
Thank you for contacting eBay regarding the item "Xiaomi Redmi Note 9T 5G" Item Id "402894423229" for which you had opened the return case on 19th Feb 2022. Rest assured,I am happy to assist you today.
First of all, I hope you are safe with your whole family. I can strongly feel your situation and please be assured that I will surely take this case to my vision and I will personally look after this.
I am really sorry to know that you have received the item not as described but there is no need to worry as I am here to resolve this issue for you.
Being a buyer I can understand your situation in this case and On eBay we have always work for the customer satisfaction and I personally do not want that your buying experience gets ruined due to this transaction. I have reviewed the details and can see that you and seller both were agreed on 30$ partial refund that is the Reason we have decided to closed this case in your favour with partial refund ($30) and you will get this refund in your original payment source within next 3 to 5 business days.
Amount to be refunded ($30)
I am happy as this issue is resolved for you now,Thank you for being a valuable part of the eBay community Since 3rd July 1999. I wish success with your future transactions on eBay..
Stay Safe. Stay Healthy.
That's maybe over the top, but it certainly makes a change.
Are they going to take action against the seller? As usual, they're not saying, but I hope so.
| Thursday, 3 March 2022 | Dereel | Images for 3 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Phở again
|
Topic: food and drink, opinion | Link here |
We had some steak left over from a couple of days ago. Eat it for breakfast? How? It's been nearly a year since I last tried to fake Phở. Part of that is that I simply don't know the dish well enough, and to do it right I hae to cook a meat broth for hours. So of course I cheat with this paste:
|
|||||||||||||||||||||||
Last time I put very little else in, but surely it needs noodles or some such? Took a look in my Vietnamese cookbooks and found yes, noodles and vegetables. The result wasn't bad, but it still needs tweaking:
|
|||||||||||||||||||||||
|
Incremental garden improvements
|
Topic: gardening, opinion | Link here |
As expected, the Hibiscus rosa-sinensis cutting that I got from Chris Bahlo is already flowering, only 6 weeks after planting. Here on 23 January and today:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It's not exactly big, but it's surprising in that it appeared so quickly.
And the Buddleja x weyeriana that had nearly died has also recovered well after being repotted. Here a month ago and today:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Russians raise white flag
|
Topic: politics, animals, opinion | Link here |
The Russian invasion of Ukraine is the theme of the year. They're not doing well in their objectives. Time for a white flag, or a yellow feather?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
eBay customer service again
|
Topic: technology, general, opinion | Link here |
Call on the phone today, asking me about my “customer experience” with the mobile phone. That's really quite extraordinary. Have they changed, or have I achieved some special status through having been with them for nearly 23 years? At least it gave me the opportunity to suggest a change in policy where they don't unilaterally decide to close a case without contacting the other party.
| Friday, 4 March 2022 | Dereel | Images for 4 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Walking with George again
|
Topic: animals | Link here |
Peggy Naumov along this afternoon with Larissa and another greyhound, Lucy, for our weekly walk.
They weren't here last week, and it seems that George missed it. For the first time he showed interest in playing with our dogs, unfortunately not long enough for me to get a good photo. But he carried on sniffing around with interest:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
All would have been well if they hadn't come up to him and infringed on his private space, causing him to snarl at them. But clearly things are getting better.
We went a little further than normal today, half way down Westons Road.
|
|||||||||||||||||||||||
That was clearly too far for Larissa, who just lay down and didn't want to go any further:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Seeing Helena again
|
Topic: health, general, animals, opinion | Link here |
Another person we haven't seen for a while is Helena Mirčić. She got a new puppy, Bary, over a month ago, and we were interested to see how he was getting on. Well, it seems:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
| Saturday, 5 March 2022 | Dereel | Images for 5 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Autumn!
|
Topic: general, gardening, opinion | Link here |
Summer is over It was a cool summer, or at least it had the lowest maximum temperature that we have experienced:
mysql> select year(date), min(outside_temp), avg(outside_temp), max(outside_temp)
from observations
where month(date) < 3
group by year(date);
+------------+-------------------+--------------------+-------------------+
| year(date) | min(outside_temp) | avg(outside_temp) | max(outside_temp) |
+------------+-------------------+--------------------+-------------------+
| 2010 | 5.8 | 20.23 | 44.9 |
| 2011 | 4.5 | 19.28 | 41.9 |
| 2012 | 4 | 20.16 | 45.4 |
| 2013 | 5.4 | 20.28 | 43.2 |
| 2014 | 6.4 | 20.85 | 46.6 |
| 2015 | 6.3 | 18.99 | 40.6 |
| 2017 | 4 | 17.93 | 41 |
| 2018 | 6.4 | 20.45 | 43.5 |
| 2019 | 4.1 | 20.51 | 43 |
| 2020 | 4.1 | 18.75 | 43.2 |
| 2021 | 3.7 | 18.45 | 40.9 |
| 2022 | 6.5 | 20.71 | 38.2 |
+------------+-------------------+--------------------+-------------------+
12 rows in set (4.32 sec)
On the other hand, the average temperature was close to the highest we have had, and the rainfall in February was only 4 mm; the average for Ballarat in February is 42.8 mm.
All that changed suddenly on 1 March, the first day of autumn, when we had 10.5 mm of rain. And this morning we measured another 36.3 mm for a total of 46.8 mm so far, already over the monthly average of 42 mm. And the plants are also looking happier. Could it be that a lot of our issues with the garden are simply a matter of not enough water?
| Sunday, 6 March 2022 | Dereel | Images for 6 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Testing E-M1 water resistance
|
Topic: photography, general, opinion | Link here |
While checking something in the garden today, discovered an Olympus OM-D E-M1 on top of the air conditioner. It was not dry:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|
|||||||||
How did it get there? It was mine, of course, and it seems that I put it on the air conditioner while changing a gas bottle some time yesterday. And of course it wasn't damaged: both body and lens are water resistant. But how could I just have forgotten it for 24 hours?
|
Google “security”: just to annoy me
|
Topic: technology, opinion | Link here |
Message from Google today:
To help keep your account secure, Google will no longer support the use of third-party apps or devices which ask you to sign in to your Google Account using only your username and password. Instead, you’ll need to sign in using Sign in with Google or other more secure technologies, like OAuth 2.0.
Email software, like Outlook 2016 or earlier, has less secure access to your Gmail. Switch to Office 365, Outlook 2019 or newer, or any other email software where you can sign in using Sign in with Google.
Bloody Google! And look at that second paragraph. What makes it assume that I'm using Microsoft?
This all relates to my Gmail account, which at one time I used to filter spam, until it became too totalitarian. I currently collect my mail using fetchmail, invoked at regular intervals via a cron job, and it does the right thing, just not what Google wants.
Spent some time asking on IRC and doing Google searches, most of which made assumptions that I was using some toy graphic MUA or even reading my mail on my phone. Finally came up with stuff like Send and receive Gmail from the Linux command line, which specifically refers to mutt. But it was very long and appears to assume that I want to use IMAP.
OK, on. All I really need to do (I hope) is to authenticate, and for that they wawnt What does Google itself say? Google 2-Step Verification tells me:
Aaargh! Mobile phone! The antithesis of security! So while only I know my password, anybody who has access to my phone can perform the second part of the authentication. How does that improve security? More importantly though, how does a cron job access it?
Somehow Google is getting further and further from my world view and making life more and more difficult. The alternative of closing down the Gmail account looks more and more attractive.
|
Where are my flutes?
|
Topic: music, opinion | Link here |
Not quite 20 years ago I gave Terry McGee two flutes for repair. He still has them, and I haven't had a reply to email from him for nearly 2 years.
OK, time's up. I want the flutes back, in whatever condition, by the 20th anniversary on 17 April. Sent him an email with the Subject: “Please return my flutes IMMEDIATELY”. And how about that, I got a response within 30 minutes! Will do, but what does the other flute look like? Out looking for photos, but it seems I never took any. It's a fairly generic 8 key flute with big holes. Hopefully he doesn't have too many of them hanging around.
| Monday, 7 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More flutes on the way?
|
Topic: music | Link here |
More mail from Terry McGee today—I've had more mail from him in the last 2 days than in the previous 5 years—saying that he had identified my second flute, and that he hadn't started the repair yet, but would do so.
This is quite a different response rate. Can he repad my second Siccama flute? Asked him that, and so far no reply. I should maybe postpone that until I get the other flutes back.
|
Google: MY way
|
Topic: technology, opinion | Link here |
Gmail continues to annoy me. Yes, I can fix the authorization this time—maybe—but what will they come up with next? The real issue is that they want me to do things their way, and that seems to increasingly mean Microsoft or Apple. Not my way.
The simplest solution is clearly to have nothing to do with Gmail. In fact, I don't have very many mail subscriptions that point to Gmail. Let's change the address. The biggest is Hugin, also the one that Gmail loves to classify as spam the most (and that despite it originating within Google!). Off to change the email address.
Oh. It insists on a Gmail address! I can't get it to send anywhere else.
This seems to be yet another indication that Google wants the world to do things its way. I'm getting more and more irritated.
|
Where's my air conditioner man?
|
Topic: Stones Road house, general | Link here |
It's been nearly 5 weeks since Atmos promised to send Tony along to repair the air conditioner within 2 or 3 weeks. Called up again on 5331 9436. Once again Leeann was apologetic, and Tony will get back to me tomorrow.
|
Origins of ICMP
|
Topic: technology, history, opinion | Link here |
Recently saw a link to “Ping! I love that duck! on the Unix Heritage Society mailing list. Fascinating reading. Who would have thought that it has such a long history?
| Tuesday, 8 March 2022 | Dereel | Images for 8 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another anniversary
|
Topic: history, opinion | Link here |
Reading my old diary entries shows that it's been 25 years since I returned to Australia. Well, roughly. Yes, I arrived at Tullamarine 25 years ago today, but I spent most of March and April 1997 moving around, and it wasn't until 4 May 1997 that I finally returned to stay. But how time flies!
|
More chili sauce experiments
|
Topic: food and drink, opinion | Link here |
I've been looking for some time for a mid-range chili-like sauce for huevos rancheros. But somehow all the pastes I have tried have been too hot, not hot enough, salty, too liquid.
There's one paste I haven't tried: gochujang. Last week Yvonne bought some, and today was the day.
Oh:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
How could that happen? Simple: it was a pot that I had had for (presumably) 14 years and only eaten a small part of it. The other one expires at the end of this year.
|
|||||||||||||||||||||||
Should I use the old paste? No, I think not. I'm happy to eat preserved foods long past their date of expiry, but this has clearly really passed its use-by date. I tried it anyway, though, and it tasted normal enough.
And for huevos rancheros? Still too hot by itself, so once again I mixed it with tomato sauce. There must be a better way.
|
Still more CNAV nonsense
|
Topic: general, opinion | Link here |
It's been over a month since we sent off the required documents to CNAV to get Yvonne's French pension reinstated. So far no money has come in, but today we received a letter:
|
||||||||||||||||||||||
I have not received (empty) requested on 10/12/2021.I have received your letter ofThe payment of your pension will be reinstated after reception of .
Can they do nothing right? Ah, the letter was written only a couple of days after Yvonne returned the last one. It would be reasonable to assume that Mme. Croza was not aware of the fact that CNAV takes nearly 2 months to deliver a letter. Once Yvonne has calmed down a little, she can write another letter. It's a good thing we sent the last one registered.
I wonder how you should address complaints to these people.
|
man pages again
|
Topic: technology, opinion | Link here |
I'm preparing to finally commit something to the FreeBSD source tree, which involves using git for the first time. OK, first try with a man page, specifically date(1), which still doesn't know anything about time zones! And how about that, I have forgotten all my mdoc-foo. Where's the documentation for the documentation? Looking at mdoc(7) wasn't very helpful: it shows standard GNU macros.
Off searching on the FreeBSD web site, and found the information in the Documentation Project Primer, which pointed me to... mdoc(7). But it bears no relationship to mine.
More searching shows that it has been completely replaced since I installed eureka. I have it on dereel and friends. But what a run-around!
|
Catching up with workarounds
|
Topic: technology, opinion | Link here |
Over the years, I have run into a number of issues that have irritated me, and I've found workarounds for them. But they're adding up.
Somehow I can no longer enter a file name into a selection box in firefox. Instead it gives me a tree that I need to climb to find anything. Particularly when uploading photos, this is a real pain: my ~/Photos/ directory has nearly 6,000 entries, and it expects me to scroll through them! There used to be a way to use a bash-like file name completion, but that was clearly too convenient. I still haven't found a solution to this one.
And then there's paging terminal output. Once there was a program called more. Then less came around: it's more than more. And for decades I got by with it with no problems.
But then people came along with control codes. And somehow nroff output of man pages looks like this:
DATE(1) FreeBSD General Commands Manual DATE(1)
ESC[1mNAMEESC[0m
ESC[1mdate ESC[22m-- display or set date and time
ESC[1mSYNOPSISESC[0m
ESC[1mdate ESC[22m[ESC[1m-nRuESC[22m] [ESC[1m-IESC[22m[ESC[4mFMTESC[24m]] [ESC[1m-r EESC[4mESC[22mfilenameESC[24m] [ESC[1m-r ESC[4mESC[22msecondsESC[24m]
[ESC[1m-vESC[22m[ESC[1m+ESC[22m|ESC[1m-ESC[22m]ESC[4mvalESC[24m[ESC[1myESC[22m|EESC[1mmESC[22m|ESC[1mwESC[22m|ESC[1mdESC[22m|ESC[1mHESC[22m|ESC[1mMESC[22m|ESC[1mSESC[22m]] [ESC[1m+ESC[4mESC[22moutput_fmtESC[24m]
I've already discovered that I needed to add options to the LESS environment variable, but clearly they weren't enough. What I had was:
LESS="-a -F -i -j5 -M -X"
The options are:
But clearly that's not strong enough. With some man page reading, came up with
LESS="-a -F -i -j5 -M -R -s -X"
Here -R outputs the control characters. There's also an -r, which seems to make more sense, but it doesn't work for me the way it's described. -s suppresses multiple long lines, which sounded like a good idea.
And that works fine for nroff output. But git doesn't want to know. What I had was:
ESC[33mcommit e031614d59288e898e17a2bd3eeb4fb06b799798ESC[mESC[33m (ESC[mESC[1;36mHEAD -> ESC[mESC[1;32mmainESC[mESC[33m, ESC[mESC[1;31mfreebsd/mainESC[mESC[33m, ESC[mESC[1;31mfreebsd/HEADEESC[mESC[33m)ESC[m
Author: Alfonso S. Siciliano
Date: Sun Mar 6 22:44:17 2022 +0100
bsddialog: new utility for TUI dialogs and widgets
And afterwards I have:
|
|||||||||||||||||||||||
It sets the foreground colour, but not the background colour, making things illegible. WHEN will people realize that half a job isn't worth doing? More to the point, though, how do I get less not to display colours? It would seem that I want the -r option and not the -R option, but that doesn't work for me.
Then there's the issue with resizing xterms. When I do, bash doesn't want to know. Yes, there's a program resize that you're supposed to hand to it, but it no longer seems to work. There's a trick, of course: resize just outputs environment variables to feed into bash, so you need something like this to do it right:
eval `resize`
And finally there's bash (again!) on tiwi. bash has Emacs-like bindings, so Alt-f should move one word forward. But on tiwi, for some reason, it capitalizes the word on the way. And in general the codes are interpreted differently.
Why? It's not bash itself—the same thing happens with a bash from another system running on tiwi. It's also not X: it happens when the xterm is running on a different X server. And it's not the keyboard codes, which are the same.
Environment? Tried outputting the environments from teevee and tiwi and comparing them. There was a likely-looking one:
-XTERM_LOCALE=C.UTF-8
+XTERM_LOCALE=C
- is tiwi, + is teevee. OK, set XTERM_LOCALE to C. No change.
Why is this all so difficult?
| Wednesday, 9 March 2022 | Dereel | Images for 9 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More control character pain
|
Topic: technology, opinion | Link here |
Spent a considerable period of time today trying to corner the bash control code issue. Where does it come from? Yesterday's investigation didn't address all the possibilities. There were more environment variables which, though unlikely, could have had a bearing on the problem. Ended up changing them all to the versions on teevee. It didn't make any difference.
So what do I have?
What is there? I'm running out of options.
|
Multiple power failures
|
Topic: Stones Road house, general | Link here |
No fewer than 5 grid power failures today. None of them were long—between 2 and 12 seconds—but all would have caused a complete failure of all computers had there been no protection.
|
Air conditioner: progress?
|
Topic: Stones Road house, general, opinion | Link here |
Call from Tony at Atmos today, phone 0419 110 370, about the air conditioner. It's only been 6 weeks. Discussed the problem with him, and of course there was little he could do. But he did call Actron about the problem and came to the hopefully more reliable conclusion that it was probably an external sensor, something that wasn't immediately obvious from description of the status E25: “Ambient Temperature not updated“. OK, first step: replace the ambient temperature sensor.
Well, no, if we've identified the thing, and it doesn't happen very often, how about checking the contacts? It seems to be attached to this board, which is quite adequately documented:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The ambient temperature sensor is at bottom left. So power down the machine, wobble on the contacts, see if it happens again? Tony agreed.
By chance, identified another behaviour. Today, for the first time in weeks, I “soft” powered down the unit (at the wall controller), and when I turned it back on, it displayed E25 immediately. And that's not the first time. Does that help identify the issue? I don't see anything obvious.
| Thursday, 10 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Fake news?
|
Topic: politics, opinion | Link here |
Woken up at 6:00 this morning to hear the news: the Russians have totally destroyed a maternity hospital in Mariupol, corpses buried under the remains of the hospital.
Horror! I couldn't get back to sleep, and heard the same report at 7:00. The whole Russo-Ukrainian War is horrifying as it is, but this shows a lack of respect for the most basic rules of war. Clearly a war crime, and clearly one sanctioned by Putin, and thus a crime against humanity. I couldn't get over it.
Then I got up and read the news. Yes, reported by the ABC and Nine News, both Australian news sources. And elsewhere? Nothing.
It took a while to find more information, by which time the hospital had not been reduced to rubble, just seriously damaged. Reports of injuries, but no deaths (they came later, currently 3 after 24 hours).
Fake news? It seems that Volodymyr Zelenskyy tweeted something briefly that gave rise to the assumption. How easy it is to come to the wrong conclusion!
The Russians don't need that kind of excuse, of course: they attacked the hospital because it was used for war preparations. How can they expect people to believe anything? Do people in Russia believe this nonsense? And now they're claiming that the USA and Ukraine were developing biological weapons in Ukraine. Can we assume that they're going to find some of these non-existent weapons and accidentally unleash them?
When all this horror is over, the United Nations are going to have to reform and ensure that countries with such abhorrent governments have less power within the organization. At the very least the issue of veto power needs to be addressed.
|
More bash bug investigations
|
Topic: technology, opinion | Link here |
Followed on investigating yesterday's investigation of the bash problems on tiwi. So far I have practically eliminated bash itself, X, the keyboard drivers and my environment. What's left?
In each case, I start the bash in an xterm. What happens if I use ssh to access the system? Tried that:
=== grog@eureka (/dev/pts/34) ~ 1 -> ssh tiwi
...=== grog@tiwi (/dev/pts/5) ~ 1 -> ps aux
And how about that, after the output, I can edit the command line normally. Same xterm, same bash, same environment, just not started via the window manager.
Now clearly it's not the window manager itself, since it's been doing that job for decades, and it still does for other systems. Time to look at the invocation line (in ~/.fvwm/xterms-eureka:0.2). The entry is:
+ "tiwi" Exec ssh -A tiwi /usr/local/bin/xterm -name "xterm-lx" -bg BlanchedAlmond -s -sl 2048 -sb -ls -j -rw -display eureka:0.2 -geometry 100x65+53+0 -fn 10x20 -e /usr/local/bin/bash &
What in that could cause the problem? Tried starting it directly with the ssh invocation, and sure enough, I ran into the problem again. None of the options look obviously like the problem, but try this anyway:
=== grog@eureka (/dev/pts/34) ~ 4 -> ssh tiwi /usr/local/bin/xterm -display eureka:0.2
Still the same problem. Is it something to do with passing the keystrokes across the network?
|
Another power outage
|
Topic: Stones Road house, general | Link here |
Another grid power outage today, 9 seconds long and thus in some ways similar to yesterday's outages, but also different from what I normally experience. Is this some maintenance person who decides that the outage is too short to be worth reporting? It would be a very poor match for modern electronic equipment.
| Friday, 11 March 2022 | Dereel | Images for 11 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Civilizing git
|
Topic: technology, opinion | Link here |
Mail from Jashank Jeremy about my problems with how less handles gratuitous colour. Specifically, he pointed to how to stop git from misbehaving.
Yes, I knew that. I thought. But he pointed to a man page git-config, only 93 pages of printed output, clearly too little for them to include examples of how to just shut the bloody thing up. Jashank included that:
To set it for one invocation::$ git -c color.ui=no [... the remainder of the invocation ...]To persist it::
$ git config --global color.ui no
Now isn't that a lot of work to set what should be the default?
Jashank also pointed to the Emacs mode for git, which I should have noticed, since it was waiting just beneath my fingers. Instead of all this mess with the git program, just use the same old key bindings I used with svn! And how about that, though the log comes up in colour, it's readable:
|
|
|||||||||
The green bars are highlighted trailing spaces (from my Emacs config). It seems that the log includes messages about other files, but until proof of the contrary I'll assume that that's just sloppy committing.
|
Upgrading tiwi
|
Topic: technology, opinion | Link here |
Despite all the frustration, I'm making progress with my TV computers. And my disks on tiwi and teevee are filling up. Time to continue the stalled migration from teevee to tiwi. OK, check out the latest stable/13 tree, not made any easier by the difficulty of finding the branch abbreviations. And somehow I ended up with the tree in /usr/stable/13 instead of /usr/src.
OK, I can rename it. But git doesn't like that. It knows the name, and I had to stick to it. Nothing for it: a symlink to /usr/src.
3 hours of building clang later and it booted happily, so much so that I checked that it was really the new kernel:
=== root@tiwi (/dev/pts/2) ~ 2 -> uname -a
FreeBSD tiwi.lemis.com 14.0-CURRENT FreeBSD 14.0-CURRENT #0 main-n253735-832acea92fc8: Fri Mar 11 17:07:51 AEDT 2022 grog@tiwi.lemis.com:/usr/obj/usr/stable/13/amd64.amd64/sys/GENERIC amd64
Oh. What went wrong there? How do I find out what branch I have in the tree? I didn't find out. But a little investigation found a bug in my igitt script. Tomorrow I'll try to fix it, though it means going down to a previous release.
| Saturday, 12 March 2022 | Dereel | Images for 12 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Making friends
|
Topic: animals | Link here |
Seen this morning:
|
|||||||||||||||||||||||
For some reason Larissa is particularly friendly with Denzel, the Swifts' no longer so new dog (they have just received another one, Mandy, only 8 weeks old, whom I haven't seen yet). Lena is also interested, but not nearly as much as Larissa.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Dry omelette, next attempt
|
Topic: food and drink, opinion | Link here |
Another fake Phat Thai for breakfast this morning, requiring a “dry omelette”, one fried with no oil. My previous attempts, most recently last month, have still not quite been successful.
But then it occurred to me: why should they fill the pan? How about a halo omelette, with a hole in the middle where the previous attempts had dried out? Tried that. Still not quite what I expected:
|
|||||||||||||||||||||||
Still, it will be cut up, so who cares? And how about that, it cooked much more evenly, though there was still minor sticking at the edge, which didn't seem to make much difference:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I can live with that.
|
Larissa problems?
|
Topic: animals, opinion | Link here |
Larissa didn't want to come for a walk today. She was in the bedroom, and when I called her, she demonstratively went to her bed and lay down. She took a lot of persuasion to get up and come with us, and when she did, she stumbled a couple of times.
It wasn't like Lena's issues last month. Once she was outside, she was fine, and there were very specific places where she stumbled, noticeably at the corner from the bedroom hallway to the kitchen, where Leonid had had problems before. In each case, I think it was some kind of fear, maybe based on bad experiences in the past.
But Yvonne thought it might be simply that the floors are too slippery. Why now? Who knows? In any case, she put a whole lot of doormats down the main hallway, where Lara had stumbled the second time. I took a look at how she reacted. The first time she avoided and went around them, but after that she just walked straight over them. Clearly nothing to make life easier for her.
I put a rug round the corner between bedroom hall and kitchen. There's no way to avoid going over that, but also no reason to assume that it makes any difference for her.
|
tiwi “upgrade”, yet again
|
Topic: technology, opinion | Link here |
Finally managed to check out the correct branch of the FreeBSD repository today, and despite reservations it
built cleanly in less than 2½ hours, considerably less than the time it took
for -CURRENT main. Rebooted nicely,
then make installworld.
But next time I tried, I couldn't access the machine.
=== grog@eureka (/dev/pts/35) ~ 1 -> ssh tiwi
ssh_exchange_identification: Connection closed by remote host
Disconnected at Sat 12 Mar 2022 16:28:10 AEDT
Huh? I still had one ssh session open to tiwi, but I couldn't connect any more. ssh -v didn't help much (it never seems to). What's in /var/log/auth.log?
Mar 12 16:16:21 tiwi sshd[1273]: Accepted publickey for grog from 192.109.197.137 port 33094 ssh2: RSA SHA256:S7sZHLcY4dgw53/rF70vrScdPuGef3enHdJzuYA1WDo
Mar 12 16:28:10 tiwi sshd[54238]: fatal: recv_rexec_state: parse config: incomplete message
Huh? What about the other side? Nothing, of course, but then I saw the output of make installworld:
--------------------------------------------------------------
>>> Installing everything completed on Sat Mar 12 16:24:17 AEDT 2022
--------------------------------------------------------------
Scanning /usr/share/certs/blacklisted for certificates...
Scanning /usr/share/certs/trusted for certificates...
Skipping blacklisted certificate /usr/share/certs/trusted/Camerfirma_Chambers_of_Commerce_Root.pem (/etc/ssl/blacklisted/f90208f7.0)
...
ion_Authority_-_G3.pem (/etc/ssl/blacklisted/dc45b0bd.0)
Skipping blacklisted certificate /usr/share/certs/trusted/VeriSign_Universal_Root_Certification_Authority.pem (/etc/ssl/blacklisted/c01cdfa2.0)
Scanning /usr/local/share/certs for certificates...
https://accounts.google.com/AccountChooser?Email=groogled@gmail.com&continue=https://support.google.com/accounts/answer/112802?rfn%3D1646361360139%26anexp%3Dnret-fa
What's all that? I've never seen that before. Is it something new? In any case, none of the “blacklisted” certificates seemed relevant, but it seemed reasonable to assume that there was a a relationship. Try restarting sshd? Yes!
I'm still puzzled. What was happening here? Is it a bug?
|
Porc en croute
|
Topic: food and drink, opinion | Link here |
For this evening's meal Yvonne had thought of porc en croute, pork filet with spinach baked in puff pastry:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Time for me to relax? Well, no, she had neither oven temperature, cooking time nor end temperature. And that's an area where I have been searching for answers, as the cooking times page shows. The meat weighed 330 g, and I had guessed that, in pastry in a 200° fan-forced oven, it would take about 45 minutes to reach 80°, which I thought about right.
It only took 26 minutes:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But the results were good. Apart from the cooking time, nothing needs changing.
| Sunday, 13 March 2022 | Dereel | Images for 13 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Dog's birthday
|
Topic: animals | Link here |
Larissa and Lena are one year old today.
In fact, I got the date wrong. Their birthday was 16 March 2021.
Yvonne had the idea of a birthday present, as if they would understand the concept of birthdays:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Animal feed topics
|
Topic: animals, opinion | Link here |
Apart from birthdays, how much food should animals get? It's easy to overfeed them, and for some months we have been keeping an eye on both the dogs and Piccola. The results are surprising: until recently we were giving Piccola 50 g of food (roughly 40 g of canned food and 10 g of pellets) twice a day, and the dogs were getting 180 g of pellets and a chicken frame once a day.
A chicken frame has very little nutritional value, so basically our 35 kg dogs are getting 180 g of food a day, while our 3.5 kg cat is getting 100 g of food a day. And she's not fat. We've decided to give her 70 g twice a day. And Yvonne thought that the dogs should get more too. But today Larissa left a considerable quantity behind: she only ate 96 g. That's less than Piccola eats.
There's the question as to whether any of the animals are sick, of course, but Lara plays happily outside, and it's not the first time that she has left food behind, while Lena never does.
|
Crown history
|
Topic: history, multimedia, opinion | Link here |
Started watching a TV series today, Royals: Keeping the Crown. The interesting thing was that they had a number of original video clips from up to 120 years old, restored and colourized.
It screams “British”. The first episode, which I haven't finished watching, is about George V of the UK, Wilhelm II of Germany and Nikolai II of Russia, all cousins. But it emphasizes the difference between Wilhelm and the other two and paints Wilhelm in a very negative light.
That could be an opinion, but the way the commentary ran suggests a wise-after-the-event British viewpoint. What's definitely wrong is the suggestion that all three were grandchildren of Victoria: Nikolai was unrelated and also not a cousin of Wilhelm: he was a cousin of George on the maternal side, though his wife Alix von Hessen-Darmstadt was a granddaughter of Victoria.
Should I carry on watching? It seems very biased and is clearly not overly accurate. The real reason that I wanted to watch it was because of the old video clips, which were in fact quite interesting. Strangely, though, they all had a considerable amount of what appears to be digital noise. I wonder if it's deliberate.
But apart from that, even biased documentaries can contain useful information. I just need to watch with a sufficiently critical eye.
| Monday, 14 March 2022 | Dereel | Images for 14 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
First git commit
|
Topic: technology, opinion | Link here |
So finally I'm in a position to commit to the FreeBSD source tree again. Asked Peter Jeremy to review it (“OK, but why don't you use Phabricator?”) and hold my hand while I committed it.
First surprise: you don't just commit, you also push. I'm still trying to find out the distinction, but so far it looks like “push” is the git word for “commit”, and git “commit” doesn't actually affect the copy in the repository. But I needed to document the commit details in the “commit” stage, and the template I was presented didn't match what I should have got.
Peter pointed me to the git primer, which is what I had been using, but it seems that there's some parameter that should be global but which needs to be done on a much more frequent basis. Yes, I can do what he says, but I'd rather find out why, despite my (I thought) careful attention to detail, I missed this. I'll have to fight my way through the primer again.
|
The Russian army
|
Topic: history, opinion | Link here |
Watched the rest of the first episode of Royals: Keeping the Crown today, confirming my opinion of how Britain-centric the series is. But one scene was interesting. From the subtitles:
But thanks to the skill of Wilhelm's generals, Germany defeats Nicholas's troops, despite Russia having the biggest army in the world. Russia is able to mobilise a huge army, but this does not mean that it's a good army, or that it's effective. It is really badly trained. It is a conscript army, essentially consisting of under-equipped and underfed peasants that do not have good training, do not have good morale, and expect this war to be very short and want to go back to harvest their fields. To make things even worse, Nicholas takes full charge of the Russian Army and only takes military advice from God.
In passing, it's once again interesting that they use the names Wilhelm and Nicholas, and not, for example, William and Nikolai. But it's even more interesting to compare the situation then with the current situation with Russia's invasion of Ukraine.
|
Upgrading dereel
|
Topic: technology, opinion | Link here |
Now that I have git more or less working, I can upgrade some of my secondary systems. Today it was dereel, which is now effectively on the same release as tiwi:
=== grog@tiwi (/dev/pts/0) ~ 2 -> uname -a
FreeBSD tiwi.lemis.com 13.1-STABLE FreeBSD 13.1-STABLE #0 stable/13-n249989-b85d0d603c57: Sat Mar 12 13:23:39 AEDT 2022 grog@tiwi.lemis.com:/usr/obj/eureka/home/src/FreeBSD/tiwi/amd64.amd64/sys/GENERIC amd64=== grog@dereel (/dev/pts/2) ~ 3 -> uname -a
FreeBSD dereel 13.1-STABLE FreeBSD 13.1-STABLE #0 stable/13-n249995-80bab8aa7ed8: Mon Mar 14 13:20:09 AEDT 2022 grog@dereel:/usr/obj/eureka/home/src/FreeBSD/git/stable-13/amd64.amd64/sys/GENERIC amd64
Why does tiwi report itself as tiwi.lemis.com, while dereel is simply dereel? It's returned by hostname, but why?
So: do I experience the same strangenesses? Yes and no: after make installworld, once again I couldn't access dereel with ssh; first I needed to restart sshd. But I didn't have the issues with bash control characters. I wonder where that bug is hiding.
| Tuesday, 15 March 2022 | Dereel → Ballarat → Dereel | Images for 15 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Clematis progress
|
Topic: gardening | Link here |
It's autumn, but some flowers are only just coming into their own. The Clematis “Edo Murasaki” has survived Bryan Ross' mutilation and is looking almost normal:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Not so the “General Sikorski”, which seemed to struggle to stay alive. After the last feeble flowering, it died back, tried to grow again, but the incipient shoots died. I don't expect to see any more activity this year, and maybe never. What's the problem?
|
Haircut!
|
Topic: general, health, opinion | Link here |
My last haircut was in October. I have a haircut every 3 months, so the next one was due at the end of January. But the increase in COVID-19 numbers in December made it sound a good idea to put off the next haircut until the surge was over.
But somehow the surge didn't finish. In mid-February we thought it was on its way down, but then it started up again. My hair couldn't wait: into town today to get Catherine to cut it.
|
git, how I hate thee!
|
Topic: technology, opinion | Link here |
Yesterday's attempt to commit to the FreeBSD repository using git failed due to an incorrect template. Back to look at the git primer today, and it seems that the conflict was my fault: instead of using the script that I had carefully set up, I did it by hand and lost the checkin template.
But while reading the primer, ran into some strangenesses:
Clone the repository:
% git clone -o freebsd --config remote.freebsd.fetch='+refs/notes/*:refs/notes/*' https://git.freebsd.org/${repo}.gitThen you should have the official mirrors as your remote:
% git remote -v
freebsd https://git.freebsd.org/${repo}.git (fetch)
freebsd https://git.freebsd.org/${repo}.git (push)OK, so far, so good.
Configure the FreeBSD committer data:
The commit hook in repo.freebsd.org checks the "Commit" field matches the committer’s information in FreeBSD.org. The easiest way to get the suggested config is by executing /usr/local/bin/gen-gitconfig.sh script on freefall:
% gen-gitconfig.sh
[...]
% git config user.name (your name in gecos)
% git config user.email (your login)@FreeBSD.orgSet the push URL:
% git remote set-url --push freebsd git@gitrepo.freebsd.org:${repo}.gitWait a while, that contradicts what was written above!
Then you should have separated fetch and push URLs as the most efficient setup:
% git remote -v
freebsd https://git.freebsd.org/${repo}.git (fetch)
freebsd git@gitrepo.freebsd.org:${repo}.git (push)Why was this necessary?
Again, note that gitrepo.freebsd.org will be canonicalized to repo.freebsd.org in the future.
Install commit message template hook:
% fetch https://cgit.freebsd.org/src/plain/tools/tools/git/hooks/prepare-commit-msg -o .git/hooks
% chmod 755 .git/hooks/prepare-commit-msgThis was the step that I had missed. But why the chmod? fetch doesn't change the permissions (which were 644). That could be done at the source. But why? It's just text.
While looking at my script, I discovered this line:
git remote set-url --push freebsd ssh://git@gitrepo.freebsd.org/doc.git
Where did that come from? It's clearly intended for documentation checkins only, not including man pages. It was harmless, because it came before another set-url, but I can't imagine that I set it there. My guess is that it was accidentally in the primer and had since been removed. For the future, it's clearly a good idea to quote the entire instructions, as I have done above.
OK, now I understand a little more. Check the primer. Ah, simple! git commit date.1. And how about that, the correct template came up, and I was able to commit with no issues.
Ah, says Peter Jeremy, you're not done yet. Now you need to push. That's when the change actually makes it to the main repository. No, I don't have time now, need to go for a haircut. That can cause problems, says Peter. You need to do them before there's a change to the main repository. No worries, this file hasn't been updated for months, and if there are any issues, at least I can test the recovery mechanisms.
Back after my haircut and tried to push:
|
|
|||||||||
Ugh! Who can read that? Spent some time discussing on IRC, during which I discovered that there is absolutely no way to get this message in legible form. All the no-color settings fail. So does setting a different terminal type, no terminal type or an invalid terminal type, and resetting the TERMCAP environment variable:
|
|||||||||||||||||||||||
That's at least two bugs: firstly, ignoring the no-color settings, and secondly selecting what appear to be hard-coded colour settings.
OK, it's reasonable to assume that this is the situation that Peter Jeremy warned me of, hut which isn't mentioned in the primer (which doesn't even mention that the push should follow commit). Asked Peter what he does:
My approach was "zfs rollback ..." and start again.
Clearly the documentation is of very little help.
OK, back to the yellow on beige message. Fortunately we have cut and paste:
=== root@dereel (/dev/pts/1) /usr/src-main/bin/date 32 -> git push
To gitrepo.freebsd.org:src.git
! [rejected] main -> main (fetch first)
error: failed to push some refs to 'gitrepo.freebsd.org:src.git'
hint: Updates were rejected because the remote contains work that you do
hint: not have locally. This is usually caused by another repository pushing
hint: to the same ref. You may want to first integrate the remote changes
hint: (e.g., 'git pull ...') before pushing again.
hint: See the 'Note about fast-forwards' in 'git push --help' for details.
That's worth following up, but I had no time any more. I still think that it's worth investigating, since the current situation is an unmitigated disaster.
|
CNAV again
|
Topic: general, opinion | Link here |
Another letter to CNAV today, with Yet Another attestation d'existence (the fifth if I'm counting correctly), this time with avis de réception, apparently not well known in English-speaking countries. But at least this time we will get some proof of what we're doing.
And a few days ago received another message from Michel Lustig telling us how to complain:
You may use this link "Faire appel au médiateur de votre caisse régionale"
Then, in "Comment ça marche ?", "Etape 2 : saisissez le médiateur en complétant ce formulaire" You have to fill in a form.
Then send the pdf file to mediateur-retraite@cnav.fr
Hopefully we won't need to do that, but it's good to know the details, even if we really wanted to complain rather than to appeal.
| Wednesday, 16 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
What is standard time?
|
Topic: language, general, opinion | Link here |
Decades ago, in Porting UNIX Software, I wrote:
UNIX timekeeping is an untidy area, made more confusing by national and international laws and customs.
blame tells me that that was committed on 16 December 1994. Things haven't improved. Currently the US Americans are up in arms about “permanent daylight saving time”, really an abolition of DST and a shift of the time zone offset. In the USA they use the term “standard time”. Where else? When I was a lad our time was referred to as “Australian Eastern Time”. But now it's “Australian Eastern Standard Time”, at least in the tzdata database. OED writes:
3. f. standard time, a standard system of reckoning time adopted throughout a country or region, now based on the time zone in which it is situated;
But none of the references suggest that the term has any official meaning in the United Kingdom, where the “standard” time is GMT. Is “standard time” an Americanism? I asked Quora and got more interesting answers than I expected. This one says:
Yes. If the definition of Americanism is a word or phrase in the English language that was first coined in the United States.
Webster’s New World College Dictionary of American English puts a star beside all the words and phrases that were coined in the USA, and it has a star beside standard time.
The definition it gives is:
1 the time in any of the 24 time zones, each an hour apart, into which the earth is divided: it is based on distance east or west of Greenwich, England,...
That's interesting, but how can an authoritative document claim that there are only 24 time zones? That's worth another rant, but it's not the first time where I've seen inaccurate claims in Webster.
And of course there were answers that were just plain wrong, like
Sometimes Military planners will say standard time instead of Zulu time, which is Greenwich Standard time. It is never used when talking about local time.
But it made clear that my question was wrong: I should have asked about its use specifically to distinguish from Daylight Saving Time.
But to get an idea about how inaccurate time information is, we just need to go to the Australian government, which states:
With a land mass close to 7.7 million square kilometres, Australia is the world's sixth largest country and is divided into three [my emphasis] separate time zones.
Time zone State or territory City Australian Eastern Standard Time (AEST) Queensland, New South Wales (except Broken Hill), Victoria, Tasmania, Australian Capital Territory Brisbane, Sydney, Melbourne, Hobart, Canberra UTC+10:00 Australian Central Standard Time (ACST) South Australia, Northern Territory and the town of Broken Hill in western New South Wales Adelaide, Darwin, Broken Hill UTC+09:30 Australian Western Standard Time (AWST) Western Australia Perth UTC+08:00 Norfolk Island Standard Time (NFT) Norfolk Island Kingston UTC+11:00 Christmas Island Standard Time (CXT) Christmas Island UTC+07:00 Cocos (Keeling) Islands Standard Time (CCT) Cocos (Keeling) Islands - UTC+06:30
|
Phone answering pain, yet again
|
Topic: technology, general, opinion | Link here |
Called Yvonne (who was shopping in Ballarat) on the phone today. No answer.
I've seen this too often. Why can't she do something as simple as answer the phone? She called back. My phone rang, but nothing appeared on the screen. “Swipe” down from top right. No answer indication. Hoist with my own petard!
Finally she got through to me, but clearly something is seriously wrong here. I can't imagine that it's both of us. Could it be Xiaomi or MIUI? Maybe, but they're a very widely distributed make. Other possibilities include:
It still doesn't make any sense to me, but if anything, things are getting worse. Why am I the only person complaining about this?
|
The bombing will continue until we're friends again
|
Topic: politics, opinion | Link here |
Clearly the Russian invasion of Ukraine is not going as the Russians have planned. Now they're doing what they can: attacking civilians.
What are they trying to achieve? What they're doing amounts to war crimes, and it's hardly conducive to the Ukrainians making friends with the Russians the way they were before Russian aggression started. This article preparing for defeat, is interesting. I just hope they do it without too much more loss of life.
|
Calculating exposures
|
Topic: photography, history | Link here |
58 years ago, when I got my first camera, I had no way of measuring light intensity. Instead I used various guides based on descriptions of the lighting and the time of day. It was a pain, and when I got my first SLR I ensured that it had a (clip-on) light meter. But the bad old days are interesting, and I found this link to a long history of the things, starting in 1840 with Daguerreotypes (midday, bright sun, 5 minute exposure). I should put it somewhere, but for the time being I have a note here.
|
Another grid power failure
|
Topic: Stones Road house, general | Link here |
Lately we've been having a kind of grid power failure that I haven't seen much before: longer than instantaneous (1 second), but less than a minute. Today it was 3 seconds.
|
git?
|
Topic: technology, opinion | Link here |
It's clear that I need to take several steps back to really understand git, or at least what was going on in the minds of the people who wrote it. Received another mail message from Jashank Jeremy in the evening, basically confirming my suspicions, and also offering suggestions that I will follow up on Some Time Soon. This xkcd cartoon seems particularly appropriate:
|
Lara in pain
|
Topic: animals, opinion | Link here |
Just before going to bed, discovered Larissa in pain. Why? We've had our concerns about her ever since August last year, when she suffered from what was diagnosed as hypertrophic osteodystrophy. Is this related? I don't think so. An hour early she was trying to play with Lena in the lounge room, and we put them out into the dog run, which they have now transformed into an obstacle course:
|
|||||||||||||||||||||||
She was in the right-hand (deepest) hole when I called her out, and my guess is that she somehow twisted her leg getting out.
In passing, the dogs seem to be getting overweight. I've been giving them 200 g of pellets per day lately, clearly too much. Back to 180.
| Thursday, 17 March 2022 | Dereel → Bannockburn → Dereel | Images for 17 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Micro pok choi
|
Topic: food and drink, opinion | Link here |
Yvonne went shopping yesterday, and she can never resist cheap vegetables. She came back with a bag of carrots, radishes, some kind of broccoli and these vegetables, shown next to normal pok choi on the right.
|
|||||||||||||||||||||||
What are they? In principle they're pok choi, just half sized. They also have the same problem as normal pok choi that they contain lots of dirt between the stalks and at the base:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
They tasted OK, just not very different from pok choi.
|
Smoke in the kitchen
|
Topic: photography, food and drink, opinion | Link here |
KL Hokkien mee for breakfast today. That involves frying squid at a very high temperature, producing lots of smoke:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And yes, that was a lot of smoke. But it doesn't look like it in the photo. Here crops of two photos:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
How do I make it more obvious?
|
More USB camera attachment issues
|
Topic: technology, photography | Link here |
Took photos with the Olympus OM-D E-M1 and the OM-D E-M1 Mark II today. Read them in to dereel, which is conveniently located. The photos from the E-M1 Mark I came in fine. No photos from the Mark II.
Cable problems? Checked the connections, plugged into the other USB port (where the Mark I had been). Same thing:
Mar 17 11:28:43 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_TEST_UNIT_READY set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:43 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:43 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_SYNC_CACHE set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:44 dereel syslogd: last message repeated 1 times
Mar 17 11:28:44 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:44 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_TEST_UNIT_READY set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:44 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_START_STOP set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:28:44 dereel kernel: usb_alloc_device: Failure selecting configuration index 0:USB_ERR_STALLED, port 20, addr 6 (ignored)
Mar 17 11:28:44 dereel kernel: ugen0.5: <OLYMPUS E-M1MarkII> at usbus0
What's that? It should have identified the camera and created two mass storage devices, one for each SD card.
OK, when did I last connect this camera to this computer (using the same cable, as it happened)?
Mar 14 09:35:46 dereel kernel: ugen0.3: <OLYMPUS E-M1MarkII> at usbus0
Mar 14 09:35:46 dereel kernel: umass0 on uhub2
Mar 14 09:35:46 dereel kernel: umass0: <OLYMPUS E-M1MarkII, class 0/0, rev 3.10/1.00, addr 31> on usbus0
Mar 14 09:35:46 dereel kernel: umass0: SCSI over Bulk-Only; quirks = 0xc000
Mar 14 09:35:46 dereel kernel: umass0:3:0: Attached to scbus3
Mar 14 09:35:46 dereel kernel: da0 at umass-sim0 bus 0 scbus3 target 0 lun 0
Mar 14 09:35:46 dereel kernel: da0: <OLYMPUS E-M1MarkII 1.00> Removable Direct Access SCSI-2 device
Mar 14 09:35:46 dereel kernel: da0: Serial Number BHUA20626
Mar 14 09:35:46 dereel kernel: da0: 400.000MB/s transfers
Mar 14 09:35:46 dereel kernel: da0: 29935MB (61306880 512 byte sectors)
Mar 14 09:35:46 dereel kernel: da0: quirks=0x3<NO_SYNC_CACHE,NO_6_BYTE>
Mar 14 09:35:46 dereel kernel: da1 at umass-sim0 bus 0 scbus3 target 0 lun 1
Mar 14 09:35:46 dereel kernel: da1: <OLYMPUS E-M1MarkII 1.00> Removable Direct Access SCSI-2 device
Mar 14 09:35:46 dereel kernel: da1: Serial Number BHUA20626
Mar 14 09:35:46 dereel kernel: da1: 400.000MB/s transfers
Mar 14 09:35:46 dereel kernel: da1: 7600MB (15564800 512 byte sectors)
Mar 14 09:35:46 dereel kernel: da1: quirks=0x3<NO_SYNC_CACHE,NO_6_BYTE>
Now that looks completely different. What happened?
=== grog@dereel (/dev/pts/2) ~ 13 -> uname -a
FreeBSD dereel 13.1-STABLE FreeBSD 13.1-STABLE #0 stable/13-n249995-80bab8aa7ed8: Mon Mar 14 13:20:09 AEDT 2022 grog@dereel:/usr/obj/eureka/home/src/FreeBSD/git/stable-13/amd64.amd64/sys/GENERIC amd64
Now that's a smoking gun. I installed a new kernel between the two attempts. What do things look like elsewhere?
Mar 17 11:37:14 tiwi kernel: usb_msc_auto_quirk: UQ_MSC_NO_TEST_UNIT_READY set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:37:14 tiwi kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:37:14 tiwi kernel: usb_msc_auto_quirk: UQ_MSC_NO_SYNC_CACHE set for USB mass storage device OLYMPUS E-M1MarkII (0x07b4:0x012e)
Mar 17 11:37:15 tiwi kernel: ugen1.3: <OLYMPUS E-M1MarkII> at usbus1
Mar 17 11:37:15 tiwi kernel: umass0 on uhub2
Mar 17 11:37:15 tiwi kernel: umass0: <OLYMPUS E-M1MarkII, class 0/0, rev 2.10/1.00, addr 3> on usbus1
Mar 17 11:37:15 tiwi kernel: umass0: SCSI over Bulk-Only; quirks = 0xc001
Mar 17 11:37:15 tiwi kernel: umass0:3:0: Attached to scbus3
Mar 17 11:37:15 tiwi kernel: da0 at umass-sim0 bus 0 scbus3 target 0 lun 0
Mar 17 11:37:15 tiwi kernel: da0: <OLYMPUS E-M1MarkII 1.00> Removable Direct Access SCSI-2 device
Mar 17 11:37:15 tiwi kernel: da0: Serial Number BHUA20626
Mar 17 11:37:15 tiwi kernel: da0: 40.000MB/s transfers
Mar 17 11:37:15 tiwi kernel: da0: 29935MB (61306880 512 byte sectors)
Mar 17 11:37:15 tiwi kernel: da0: quirks=0x3<NO_SYNC_CACHE,NO_6_BYTE>
Mar 17 11:37:15 tiwi kernel: da1 at umass-sim0 bus 0 scbus3 target 0 lun 1
Mar 17 11:37:15 tiwi kernel: da1: <OLYMPUS E-M1MarkII 1.00> Removable Direct Access SCSI-2 device
Mar 17 11:37:15 tiwi kernel: da1: Serial Number BHUA20626
Mar 17 11:37:15 tiwi kernel: da1: 40.000MB/s transfers
Mar 17 11:37:15 tiwi kernel: da1: 7600MB (15564800 512 byte sectors)
Mar 17 11:37:15 tiwi kernel: da1: quirks=0x3<NO_SYNC_CACHE,NO_6_BYTE>
tiwi's kernel is a couple of days older than dereel's. Here we have the “auto_quirks”, but it finds the device, as dereel does for the E-M1 Mark I:
Mar 17 11:29:12 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_GETMAXLUN set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:12 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_TEST_UNIT_READY set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:12 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:12 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_SYNC_CACHE set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:14 dereel syslogd: last message repeated 1 times
Mar 17 11:29:14 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_PREVENT_ALLOW set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:14 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_TEST_UNIT_READY set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:14 dereel kernel: usb_msc_auto_quirk: UQ_MSC_NO_START_STOP set for USB mass storage device OLYMPUS E-M1 (0x07b4:0x012e)
Mar 17 11:29:14 dereel kernel: ugen0.4: <OLYMPUS E-M1> at usbus0
Mar 17 11:29:14 dereel kernel: umass0 on uhub1
Mar 17 11:29:14 dereel kernel: umass0: <OLYMPUS E-M1, class 0/0, rev 2.00/1.00, addr 7> on usbus0
Mar 17 11:29:14 dereel kernel: umass0: SCSI over Bulk-Only; quirks = 0xc105
Mar 17 11:29:14 dereel kernel: umass0:3:0: Attached to scbus3
Mar 17 11:29:14 dereel kernel: da0 at umass-sim0 bus 0 scbus3 target 0 lun 0
Mar 17 11:29:14 dereel kernel: da0: <OLYMPUS E-M1 1.00> Removable Direct Access SCSI-2 device
Mar 17 11:29:14 dereel kernel: da0: Serial Number BHP228512
Mar 17 11:29:14 dereel kernel: da0: 40.000MB/s transfers
Mar 17 11:29:14 dereel kernel: da0: 30176MB (61800448 512 byte sectors)
Mar 17 11:29:14 dereel kernel: da0: quirks=0x3<NO_SYNC_CACHE,NO_6_BYTE>
Mar 17 11:29:14 dereel kernel: (da0:umass-sim0:0:0:0): READ(10). CDB: 28 00 03 ae ff ff 00 00 01 00
Mar 17 11:29:14 dereel kernel: (da0:umass-sim0:0:0:0): CAM status: SCSI Status Error
Mar 17 11:29:14 dereel kernel: (da0:umass-sim0:0:0:0): SCSI status: Check Condition
Mar 17 11:29:14 dereel kernel: (da0:umass-sim0:0:0:0): SCSI sense: NOT READY asc:3a,0 (Medium not present)
Mar 17 11:29:14 dereel kernel: (da0:umass-sim0:0:0:0): Error 6, Unretryable error
So it seems that some change has caused the USB subsystem to find problems where none were before. I particularly like the way it claims that the medium is not present after reporting the size. And yes, teevee, running a 2 year old 12.1 system, had no difficulty with either of them, so it's not the cable.
What do I do? First, I suppose, upgrade the kernel again and see if it's still there. But this shouldn't happen on a STABLE kernel.
|
To the vet again
|
Topic: animals | Link here |
Larissa still seems to have some problems walking. Yvonne was worried enough that we set off to Bannockburn to have her looked at. Zoe came to the conclusion that she had hurt her left knee (it seems that dogs only have knees on their hind legs; presumably the front ones are elbows), and prescribed Meloxicam for 5 days, during which there is to be no running around. Lara and Lena will love that.
While we were there, we weighed them and talked about feeding. Lara now weighs 31.14 kg and Lena 32.44 kg. That's 0.24 kg more for Lara and 0.41 kg less for Lena than 3 weeks ago. But somehow Lara looks overweight, though she also eats less. So for the time being we'll give her 180 g of pellets per day, and Lena will get 200.
|
More Google Maps fun
|
Topic: technology, opinion | Link here |
There are two vets in Bannockburn: the Bannockburn Veterinary Clinic and the Golden Plains Animal Hospital, apparently without a web site. We wanted to go to the Bannockburn Veterinary Clinic, so I said to my phone “Hey Google, take me to the Bannockburn Vet Clinic”. Yes, no problem. But when we got to Bannockburn, it tried to take me to the animal hospital.
That wasn't a big issue, since we don't need guidance to get there. But how do you tell Google that it's wrong? You can enter feedback after arriving, but it didn't think we had arrived. And to make matters worse, when I tried it again later, it worked correctly.
| Friday, 18 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Heart problems?
|
Topic: health, opinion | Link here |
Woke up in the middle of the night with palpitations that took over 30 minutes to go away. Not for the first time, but the older I get, the more attention I pay to what I previously mainly ignored.
What is it? I'm sure that I spoke to a doctor about similar issues a few years back, but I forget what he said. Maybe it was Neil Philips, then at the Eureka Medical Centre, who once told me “I don't test for things that I don't know how to handle”. That would also put it at at least 11 years ago.
After getting up, revived my ancient blood pressure sphygmomanometer, which gave me:
| Time | Systolic | Distolic | Pulse | |||
| 8:50 | 119 | 81 | 106 | |||
| 9:20 | 111 | 74 | 100 | |||
| 12:15 | 111 | 66 | 86 | |||
| 12:19 | 110 | 65 | 83 | |||
The blood pressures look fine, but why the high pulse? Still, it dropped in the course of the morning; something to talk to Paul Smith about next time I see him.
|
Noodle cooking times, revisited
|
Topic: food and drink, opinion | Link here |
Phở for breakfast again today. Last time I had found some unmarked rice noodles and established that they only needed soaking in cold water for 5 minutes, and marked it on the package. Today I used the rest of them.
But either I confused them with something else, or they had changed completely between the two occasions. Today they were still relatively firm after 25 minutes. The hot phở softened them sufficiently, but what a surprise that noodles that look just the same can take such different times to soften.
One thing that I did establish: last time I used 40 g of paste for 360 g of water, and noted that next time I should use 48 g. In the end I used 40 g and 300 g of water, the same proportion, and despite my reservations it seemed to be right.
|
Dog feeding again
|
Topic: animals | Link here |
Since yesterday we've been feeding Lena 200 g of pellets and Larissa 180 g. The difference seems to be appropriate: today Lena ate her pellets as always, but Lara left 65 g behind: only 115 g, actually less than Piccola ate. And that not for the first time.
| Saturday, 19 March 2022 | Dereel | Images for 19 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More health issues?
|
Topic: health | Link here |
Yesterday's mild health concerns didn't repeat themselves today, though Juha Kupiainen expressed concern. I don't see that there's anything special about my health at the moment. It's only that I have decided to make notes of it in my diary; after all, that's what the diary is for.
It does seem worthwhile, though, to keep a log of my blood pressure readings. In that connection, it seems that yesterday's readings were a little lower than normal.
|
30 years of fun
|
Topic: technology, history, opinion | Link here |
Read in my calendar today:
Mar 19 Greg installs BSD/386 0.3.1, 1992
Those were the days. My copy of BSD/386 was “beta” (thus the < 0 release number), and in those days I took beta releases seriously; it was less than 3 weeks since I had left Tandem, and at Tandem people expected reports, which I dutifully provided.
How times have changed!
|
USB issues again
|
Topic: technology, photography, opinion | Link here |
Reply from Hans Petter Selasky, the FreeBSD USB man, about my USB issues with the latest version of FreeBSD.
You should add the quirks that appear in dmesg to: /usr/src/sys/dev/usb/quirk/usb_quirk.c
You can test the quirk by adding:
hw.usb.quirk.0="0x07b4 0x012e 0 0xffff UQ_MSC_NO_TEST_UNIT_READY,UQ_MSC_NO_PREVENT_ALLOW,UQ_MSC_NO_SYNC_CACHE,UQ_MSC_NO_TEST_UNIT_READY,UQ_MSC_NO_START_STOP"To /boot/loader.conf
Also you can use usbconfig's add_quirk command to test these quirks for your device.
Three different workarounds. The first clearly requires rebuilding the system, the second rebooting, and the third RTFM. And that gave me ample opportunity to think about man page updates. If I get it right, I need to specify not the vendor and product codes, but the USB address to which the device is currently connected. And it seems that I can only specify one quirk at a time.
Ultimately I had, with the Olympus OM-D E-M1 Mark II connected as ugen0.4:
usbconfig -d ugen0.4 add_quirk UQ_MSC_NO_TEST_UNIT_READY
usbconfig -d ugen0.4 add_quirk UQ_MSC_NO_PREVENT_ALLOW
usbconfig -d ugen0.4 add_quirk UQ_MSC_NO_SYNC_CACHE
usbconfig -d ugen0.4 add_quirk UQ_MSC_NO_START_STOP
And that did the job. While I was at it, I checked out my other cameras. I have at least 4 old non-Olympus cameras with USB connection, but without exception they want some strange USB cable that I couldn't be bothered to look for. I also have no fewer than 9 Olympus cameras with only 3 USB cables, and only one of which is proprietary. The results of connecting them were:
| Camera | Product code | Known quirks | Quirk | |||
| E-1 | 0x102 | 0xc301 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| E-300 | 0x118 | 0xc101 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| E-330 | 0x118 | 0xc105 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| UQ_MSC_NO_START_STOP | ||||||
| E-30 | 0x118 | 0xc101 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| E-PM1 | 0x12c | 0xc105 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| UQ_MSC_NO_START_STOP | ||||||
| E-PM2 | 0x12e | 0xc005 | none | |||
| E-M1 | 0x12e | 0xc105 | UQ_MSC_NO_GETMAXLUN | |||
| UQ_MSC_NO_TEST_UNIT_READY | ||||||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| UQ_MSC_NO_START_STOP | ||||||
| E-M1 Mark II | 0x12e | 0xc005 | UQ_MSC_NO_TEST_UNIT_READY | |||
| UQ_MSC_NO_PREVENT_ALLOW | ||||||
| UQ_MSC_NO_SYNC_CACHE | ||||||
| UQ_MSC_NO_START_STOP | ||||||
| E-M5 Mark III | 0x12e | 0xc005 | none | |||
That list shows a number of surprises:
But the bottom line is: why do we need all this? It worked fine before. Sure, I can fix for the cameras I have, but how many other people can do it? And I only found out by contacting the developer.
| Sunday, 20 March 2022 | Dereel | Images for 20 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
The last Begonia
|
Topic: gardening, opinion | Link here |
It's been round 10 years since I was really active in the garden, and since then I've been mainly watching my plants die.
Today we needed to replace a painting in the corner of the dining room behind our last Begonia, which we bought over 11 years ago. It's a low-light plant, and it certainly didn't get much, but it grew well, and I was continually cutting it back. It now looks a mess, so for the time being I've put it outside:
|
|||||||||||||||||||||||
The original stems have long turned to wood:
|
|||||||||||||||||||||||
So it looks like time for replanting.
Another flower that just popped up is a bit of a surprise:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Somehow I have a recollection that they're a very wide-leafed bulb, with a name something like “elephant leaf”, but I can't find any reference, just that I found them in the same place this time last year:
|
|||||||||||||||||||||||
| Monday, 21 March 2022 | Dereel | Images for 21 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Fixing Olympus quirks
|
Topic: technology, opinion | Link here |
More looking at the camera USB issues. It's been a while since I've messed around in the FreeBSD kernel, but it was less work than I expected. In this particular case, I needed to update /usr/src/sys/dev/usb/quirk/usb_quirk.c with the quirks, and also /usr/src/sys/dev/usb/usbdevs to define the devices.
A typical entry looks like:
USB_QUIRK(OLYMPUS, E_M1MarkII, 0x0000, 0xffff, UQ_MSC_NO_GETMAXLUN,
UQ_MSC_NO_TEST_UNIT_READY, UQ_MSC_NO_PREVENT_ALLOW,
UQ_MSC_NO_SYNC_CACHE),
That could almost be generated from the dmesg output. But of course I needed to define E_M1MarkII, and that's why I needed to change usbdevs:
product OLYMPUS E_M1MarkII 0x012e
Problem: Olympus is too stingy to allocate a new product code for every product code, as I noted on Sunday. The Olympus OM-D E-M5 Mark III has the same product code as the Olympus OM-D E-M1 Mark II:
product OLYMPUS E_M5MarkIII 0x012e
But it has no quirks. What happens if I apply them anyway? I'll just have to try them.
OK, git pull, apply the patches, build, reboot, test. All went smoothly, and now the cameras connect with no complaints, and the E-M1 Mark II now also transfers properly. And the E-M5 Mark III? Works. But somehow it seems tacky.
| Tuesday, 22 March 2022 | Dereel | Images for 22 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
37 years of marriage!
|
Topic: history, general | Link here |
37 years ago today I got married to Yvonne Ködderitzsch:
|
|||||||||||||||||||||||
Not a round number, but a milestone all the same: over half our lives ago!
|
git, how I hate thee!
|
Topic: technology, opinion | Link here |
In preparation for committing the USB changes, I need to have an up-to-date CURRENT source tree. OK, I can do that:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main 21 -> Log igitt up
===== Tue 22 Mar 2022 12:56:52 AEDT on dereel: igitt up
This seems to be a src clone
From https://git.freebsd.org/src
8e72f458c6d3..fd6ca665d206 main -> freebsd/main
210991b1f28b..b8ae329db949 releng/13.0 -> freebsd/releng/13.0
fa67c45842bb..928e649f6818 releng/13.1 -> freebsd/releng/13.1
8342c11f27eb..931c75bf01d6 stable/11 -> freebsd/stable/11
6fa8af618475..bf8cb490e27d stable/12 -> freebsd/stable/12
9f600a260a73..812638797d59 stable/13 -> freebsd/stable/13
error: Your local changes to the following files would be overwritten by merge:
contrib/libcxxrt/atomic.h
contrib/libcxxrt/exception.cc
contrib/libcxxrt/guard.cc
contrib/libcxxrt/memory.cc
lib/libcasper/services/cap_fileargs/cap_fileargs.c
lib/libkvm/kvm_proc.c
lib/libmixer/mixer.3
sbin/sysctl/sysctl.8
sys/compat/linuxkpi/common/include/linux/pci.h
sys/compat/linuxkpi/common/src/linux_pci.c
sys/kern/kern_tc.c
sys/kern/kern_time.c
sys/kern/kern_tslog.c
sys/kern/kern_umtx.c
sys/kern/sysv_shm.c
sys/powerpc/mpc85xx/pci_mpc85xx.c
usr.sbin/mixer/mixer.8
usr.sbin/mixer/mixer.c
Please commit your changes or stash them before you merge.
Aborting
What kind of nonsense is that? I didn't change anything in that tree, and those files have nothing to do with what I'm working on. OK, as xkcd says:
But why is everything to do with git so frustrating?
|
VirtualBox again
|
Topic: technology, photography, opinion | Link here |
In preparation to commit the quirk entries for Olympus cameras, I need to build a FreeBSD kernel against CURRENT main.
But I don't have a system running CURRENT any more.
OK, that's what virtual machines are for. Fire up VirtualBox and create a virtual machine with a recent CURRENT kernel and build the kernel with my changes.
And once again I ran into issues with networking. The workarounds I established last time didn't work, at least not the way I described them.
The issue seems to be which klds to load. The important one is ng_ether, which in fact has nothing directly to do with VirtualBox. But just loading it doesn't seem to be enough; today it didn't work until I stopped VirtualBox, unloaded the kernel modules and loaded them again. Afterwards I had:
=== root@eureka (/dev/pts/10) /src/ISOs/FreeBSD 85 -> kldload vboxdrv=== root@eureka (/dev/pts/10) /src/ISOs/FreeBSD 87 -> kldload vboxnetflt=== root@eureka (/dev/pts/10) /src/ISOs/FreeBSD 88 -> kldload vboxnetadp=== root@eureka (/dev/pts/10) /src/ISOs/FreeBSD 89 -> kldstat
Id Refs Address Size Name
1 32 0xffffffff80200000 179cc10 kernel
3 2 0xffffffff81a17000 ad1c0 linux.ko
4 1 0xffffffff81ac5000 f08f98 nvidia.ko
6 3 0xffffffff829d6000 159a0 netgraph.ko
7 1 0xffffffff82c11000 56df fdescfs.ko
8 1 0xffffffff82c17000 359d ums.ko
9 1 0xffffffff82c1b000 2ba9 uhid.ko
10 2 0xffffffff82c1e000 114ed ipfw.ko
11 1 0xffffffff82c30000 4cba ipdivert.ko
12 1 0xffffffff82c35000 412b ng_ether.ko
13 3 0xffffffff82c3a000 4e436 vboxdrv.ko
14 2 0xffffffff82c89000 2a02 vboxnetflt.ko
15 1 0xffffffff82c8c000 3f64 vboxnetadp.ko=== root@eureka (/dev/pts/10) /src/ISOs/FreeBSD 90 ->
Do I need both vboxnetflt and vboxnetadp? To be checked next time.
The other, simpler thing is: when creating a virtual machine, give it a bridged Ethernet adapter connected to em0. The name is of the interface on the host machine, not on the guest machine.
And then I was able to install CURRENT, point it to the repository on eureka, and start a build. It ran all day.
|
Garden flowers in early autumn
|
Topic: gardening, opinion | Link here |
It's the autumnal equinox, almost, and time to take the monthly garden photos. It was really yesterday, but the light wasn't good (too sunny), so I put it off until today. Atypically, it was even foggy:
|
|||||||||||||||||||||||
The combination of better irrigation and relatively cool weather means that the trees in the east garden are looking better:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The Schinus molle that looked as if it were dying seems to be holding its own, though it's unclear whether it's getting any better:
|
|||||||||||||||||||||||
It's not clear what is going to happen to the Ginkgo:
|
|||||||||||||||||||||||
It has just been transplanted, so the “autumn” look is not overly surprising. It has fought its way through worse conditions for nearly 7 years now, so I still have hope.
The plants mutilated by Bryan Ross have mainly come back. The Clematis “Edo Murasaki” is flowering nicely, almost as good as in the first year:
|
|||||||||||||||||||||||
But the other one, the “General Sikorski”, seems to be dead. We'll see if we can resurrect it in the spring.
Apart from that, the “Vietnamese Mint”, cut to the ground, is now almost as big as it was a year ago:
|
|||||||||||||||||||||||
The Alyogyne huegelii has come back, but it isn't flowering. It's not clear whether that's Bryan's fault or whether it's the competition from the cannas around it.
The abutilons that we propagated last year are settling in, and the one I planted in the ground is now looking happier and is producing flowers:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And the almost-dead Buddleja x weyeriana that nearly died, and which I repotted and propagated last month, are now coming back with a vengeance. Here the main plant last month and now:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I'm hoping that this one will still flower this year. Even the twigs are doing much better:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The lime tree to the north of the house has gradually been improving—not before time—and now some of the fruit are getting quite large:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Even the second one, relatively new, has at least one fruit:
|
|||||||||||||||||||||||
And our volunteer tomato is bearing fruit. We ate the first two ripe ones on 21 March before I thought of taking a photo:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The roses are not doing overly well. Despite best efforts, the rose bed is filling up with weeds, and roses like “M. Tillier” are looking quite unhappy:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Still, there are some, like the “Gruß an Aachen” and the climbing Icebergs, that defy all that:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And to my surprise, the Trachelospermum jasminoides is still alive and (barely) flowering:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Hopefully things will improve now that the irrigation is more or less in shape.
One unexpected casualty was the bridge across our “river”, only 4 years old:
|
|||||||||||||||||||||||
| Wednesday, 23 March 2022 | Dereel | Images for 23 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Mulch!
|
Topic: gardening, opinion | Link here |
It has become clear that we need to mulch the garden beds, and somehow every time Yvonne finds somebody trimming trees and prepared to deliver mulch, something goes wrong. Nothing for it: today's the day to order some mulch and pay real money for it.
And the Yvonne discovered some people trimming trees two doors down, the other side of the Marriotts:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, they were prepared to deliver, even for free:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And why the mulching? The property is the southernmost of Garry Marriott's parcel of three. He lives in the middle one, we bought (and live on) the northern one, and the other is currently unused. Has he sold it? No, it seems that it's Powercor maintenance for the same line that has caused us so much pain with the covered riding arena.
|
Larissa still not OK
|
Topic: animals | Link here |
Yesterday was the last day that we gave Larissa Meloxicam for her pain. Today she should have been normal. Was she? No! She was limping again and in some pain. Damn! Now we're off to Bannockburn again tomorrow
|
igitt!
|
Topic: technology, opinion | Link here |
Yesterday git annoyed me by coming up with (or at least claiming) spurious changes in my source tree. I still can't understand what went round there, though I should probably go and examine the differences that it claims. But the xkcd cartoon won, I cloned Yet Another source tree and added my changes. Build Yet Another kernel. All well.
OK, the point of no return. Commit and push.
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main 8 -> git commit
On branch main
Your branch is up to date with 'freebsd/main'.
Huh? I had just changed two files! It seems that I had to dig to find them:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main 9 -> cd sys/dev/usb/=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main/sys/dev/usb 10 -> git diff
diff --git a/sys/dev/usb/quirk/usb_quirk.c b/sys/dev/usb/quirk/usb_quirk.c
index cab6f1b7c074..d86ca5643515 100644
--- a/sys/dev/usb/quirk/usb_quirk.c
+++ b/sys/dev/usb/quirk/usb_quirk.c
@@ -353,6 +353,39 @@ static struct usb_quirk_entry usb_quirks[USB_DEV_QUIRKS_MAX] = {
USB_QUIRK(OLYMPUS, C1, 0x0000, 0xffff, UQ_MSC_FORCE_WIRE_BBB,
UQ_MSC_FORCE_PROTO_SCSI, UQ_MSC_WRONG_CSWSIG),
USB_QUIRK(OLYMPUS, C700, 0x0000, 0xffff, UQ_MSC_NO_GETMAXLUN),
+ /* Selected Olympus DSLR and EVIL models. See ../usbdevs for more
+ * details.
...
Can I now get it to commit the changed files? No:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main/sys/dev/usb 11 -> git commit
On branch main
Your branch is up to date with 'freebsd/main'.
It seems that I have to specify them by name:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main/sys/dev/usb 14 -> git commit usbdevs quirk/usb_quirk.c
hint: Waiting for your editor to close the file... Waiting for Emacs...
[main 31e076d70b5f] Add devices and quirks for some Olympus cameras.
2 files changed, 48 insertions(+), 1 deletion(-)
That worked. The references to Emacs were for the commit message.
OK, now push:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main/sys/dev/usb 15 -> git push
FATAL: invalid repo name: ''
fatal: Could not read from remote repository.
Please make sure you have the correct access rights
and the repository exists.
Huh? Back to the git primer, and tried:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main/sys/dev/usb 16 -> git push freebsd
FATAL: invalid repo name: ''
...
Huh? That's exactly what the primer told me to do. What does the man page say?
SYNOPSIS
git push [--all | --mirror | --tags] [--follow-tags] [--atomic] [-n | --dry-run] [--receive-pack=<git-receive-pack>]
[--repo=<repository>] [-f | --force] [-d | --delete] [--prune] [-v | --verbose]
[-u | --set-upstream] [-o <string> | --push-option=<string>]
[--[no-]signed|--signed=(true|false|if-asked)]
[--force-with-lease[=<refname>[:<expect>]] [--force-if-includes]]
[--no-verify] [<repository> [<refspec>...]]
OK, try the --repo option:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main 29 -> git push --repo=src freebsd main
FATAL: invalid repo name: ''
...
I've given it a repo name, and it seems to have ignored it. What is this nonsense? Some discussion on IRC. Daniel O'Connor wanted to see my .git/config:
[remote "freebsd"]
fetch = +refs/notes/*:refs/notes/*
url = https://git.freebsd.org/src.git
fetch = +refs/heads/*:refs/remotes/freebsd/*
pushurl = git@gitrepo.freebsd.org:.git
Ah, says Jamie Fraser, that pushurl looks strange. Shouldn't it be git@gitrepo.freebsd.org:src.git? Yes, that makes sense. Change it, then:
=== root@dereel (/dev/pts/0) /src/FreeBSD/git/main 30 -> git push freebsd
Enumerating objects: 15, done.
Counting objects: 100% (15/15), done.
Delta compression using up to 4 threads
Compressing objects: 100% (8/8), done.
Writing objects: 100% (8/8), 1.32 KiB | 676.00 KiB/s, done.
Total 8 (delta 6), reused 0 (delta 0), pack-reused 0
To gitrepo.freebsd.org:src.git
4d1ba6febfa7..31e076d70b5f main -> main
Done! My first git commit!
What went wrong here?
git insists on consistent source trees. That's why you can't push to a repo if it has changed since you cloned it: one of the files committed could invalidate your changes.
That makes sense, though it's not the way that older systems did it. But if that's the case, why does git commit not find the changes? In my case there were only two files, but I have to specify them, and if I miss one I have an invalid repo. svn knew and committed the ones that needed committing.
Why is the error message from git push so vague? “Please make sure you have the correct access rights and the repository exists”. git can do that too and report if one of the conditions is not fulfilled.
What's .git/config? It's clear, of course, but the primer doesn't mention it at all, let alone what the entries mean. But it was the clue to the problem.
Do I need the invocation git push freebsd? The primer is unclear, and maybe it's unnecessary. But I'll have to wait until the next commit to find out.
Why did I get the incorrect push URL? I did this with my igitt script, which contains:
git remote set-url --push freebsd git@gitrepo.freebsd.org:${repo}.git
Oh. Further up I have:
REPO=$2
So that part was my fault: I just copied the example in the primer without checking the variables in my script. Still, in some ways it shows the validity of having a script that takes care of the complexity of cloning a source tree.
What a mess! Why does it have to be like that?
|
Gift horses
|
Topic: general, animals, opinion | Link here |
Months ago Emma Jefferson offered a free horse to Yvonne. All she had to do was to pay for the transport from Kybeyan.
But somehow the transport fell through, and for some weeks now Yvonne has been planning to travel there to take a look at the horse and teach it to load: it's young and completely untrained. I can't say that I'm happy: it's a long journey, and she had originally planned to go by train and bus, a connection that only works twice a week, requiring her to leave Ballarat railway station round 5:30, taking all day, and risk missing the twice-weekly bus. The idea was to come back with a friend who has a horse float, so she could bring the horse with her.
Then the friend decided to postpone, and in any case she wanted two horses, so no space for Yvonne's. Yvonne decided to go by car (still 800 km and 9 hours' drive).
And now Emma can't be there. So it's postponed again. Somehow these gift horses are more trouble than they're worth, and Yvonne is wondering whether to go at all.
| Thursday, 24 March 2022 | Dereel → Bannocburn → Dereel | Images for 24 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Bannockburn vet again
|
Topic: animals, opinion | Link here |
Larissa's pain wasn't as bad this morning, but clearly there's something wrong. We let her out to play with Lena, but though she ran a little, it wasn't much, and she didn't show much enthusiasm.
Off as planned Bannockburn, taking, Lena and Piccola for good measure. Piccola has been vomiting more than usual lately, though it's not overly much, and she has a long history of it. But Yvonne was concerned and thought that she might have reached her use-by date—at 13!
Peter Bond looked at them both. Nothing obviously wrong with Piccola (who had never been there before). Much investigation of Lara, including X-rays. The results were interesting:
|
|||||||||||||||||||||||
|
|
|||||||||
Both hips, but particularly the left hip (on the right in the photo) have suboptimal Norberg Angles. Unfortunately we didn't get a copy of the X-ray without the markup, but it was clear that the hip joint doesn't fit well. These images, from hip dysplasia, show the same effect in a more extreme form. The first is abnormal: the ball doesn't fit properly into the joint, and the second is normal. Clearly Lara is somewhere in between.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Peter recognized signs of incipient osteoarthritis—in a 12 month old puppy! Fortunately it seems that we can treat it; he gave her an injection of cartrophen (1 ml), and gave us another three to give her over the next three weeks. He's fairly confident that we can get it under control, since we started early. And it seems that this is a fairly common problem with large dogs. Yvonne was just happy that it wasn't hip dysplasia, which would mean that the dog would have to be put down. But looking at the description, that's exactly what it is.
While we were there, also weighed the dogs again. It's only been a week, and of course there's a fair amount of fluctuation, but the differences are interesting:
| Date | Larissa | Food/day | Difference | Lena | Difference | Food/day | Difference | |||||||
| (g) | (g) | Lena/Lara (kg) | ||||||||||||
| 22 February 2022 | 30.90 | 32.85 | 1.95 | |||||||||||
| 17 March 2022 | 31.14 | 0.24 | 32.44 | -0.41 | 1.3 | |||||||||
| 24 March 2022 | 31.62 | 151 | 0.48 | 32.16 | -0.28 | 214 | 0.54 | |||||||
We're feeding Lena much more than Lara, but she's losing weight! Peter agreed, so for the while we'll give her 240 g per day.
|
Old stamping grounds
|
Topic: general, opinion | Link here |
We had 45 minutes to wait for the X-rays, and I had a camera with me, so decided to take a look down the old road from Bannockburn to Lara (yes, really). It's a small road, but it used to be the quickest way to get to Melbourne before they finished the Geelong bypass freeway.
I recalled a relatively pretty creek just to the east of the Midland Highway, and thought I could get a photo round there. But it was pretty dry, and it didn't seem to be worth getting out of the car. Up the road to where there's still an old shearing shed in relatively good condition:
|
||||||||||||||||||||||||
But this is no longer sheep country. They've planted vines everywhere.
| Friday, 25 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Enhancing X-rays
|
Topic: photography, animals, opinion | Link here |
Today I processed the X-reys of Larissa's hips. They were JPEGs. Should I put them through my normal processing? Yes, why not?
The results were interesting. Here the most interesting of the photos as an original, as processed with DxO PhotoLab and my standard settings, and after further processing it with PerfectlyClear (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The difference is even more pronounced here:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It seems that even the hair is visible. Is this good or not? It looks good to me, but you'd think that expensive X-ray units should be able to do this too.
|
Garden work
|
Topic: gardening | Link here |
Both Helen Miller and Jesse Walsh along today to work in the garden. Jesse in particular made a very visible change: a fence outside the laundry to keep the dogs out of the garden bed, at least from the one side. It's my guess that they won't bother from the other:
|
|||||||||||||||||||||||
|
Igitt! again
|
Topic: technology, language, opinion | Link here |
Now that I'm gradually recovering from my first experience with git, I wondered if anybody else had come up with the play on words „igitt“. Yes: IGitt is “a simple library that allows you to access various git hosting services like GitHub, GitLab and so on via one unified python interface”. And at the end of the page it confirms the etymology. But it seems to add Yet Another layer of abstraction on an already completely overloaded system.
| Saturday, 26 March 2022 | Dereel | Images for 26 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Keeping dogs out of the garden
|
Topic: animals, opinion | Link here |
Yesterday we had a fence put in round two sides of the garden bed outside the laundry. The idea is to keep the dogs out; otherwise they would cut across. And there was no reason to believe that they would go in from the other side:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I suppose a wire across the other side will keep them out until they decide not to go in any more.
|
The passage of time
|
Topic: gardening, opinion | Link here |
The view from the south side of the house is changing, mainly nature's doing. Here 3⅓ years ago and now:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
It's not just the tree; the fence round the edge of the garden bed is also new, though it's barely visible by comparison.
| Sunday, 27 March 2022 | Dereel | Images for 27 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Por Kwan Laksa
|
Topic: food and drink, opinion | Link here |
I've had this jar of laksa paste in the pantry for ever:
|
|||||||||||||||||||||||
It expired nearly 2 years ago, so it's high time to process it. Nothing very special: one part paste, one part coconut cream, 2 parts water. A total of 1,8 litres out of the jar. That's about the same as I get from my foil packages. But this jar has 454 g of paste, and the foil packages have 200 g. Why the difference?
More to the point, what does it taste like? Edible, but barely. It's far too sour. I considered that it might have been due to its age, but apart from that it tasted normal, and it's not as sour as assam laksa, so I'm relatively sure that it's intentional. I ate it, and I froze the other four portions, but I wonder if I did the right thing.
This is the second preparation I have had from Por Kwan. Two years ago I tried a Phat Thai sauce. That, too, was not as good as cheaper products. And given that it's made in Thailand, you'd expect the Phat Thai to be more authentic than a copied Malaysian laksa.
|
Roast pork again
|
Topic: food and drink, opinion | Link here |
This week Yvonne bought a pork roast that was on special at ALDI. I had expected it to be about 800 g, but it turned out to be 1.718 kg, far too much for the two of us.
But we had the choice of freezing it or eating it. So In the end we decided to cook it, eat what we could and use the rest on bread for Abendbrot.
Three months ago I cooked a similar roast and discovered that it was more complicated than I thought. In particular, getting the skin nice and crackling was more work than it seems. It took nearly 40 minutes of considerable smoke to get the skin crackling at all. Based on that experience, I decided to use the “hair dryer” “air fryer” this time. That also required a maximum temperature of 230° instead of the 250° and 240° that I had used last time.
|
|||||||||||||||||||||||
That was, indeed, different. Instead of taking 40 minutes, it started to crackle within minutes. Here it is before starting and after 5 minutes:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
By 10 minutes it had started to burn, and I tried some aluminium foil on the area. But it didn't help much: the toothpick burnt away:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
While messing around with that, the hair dryer fan stopped working. OK, in the oven after all, and I continued cooking there. Once again it took forever. Here a summary, not really worth putting in the recipe:
| Time | elapsed | temperature | oven setting | |||
| 16:27 | 230° | |||||
| 16:46 | 19 | 210° | 1 | |||
| 17:06 | 39 | 180° | 2 | |||
| 17:35 | 68 | 170° | 2 | |||
| 18:30 | 123 |
After that time, the meat temperature was 80°:
|
|||||||||||||||||||||||
And yes, it was good, though possibly it could have done with a little more time. Maybe that was at least in part due to the fact that the meat had no time to rest before we ate it.
So, what do I do next time?
Why did the fan fail on the “hair dryer”? It was only temporary (minutes), and maybe I got something jammed in it. Or maybe it's just that the thing isn't intended to run for a very long time. The timer (which must be set) only goes to 60 minutes, so it's clearly not intended for this kind of cooking.
| Monday, 28 March 2022 | Dereel | Images for 28 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More weather station problems
|
Topic: technology, general, opinion | Link here |
My Fine Offset WH-1080 weather station is not the most reliable device in the world. Apart from strange USB issues, it has great difficulty communicating with the external unit. And so it was again today: I came into the office to discover that it had not been communicating at all for over 10 hours. Normally it recovers if I disconnect the receiver and walk closer to the transmitter. But today that didn't work either.
What should I do? Buy another one? They're not made any more. Buy a different model and go through all the pain that I went through 12 years ago? I don't know if it's worth it.
Took the batteries out, and then it occurred to me: what happens if I put them back in again? It worked! What a flaky piece of hardware!
|
More git fun
|
Topic: technology, opinion | Link here |
Now that I have broken the ice using git, time to catch up on other things that I was planning. The obvious one was the update to date(1) that failed two weeks ago. I had planned to investigate how to recover from that situation, but then a week later the source tree fell apart, and I had to recreate it. No git-oriented recovery possible any more.
OK, put the file into a fresh tree, in the process finding more details to change. Push!
=== root@dereel (/dev/pts/1) /src/FreeBSD/git/main/bin/date 35 -> git push date.1
fatal: You are pushing to remote 'date.1', which is not the upstream of
your current branch 'main', without telling me what to push
to update which remote branch.
Fool! First you commit, then you push, just as the clear error message tells you!
And that worked. Gradually the pain is diminishing, but there are still things like MFCs to worry about. And to my surprise, Philip Paeps responded to the commit message with corrections and suggestions. As he said,
This is a very deep rabbit hole.
|
Those who fail to learn from history...
|
Topic: history, general, opinion | Link here |
Found in Wikipedia today:
The war marked a turning point for Russia. The war weakened the Russian Army, drained the treasury and undermined Russia's influence in Europe. Russia would take decades to recover. Russia's humiliation forced its educated elites to identify its problems and to recognise the need for fundamental reforms.
It has widely been noted that ... [it] led to a war that stood out for its ”notoriously incompetent international butchery”.
Well, almost. I have removed a couple of words to make it look more modern; in fact, though, it was a summary of the results of Russia's last war in Ukraine, the one that I wanted to call the First Crimean War.
Let's hope that this one is over sooner.
|
Goodbye Quora
|
Topic: technology, general, opinion | Link here |
I've been answering questions on Quora for years now, in the process watching its quality deteriorate from a serious question and answer site to a purveyor of nonsense and pornography. Today I received a message from their “moderation”:
Your answer has been deleted as it was found to be in violation of our Spam policy. To learn more about our Quora's policy, click here. If you think this is an error, you can appeal here.
Spam? I don't send any steenking spam! What was the question, anyway?
Q: Who invented the camera in 1888?
A: I don’t know, but I’m sure that he was annoyed when he discovered that others had beaten him to it by 60 years or so.
You might like to read History of the camera. Reading between the lines, you might be thinking of George Eastman, but your description is inaccurate.
Where's the spam in that? Nowhere, of course. But my appeal has remained unanswered, and I note that all the answers were downvoted. Presumably a clever kiddie who gets a feeling of power from that. And a “moderation” team that appears to be non-existent.
The message is clear: goodbye, Quora.
| Tuesday, 29 March 2022 | Dereel → Ballarat → Dereel | Images for 29 March 2022 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Mee goreng mamak again
|
Topic: food and drink | Link here |
It's been some time since I last made mee goreng mamak, at least partially because the recipe was for 3 portions, and I wasn't sure of some of the details. But today I finally got round to it again. Made a number of changes based on what I had. This isn't what I updated in the recipe:
| quantity | ingredient | step | ||
| 160 g | cooked bucatini | 1 | ||
| 30 g | squid tubes | 2 | ||
| 66 g | firm dofu | 2 | ||
| 70 g | fresh tomato | 2 | ||
| 57 g | cooked potato | 2 | ||
| 30 g | onion | 2 | ||
| 30 g | soya sauce | 3 | ||
| 10 g | dark soya sauce | 3 | ||
| 15 g | sambal ulek | 3 | ||
| 15 g | tomato sauce | 3 | ||
| 1 | egg | 4 | ||
| oil | 5 | |||
| 60 g | taugeh | 7 | ||
| karuveppilai (curry leaves) | 7 | |||
|
|||||||||||||||||||||||
|
The flutes are back!
|
Topic: music, food and drink, general | Link here |
Not quite 20 years ago I gave Terry McGee two flutes for repair. And now they're
back! Off to Napoleons to pick them up, along with 4 3 cartons of wine. One was damaged, the first
time I have experienced that problem with Australia Post:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
That one went back; another was marginally knocked around, and at least acknowledged by Australia Post:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
And the flutes? Still not repaired, but I knew that. I need to investigate them in more detail.
|
Braun toothbrushes: tower of Babel
|
Topic: health, opinion | Link here |
We've been using Braun Oral B electric toothbrushes for decades. They're not reliable, but they're getting better. We have had our current brushes for nearly 6 years, and as I noted at the time, that seems to be the life of the batteries: both toothbrushes are showing signs of fatigue.
OK, how about a replacement? But which? The maker's web site shows 6 different series with silly names like “iO Series”, “Genius”, “Smart”, “Pro” and “Vitality” and a total of 29 models (not including different colours). And then there are still other models on eBay, presumably ones no longer in production. How can you decide? The web site doesn't help:
Oral-B’s best ever clean with revolutionary magnetic iO technology for a professional clean feeling. The Oral-B iO Series combines powerful but gentle micro-vibrations with Oral-B’s unique dentist-inspired round brush head design.
OK, Oral-B, what does that mean? What's magnetic about it? And the brushes are the same for (almost) all models. After some searching I discovered that “magnetic” means a different kind of (faster) charger with a magnet. Wow!
Since I was in town anyway, dropped in to UFS to look for what they had there, also including angled interdental brushes. That was pretty much the opposite of Oral B: nothing. On to the Chemist Warehouse, where they had lots of them, equally confusing, but to my surprise a shop assistant who appeared well-informed. And they had models from $20 to $250! What's the difference? They all brush teeth. So I bought a pair of the cheapest.
Back home, things weren't quite as good. After fighting my way into the particularly obnoxious packaging, I discovered that they didn't even have a charge indicator! The power supply was the same as for the older units, so I was able to confirm that it was working by connecting the old brush. But why save on such a small detail?
The real issue, though, was that the thing was much slower than the ones we had. We wanted to replace the existing ones because they were slowing down, but this was worse! About the best thing about them was the guarantee on the side: money back within 180 days. We'll take advantage of that.
|
Measuring Lena
|
Topic: animals | Link here |
Yvonne measured Lena (but not Larissa) today: 78 cm. According to Wikipedia, that's the maximum height for a Borzoi bitch.
|
More ssh pain
|
Topic: technology | Link here |
I've upgraded dereel to FreeBSD 14-CURRENT, which currently seems to be very stable, and which makes it easier to build and test source upgrades. But once again I have had trouble with ssh. After make installworld I could no longer connect.
What does /var/tmp/auth.log say?
Mar 29 14:21:51 dereel sshd[68066]: userauth_pubkey: key type ssh-rsa not in PubkeyAcceptedAlgorithms [preauth]
OK, that message is (almost) clear enough. But what's PubkeyAcceptedAlgorithms? Off looking online and found a suggestion to run:
=== root@dereel (/dev/pts/0) /usr/src/bin/date 50 -> ssh -Q PubkeyAcceptedAlgorithms
unknown option: '-Q'
...
ssh-rsa
What's that “unknown option”? It's required; otherwise it would try to interpret PubkeyAcceptedAlgorithms as a system name. But why ssh? That's the client, and the issue is with the server.
What does /etc/ssh/sshd_conf say? No mention of PubkeyAcceptedAlgorithms. How do the files differ between eureka and dereel? Lots of stuff, of course, but the closest to a smoking gun was:
--- /eureka/etc/ssh/sshd_config 2013-12-24 10:34:47.000000000 +1100
+++ /etc/ssh/sshd_config 2019-11-01 15:28:52.000000000 +1100
@@ -49,11 +38,11 @@
#MaxAuthTries 6
#MaxSessions 10
-#RSAAuthentication yes
#PubkeyAuthentication yes
sshd_conf is rather silly in that it shows the defaults commented out. So this means that on eureka RSAAuthentication was enabled, and on dereel it wasn't. OK, enable it.
Mar 29 15:36:29 dereel sshd[68447]: rexec line 41: Deprecated option RSAAuthentication
Mar 29 15:36:29 dereel sshd[68447]: reprocess config line 41: Deprecated option RSAAuthentication
Mar 29 15:36:29 dereel sshd[68447]: userauth_pubkey: key type ssh-rsa not in PubkeyAcceptedAlgorithms [preauth]
Mar 29 15:36:30 dereel sshd[68447]: Connection closed by authenticating user grog 192.109.197.137 port 58214 [preauth]
Now that's really “deprecated”. It doesn't work at all.
Long discussion on IRC. Silly Groggy! As Daniel O'Connor says, you don't do that, you add
PubkeyAcceptedKeyTypes=+ssh-rsa
Of course! It's not documented anywhere except without explanation in a couple of Google hits, but it's so obvious! And, more importantly, it works.
The real issue, though, is that the authentication method is clearly obsolete. What's modern? I still need to investigate that.
| Wednesday, 30 March 2022 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Understanding tortilla dough
|
Topic: food and drink, opinion | Link here |
I've been making tortillas for well over 10 years, and initially they were frustrating and tended to disintegrate. I decided that much had to to with the composition of the dough, specifically the ratio of masa to water, and I wrote a page with detailed descriptions of the proportions for various brands of masa.
But last year we got no fewer than 5 packages of a kind of masa (“El Maizal”) that was completely different. Instead of a masa:water ratio of 1:1.5 or so, it wanted only 1:1.38. And the various packets were different! The one I'm working through at the moment wanted 1:1.45.
Or so I thought. For reasons I can't recall, I tried adding more water, and now I'm at 1:1.6, and it still works well. What does that mean? My best take is that the quantity is not as important as I thought. Maybe the real secret is to let it rest for 10 to 15 minutes before pressing.
|
Returning the wine
|
Topic: technology, language, general, opinion | Link here |
Yesterday I sent a mail to McLaren Vale Cellars about the damaged wine shipment. No response. OK, enter a “dispute”, a rather heavy-handed term for what I wanted to say. But that's modern web sites for you, and McLaren Vale Cellars have a very modern web site. First I had to choose from a limited set of reasons. “Damaged in transit”? No, nothing like that. The closest I came was:
I did not receive an item I purchased, or the item I received is dead on arrival, not working, not the one I ordered or significantly different than [sic] what was described.
Clearly a well adapted text. Even the “different than” (and not “different from” or even “different to”) makes it clear that it's a US American base.
|
Returning the toothbrushes
|
Topic: health, opinion | Link here |
Yvonne into town shopping today, taking yesterday's toothbrushes with her for return.
It didn't work. Yes, she was able to return the unopened one, but Chemist Warehouse didn't want to know about the 180 day warranty on the other one. Take it up with Braun if you want.
Is that acceptable? My understanding is that they're our interface to the maker. I could follow it up with Consumer Affairs Victoria, but do I care enough? It does, however, say something about Chemist Warehouse.
|
More system upgrade problems?
|
Topic: technology, multimedia, opinion | Link here |
Now that I have upgraded dereel to FreeBSD 14-CURRENT, I'm still discovering strangenesses. Hitting c-p to get the previous command gives me:
=== grog@dereel (/dev/pts/4) /spool/Series/Doc/01 18 ->
<t https://www.sbs.com.au/ondemand/video/1891506243952
I've seen that before, but not for decades. The < at the beginning of the line is an indication that it's not the start of the line: there's more to the left. Press c-a for the beginning of the line and I get
=== grog@dereel (/dev/pts/4) /spool/Series/Doc/01 18 -> yt https://www.sbs.com.au/ondema>
What causes that? I've forgotten. Off to search for bash line wrap. Surely there must be somewhere in the documentation about it, but I've been using bash for over 30 years now, and I don't know where to start. And why is it happening now?
Finally I found it: not a -CURRENT issue at all. YouTube download has followed the misguided example of git and other Linux-space software and introduced half-coloured output:
[download] 0.9% of ~688.73MiB at 256.49KiB/s ETA 07:09 (frag 5/552)
I wish I understood why people do that. But there's clearly a solution: reset the TERM environment variable for the invocation. In yt I have:
TERM=
...
youtube-dl $PROXY -c --restrict-filenames -o "$PREFIX%(title)s.%(ext)s" --all-subs "$URL"
That works:
[download] 1.7% of ~720.15MiB at 1.97MiB/s ETA 06:42 (frag 10/582)
But yt is a shell function, not a standalone script. And it seems that shell functions leak: TERM remains unset, causing the problems I saw.
The solution is simple:
TERM= youtube-dl $PROXY -c --restrict-filenames -o "$PREFIX%(title)s.%(ext)s" --all-subs "$URL"
There's clearly also the possibility of making the function into a standalone script, but I don't see the need for that.
|
Dog heights: the other shoe
|
Topic: animals | Link here |
Yvonne has now also measured Larissa. She's 76 cm tall, still well above the average height of 68—78 cm for Borzoi bitches.
|
Quora responds
|
Topic: technology, general, opinion | Link here |
Response from Quora about my “appeal” two days ago:
Date: Tue, 29 Mar 2022 22:39:11 +0000
We are very sorry for this experience. Your content was mistakenly removed as a violation of our spam policy, but it is now reinstated.
That's something, but I think it's still clear that my Quora time is over. There are no sensible questions to answer any more, and moderation is (probably unintentionally) on the side of the abusers. Remove answer immediately, take 2 days and an appeal to reinstate it. Why waste my time?
| Thursday, 31 March 2022 | Dereel | Images for 31 March 2022 |
| Top of page | ||
| previous day |
|
Moving house
|
Topic: general | Link here |
Seen from my office window this morning:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
DMARC?
|
Topic: technology, opinion | Link here |
Last month I set up DMARC for my external mail server, and almost immediately I started getting messages that I can't parse:
[-- Attachment #2: protection.outlook.com!lemis.com!1648512000!1648598400.xml.gz --]
[-- Type: application/gzip, Encoding: base64, Size: 0.8K --]
[-- application/gzip is unsupported (use 'v' to view this part) --]
Clearly I need software to decipher it for me. But how? A google search brought a number (well, 210,000) of hits, including The Ultimate Guide to DMARC Reports in 2022 from GLOCK APPS and Understanding DMARC Reports from EASYDMARC (no I dea why they shout). GLOCK APPS offered free analysis for less than 100,000 messages a day, which seems more than adequate for my 500 or so maximum. OK, sign up. Nice, broken web site that overlays the password entry field with the usual stupid restrictions: at least one upper case letter, one lower case letter, one special character, one digit. Entered N0 way Jose!, but it didn't like that, and out of spite erased all my other entries too.
Finally found something that it didn't complain about, but it also didn't respond to “SUBMIT”. Goodbye, GLOCK APPS.
On with EASYDMARC, which was marginally easier to set up. More DNS changes, which it wanted me to set up in the wrong place, and before long I had information that I would have to wait 72 hours for any information, and that things would be free for 14 days. And then? They were too polite to say. I did, however, get a health check of my DNS records: SPF OK, DKIM not OK (not surprising; I don't use it), DMARC also not OK!
|
|
|||||||||
Huh? That's exactly what they asked me to set! And no suggestion as to how to improve it.
Somehow these people are all a little suspect.
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 |
