|
| Greg's diary |
| June 2024 |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Saturday, 1 June 2024 | Dereel | Images for 1 June 2024 |
| Top of page | ||
| next day | ||
| last day |
|
Winter without heating
|
Topic: general | Link here |
Today's the first day of winter, an ideal time for the air conditioner to fail, and fail it did.
Well, it ran, sort of, but it barely heated. Once again I had the errors 30 (trip lock) and 6 (High Discharge Temperature). And after power cycling after an error 6, I briefly saw ridiculously high temperatures in the zone controller after the power-on sequence, up to 51°.
And that rings a bell. Isn't that the critical point of the refrigerant? Could it really be that it's getting too hot? And then again there's this strange 138° C value for error 6. Could it be that they confuse Fahrenheit and Celsius? 138° F is 58.9° C, high enough to be an issue and close enough to 51° to be plausible. It doesn't explain why the temperature appeared in the zone controller, which should have been showing the room temperature (round 17°), but I've seen something like that before under normal circumstances. 30.6° rings a bell.
Still, what does it all mean? Under what circumstances would the refrigerant get that hot? Apart from that, all points to lack of refrigerant. Hopefully I can get Tony to come quickly.
|
Another power issue
|
Topic: technology, general, opinion | Link here |
As a result of the air conditioner issues, I listened carefully to what it was doing. And then round 7:10 it just cut out, stopped dead.
That's unusual. Normally it gradually winds down. Damn, what is it this time? Into the office for some reason, where I saw the display on the ATA: only 2 LEDs on, out of 4. That's a reasonable first indication that we're off the net.
What does the NTD say? All looks normal, for once. But the display on hydra was funny: window manager gone. A look on eureka confirmed my worst suspicions: Boot:. Somehow it had restarted. Power issue? That sounds like eureka, and it would explain the air conditioner (which in the meantime had started limping on again).
The usual fun getting eureka up again, except that this time I couldn't get the web browser to connect. Squid wasn't running. Started it, getting an unexpected error in /var/log/squid/cache.log:
2024/06/01 07:28:31 kid1| cannot change current directory to /usr/local/squid/cache: (2) No such file or directory
Yes, the directory didn't exist. But that wasn't the first time the message appeared:
2024/01/24 08:40:26 kid1| cannot change current directory to /usr/local/squid/cache: (2) No such file or directory
And that's pretty much the last thing that squid had to say before today. It appears that it hasn't been running since January. The browsers on hydra don't use the proxy, and maybe the ones I have been using on eureka don't either; at one point I was using three different proxies depending on the display.
Yet another untidy end that requires more time than I want to invest.
And then, of course, tiwi had also restarted. And this time, once again, no display. Turned the TV on, rebooted, and all was well. My best guess now is that the display card can't identify the TV if it's not turned on. It doesn't explain, though, why this time I didn't see the boot messages.
|
Five years of phone smart
|
Topic: history, technology, general, opinion | Link here |
Five years ago today I grumbled about the same things that still annoy me with Android. Now I need a new phone just to be able to make calls. At the time I wrote:
I'm thinking of reverting to my old steam-driven mobile phone: at least I can always make calls and terminate them without scratching my head.
Of course, that's no longer a solution either. If my 3 year old phone can't make calls after the end of August, certainly no steam-driven phone will.
More searching. Maybe I should just buy a cheaper phone. About the only thing that irritates me about them is that it could be underpowered, like the Xiaomi Redmi 9A. Juha Kupiainen pointed me at JB HiFi, who have a couple of Xiaomi phones for sale. Maybe they've done their homework, and they're the best choice. It doesn't mean I have to buy them from them, of course.
|
Two years of core, over?
|
Topic: history, technology, opinion | Link here |
Two years ago today I became a member of the FreeBSD Core Team. It should now be over, but for reasons to which I wasn't privy (midnight teleconferences), it has been delayed. As I said at the time “Video meetings aren't going to work effectively. ”
So now we're hanging on to our power for another few days. The elections are almost over, and after several weeks, only 52% of developers have voted. Why? Apathy, I suspect. I certainly failed in one of my stated goals, to make the Core Team more visible. But it seems that I was the only one who wanted to do that. I certainly haven't nominated myself this time round. I belong to another generation, literally: the youngest member of the current core team is 47 years younger than I.
|
Porting enblend 4.1.4
|
Topic: technology, opinion | Link here |
House photo day again today, and I still haven't ported the last functional version of enblend to hydra. Tried again today, with limited success. For some reason it tried to refer to strerror_r() as returning a string, when in fact it returns an integer.
OK, not an issue, hack the source. Oh. The whole semantics are different. But it seems that the port had prepared for that, and it had alternative code for what proves to be the POSIX version (the other version is the GNU version; rms has something to answer for). But try as I might, I couldn't persuade it to accept my hacks.
OK, this is silly. The obvious thing is to use the FreeBSD Ports Collection framework, which I still have at /home/src.local-and-old/FreeBSD/svn/ports-rot0/graphics/enblend/. I wonder what horrors it has in store for me.
| Sunday, 2 June 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More air conditioner pain
|
Topic: general | Link here |
I left the air conditioner off overnight. It was another cool night, and about the only effect of me leaving it running was that it would keep me awake wondering what was happening. Even when I turned it on, it didn't seem to make much difference: it didn't go beyond 20° until 13:00. Nothing for it: call Tony Nesci first thing tomorrow morning.
|
No mail!
|
Topic: technology, opinion | Link here |
Into the office this morning to discover that mail delivery had stopped during the night. Further investigation showed that the root file system on eureka was full.
Why? Lots of searching with find didn't find anything obvious. No new large files in the last 24 hours. But looking through the backup logs gave me a clue: it happened during backup. And then looking at eureka from tiwi gave me the clue:
=== grog@tiwi (/dev/pts/9) ~ 1 -> df
Filesystem 1048576-blocks Used Avail Capacity Mounted on
eureka:/videobackup 39,662 31,535 4,954 86% /videobackup
eureka:/photobackup 39,662 31,535 4,954 86% /photobackup
eureka:/ 39,662 31,535 4,954 86% /eureka
eureka:/home 3,705,520 2,667,836 741,242 78% /eureka/home
eureka:/Photos 7,629,565 6,905,419 647,849 91% /Photos
eureka:/dump 39,662 31,535 4,954 86% /dump
That may look like 6 file systems, but in fact it's only 3. For reasons that I can no longer confirm, /videobackup, /photobackup and /dump all got mounted on the root file system. The insidious thing is that files in these directories aren't visible locally, only via NFS. And of course I've been there before; the last point in cold start reads “Check other FreeBSD computers for correctly mounted NFS file systems”.
|
Stones Road, ten years on
|
Topic: Stones Road house, history, photography, opinion | Link here |
We bought our current property in Stones Road, Dereel, some time in January 2014, though I didn't note the exact date of the transfer. I had started taking panorama photos even earlier, and last November I tried a comparison ten years later.
But that was a view that I hardly repeated. Ten years ago today I started on the weekly series that I have been keeping up ever since, with up to 520 views of the house from various perspectives. So now I have a comparison of the same view taken then, 5 years ago and yesterday.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The view of ten years ago is almost unrecognizable. The trees at the left and right edges of the photo are both gone, and other vegetation has taken its place. And then there's the difference in the photos themselves: that's only partially due to the processing. Ten years ago my widest lens was the Zuiko 9-18 mm rectilinear lens, but the other two were taken with the M.Zuiko ED 8 mm f/1.8 fisheye. Still, there are interesting differences between “then” and “now” that I can't completely explain.
|
Chicken with cashew nuts again
|
Topic: food and drink, opinion | Link here |
Chicken with cashew nuts for dinner again tonight. And of course some changes. The recipe called for 25 g of capsicum, something that we don't eat very often. What do we do with the rest? But by chance I had read an article “What to do with broccoli stems”, and we had some broccoli stems left over. So I put in a julienne of broccoli stems instead.
How was it? They're softer than I expected. They didn't taste bad, but I'm not convinced that that's the best thing for dish.
In passing: we ate 212 g of rice between us.
| Monday, 3 June 2024 | Dereel → Ballarat → Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Air conditioner woes, continued
|
Topic: general | Link here |
Another cold night without the air conditioner, which wouldn't have contributed much to keep the place warm anyway. Up and called Tony Nesci. They're on holiday until 17 June! One of the issues that we have with a small company.
Google search for "Air conditioner repair Ballarat”, which brought many hits, many with tastefully truncated phone numbers, requiring me to follow the links to find the numbers. Got as far as these notes:
| Celsius | 5333 1267 | |
| G J Bradding | 5330 2166 | |
| Mechtrol | 0412 011 320 | |
| Buggzys | 0400 007 441 | |
| Eeles | 5333 2117 | |
| Waldron | 5336 0000 | |
| SJS Servioces | 0400 154 569 | |
| JZs | 5332 0140 |
Waldron? They're the people who supplied the unit, and who were singularly unhelpful four years ago. Still, they could have changed. Called them up and spoke to Anne, who wanted to know details of when we purchased the unit and had great difficulty finding them. Then she put me on hold and came back to say that I should contact a Actron service person—just as 4 years ago. But this time she had a phone number: 0401 022 148, of Nathan Barnes of South West Electrical in Ballarat. That sounded lot better than the run-around I got last time.
Called up Nathan, and yes, he could come and take a look—on Wednesday. Two more days of shivering, but a whole lot better than the 23 days that it took last time.
But we can't continue like this. In the shed I found an electric convection heater, which made some difference, but we need more, so into Ballarat to buy a couple more. That helped a little, and we should be able to weather the next couple of days.
Why do these things always happen in extremes of weather?
| Tuesday, 4 June 2024 | Dereel | Images for 4 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More heating pain
|
Topic: Stones Road house, general, opinion | Link here |
There's no point running the air conditioner any more, so we're relying on the electric heaters, really boys sent out on a man's job. Gradually we're getting used to temperatures 2 to 3° lower than we like. I'm hoping that Nathan will be able to work a miracle when he comes tomorrow, though if it's really a refrigerant leak, finding it could take a lot of time.
|
Power fail!
|
Topic: Stones Road house, general, technology, opinion | Link here |
Heating with electricity uses a lot of power, of course, something that we don't have much of. In principle we have 11.5 kW available, but even that got exceeded just before breakfast:
tstamp Pac Status SOC VBat PacGrid PacBat PacPV FromPV W1
2024-06-04 10:42:16 0 Internal Bypass Closed 22 255 NULL 0 0 0 13750
2024-06-04 10:42:17 0 Internal Bypass Closed 22 255 NULL 0 0 0 13575
2024-06-04 10:42:19 0 Internal Bypass Closed 22 255 NULL 0 0 0 13412
2024-06-04 10:42:20 0 Internal Bypass Closed 22 255 NULL 0 0 0 13152
2024-06-04 10:42:21 0 Internal Bypass Closed 22 255 NULL 0 0 0 13099
2024-06-04 10:42:22 0 Internal Bypass Closed 22 256 NULL 0 0 0 13196
2024-06-04 10:42:24 0 Internal Bypass Closed 22 256 NULL 0 0 0 13182
The last column, W1, is the total power used in the house. In principle, both the PV system and the grid should only deliver 11.5 kW, and that's when the PV system gave up (“Internal Bypass Closed”). Fortunately the grid didn't, though the main switch is rated at 50 A.
But later a second failure occurred: at 11:39 the mains power failed, just as we were using 8.2 kW. The inverter can only deliver 6.5 kW for a short period, after which it drops back to 6 kW. What does it do? What can it do? It disconnected. Total power failure! And I received a gleeful SMS from Powercor telling me that they expected to restore power at 14:00.
They lied: power was restored after only 9 minutes, but the damage was done. All machines down except for lagoon, which managed on its UPS. I should really put UPSs on the others, but previous experience has shown them particularly sensitive to power surges.
Bringing the machines up had some surprises. I know that eureka is special about coming up, but hydra proved even worse: it stayed powered down. And eureka had strange problems with the USB disks: it didn't recognize the third one when I connected it. Lots of playing around didn't help, and when I tried to reboot it hung before loading the FreeBSD loader.
Oh. eureka is over 10 years old. It has done good service, but is it on the way out? I certainly need a backup strategy in case it fails. But fortunately, after power cycling and leaving the USB disks connected, it booted and recognized all the disks. This is the first time I recall booting with the USB disks powered off. Is there an issue there? Certainly the backup strategy is something to consider.
And hydra? Powered it on. Nothing. What's wrong there? While I was scratching my head, it came up as if nothing had happened. After an estimated 5 minutes. That's presumably something to do with the tuning that it needs for fast memory, but it's yet another reason to ensure that it doesn't lose power.
And after everything came up, start firefox. Which profile? I no longer knew. I couldn't find anything that matched what I had been running before, and it took me over half an hour before
And tiwi? It rebooted normally and was hanging on NFS mounts—and though the TV was off, I had a vty display. All wonderful, and not what I expected.
But when I came back later, after switching the TV off and then on again, I had no display! Reboot, no vty display, but after the normal time (which seems double as long in these circumstances), it came back with a normal X display. What went wrong? I don't know. It turned out that the (wireless) keyboard needed new batteries, but surely I used the mouse, which was working fine.
What a day!
|
RIP Mike Karels
|
Topic: technology, opinion | Link here |
Shock news today: Mike Karels, one of the original BSD troop, died yesterday of a heart attack on his way back from BSDCan. It was completely unexpected, and we're still trying to come to terms with the news.
| Wednesday, 5 June 2024 | Dereel | Images for 5 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Air conditioner: the verdict
|
Topic: Stones Road house, general, opinion | Link here |
Nathan Barnes of South West Electrical along this morning to take a look at the air conditioner. As we suspected, low refrigerant. Well, almost no refrigerant at all. And he showed how the distribution pipes iced up almost immediately, apparently another sign of low refrigerant. Yes, he could top it up, but it would take 6 kg at $80 a kilogram, and who knows how long it would stay in there. He checked the inside unit for obvious leaks, but didn't find anything, so he separated the piping to the inside unit from the outside unit and filled the inside to 400 psi (about 2.75 MPa in sane units), and the outside unit to 280 psi (about 1.9 MPa) with nitrogen. Why not the same pressure? He ran out of nitrogen. And now it'll have to stay like that for a week!
The unit is now 9 years old, and it's been used more than most. He assures me that it's worth repairing, but it's also on the cards that we'll have more fun like this. So, back to the Tandem philosophy: we need a backup. After some discussion we decided on a Mitsubishi Electric MSZ-AP60VG(D), 6 kW cooling and 6.8 kW heating, which he'll install in the lounge room next week. That won't help much in the offices and bedrooms, but they're relatively easy to heat with the electric heaters, and it's only as a backup measure.
In passing, I was happy to find that Nathan was part of a big company, South West Electrical. One of my concerns with Tony Nesci at Atmos was that he was by himself, and if he wasn't available (like now) there's nobody to cover for him. But that proved to be misleading: the South West Electrical that I found is in Queensland, and it has nothing to do with Nathan. Nathan is South West Electrical, though he does have an apprentice.
|
Oven fail
|
Topic: food and drink, general | Link here |
Nathan's an electrician, and one of the things that we need done is to change the switch for the oven, which failed in January:
|
|||||||||||||||||||||||
Yes, he can do that next week with the other work. But he touched the switch! Never mind, it still worked.
Until I tried to bake today's batch of bread. It started heating, but when I came out again a few minutes later, it no longer had power. Nothing that I could do kept it running. My guess is that the current fried the last remnants of contact.
What to do? I have bread waiting to be baked, already threatening to escape from the form:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Nothing for it: take the wires out of the switch and join them with a terminal block. Simple? Well, first of all Jim Lannen, my all-time favourite sparky, came up with a late greeting: the frame that should have held the switch in place was not anchored in the wall, so when I removed the screws, it fell down in the space behind. No way to replace the switch in the wall. Nathan will have to do it next week.
And then joining the oven wires? I couldn't find a terminal block! I know I had plenty when I didn't need them, but all I could find after a considerable search was another worn-out switch. But it had a connection that was large enough to take both wires, so I put them in there.
It works! Ship it!
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It's a good thing that it's only for a week.
| Thursday, 6 June 2024 | Dereel → Ballarat → Dereel | Images for 6 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
New glasses
|
Topic: health, opinion | Link here |
My new prescription glasses are ready. Into town to pick them up. Reading glasses +2.25 dioptres! I had specifically asked for 1.5.
Oh, that's right. My prescription was +0.75 dioptres spherical in each eye, from which 0.75 (left) or 1.0 (right) cylindrical are subtracted. That means that the focus in my left eye is really between 0 and +0.75 dioptres, and in the right eye between -0.25 and +0.75 dioptres. So the reading glasses are correct, as I was able to confirm when I got home.
It's funny wearing progressive (multifocal) lenses again. It has been less than 3 months since I stopped wearing them, but that's enough to forget the strangenesses. In addition, they weren't as well adjusted to my head as I thought: there's some discomfort behind the right ear, and it looks like I'll have to go and get them adjusted again.
Still, things are much better. Now I can really use smaller fonts on my computer, and I can read the text on the TV much better. The only issue with the TV is that the vertical field of view is larger than the fixed focus part of the lenses, so I have to move my head. Maybe it would be an idea to have Yet Another pair of glasses for TV.
So it's done! Only a little over 9 months since the first diagnosis of cataract. And yes, things are much better.
|
Google Maps pain
|
Topic: technology, general, opinion | Link here |
Coming home was a pain. In principle I know the way, but I let Google Maps guide me anyway, though the instructions are a bit vague round the south of Ballarat. Going down the extension of Humffray Street I saw warnings of road works up ahead, so I turned towards Barkly Street. But Google Maps wanted to take me in the wrong direction!
Finally got to the end and found significant road works that encompassed Barkly Street too. After waiting for ever for the traffic lights and not getting through when they did turn green, turned off for another alternative. Oh. Road closed.
Finally along Brittain Street through a part of town that I have never seen before, past the Ballarat Gold Mine, finally ending up at the junction of Victoria and Albert Streets in Sebastopol.
Damn Google! Why didn't they tell me? Oh. They did. They wanted to take me along a sub-optimal way back to the main Midland Highway, but the instructions were so vague that I didn't understand. That at least explains why they wanted to take me the wrong way down Barkly Street. I need to examine these things more carefully. The mobile phone format doesn't help.
|
Another power fail
|
Topic: technology, general, opinion | Link here |
On the way home, got a frantic call from Yvonne: she had plugged an electric heater into one of the power strips on the desk in her office, despite my warnings to only use the wall outlets: they were connected to the UPS for her computer, lagoon, rated at 350 W. Bang! Goodbye to the last system that didn't go down on Tuesday. But the worst thing about it was that, even after she disconnected the heater, the power didn't come back, and the UPS screamed continuously.
Nothing for it, she had to wait until I got home, 20 minutes later. The UPS is one of the ones that I bought at Officeworks last February, not my favourite choice. And it seems that it would have continued to scream until its power ran out. To turn it off required a long press on the power button, and turning it on also required a long press. Barely acceptable if you know, but panic-inspiring if you don't.
|
Mona: I love you
|
Topic: animals, opinion | Link here |
Mona is still not overly affectionate, though things are improving. But after I got home, she was all over me, purring and rubbing up against me. Why? Our guess is that it was a “my saviour” moment after I got back and stopped the screaming from the UPS. The animals had all reacted badly to the noise, going into hiding.
| Friday, 7 June 2024 | Dereel | Images for 7 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More system upgrades
|
Topic: technology | Link here |
An interesting thing about solid-state disks is that, though they're generally more expensive than rotating disks, that doesn't apply to smaller sizes. There seems to be a lower price limit for old-style disks, round about $70, but “small” SSDs can be much cheaper. Last week I bought 4 128 GB disks for $19.31 each, including postage. My intention is to use them as system disks for machines like tiwi (which currently has 17 TB of spinning disks) and lagoon (again with considerable image data on line).
Don't I have enough to do to finally complete the migration to hydra? Yes, of course I do. But I have significant performance issues with tiwi, which is no longer a match for firefox's memory hunger, and it's swapping its heart out. I should start the migration while I can still return the disks if there's anything wrong with them. So today I dragged up the remnants of dereel, once my test machine, and put one of the new disks in it as a secondary. Copy tiwi's root file system (40 GB) and partition the new disk accordingly. It's refreshing to see how I was able to run fsck on the resuscitated image and see the soft update journal get applied.
And now? Bring dereel up to date, I suppose. Then switch disks and run the tiwi root system as teevee, update it and test before putting it in the lounge room.
Oh, and it's not solid-state disk, it's solid-state drive, because clearly they have a drive motor. That's almost worse than this silly term “hard drive”.
|
Where are my lenses?
|
Topic: photography, general, opinion | Link here |
Last week I bought two genuine Leitz Wetzlar lenses: a 13.5 cm f/4.5 Hektor number 855598 (which, according to this site, means that it was made in 1951) and a 20 cm f/4.5 Telyt, number (I think; it's hard to read) 1367108 (made in 1956). Leica lenses, even old ones, command ridiculous prices. There's currently a 35 mm f/3.5 Elmar, admittedly serial number 182023 (1933) on offer for $3,400. But I got both of these lenses together for only $112.50.
But where are they? As usual I'm trying to track with Australia Post, which tells me:
|
|
|||||||||
|
|
|||||||||
And I took these screen shots at the same time, today at 11:37. The first tells me that it hasn't been posted yet, the second tells me SUCCESS: received on Wednesday, now (after 2 days storage) on its way to Melbourne airport . And for all that, it'll be between another 6 and 11 days before it gets here! I don't believe a word of it, of course, but how can they be that inaccurate?
And what's this SUCCESS? Nothing to do with their performance. It's the location of the post office, a suburb of Perth.
|
Mona thaws
|
Topic: animals, opinion | Link here |
Why was Mona so affectionate yesterday? Was it really because she had identified me as the solution to the UPS alarm noise yesterday? I wish I understood how cats think. Still, she's much friendlier now, and she spent a lot of time on my lap while I was watching TV.
| Saturday, 8 June 2024 | Dereel | Images for 8 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Leg pain?
|
Topic: health, history, opinion | Link here |
Woke up in the middle of the night with mild pain in my lower left leg. Nothing serious, but...
That's a symptom of a slipped disk! Surely that's over and done with. As it happens, it was a year ago today that I had the issue. But it healed up, and I have had no symptoms for six months. Surely it's not coming back? What can I do?
|
dereel upgrade, completed
|
Topic: technology | Link here |
Brought dereel up to date today. It's a much slower machine than hydra, and the build took nearly 4 hours. Somehow clang is an exercise in navel contemplation: it seems to take longer than the rest of the build together. Next is to revive tiwi as teevee and bring it up to the latest stable release. Mañana.
|
Relying on digital devices
|
Topic: general, technology, opinion | Link here |
How accurate are digital devices, in this case thermometers? I asked myself that question when I used the Dallas 1820 twenty years ago. I came to the conclusion that the inaccuracies were negligible in comparison with random temperature fluctuations in the object I was messing. And I've stayed of that opinion, though I admit to not wanting to challenge it.
Lately I've been measuring temperatures frequently to adjust the electric heaters we're using. And yes, there have been significant differences between the fixed displays and the infrared thermometer that I carry around with me. But then, the fixed thermometers have significant hysteresis, so I accepted that.
But today I found that the temperature in the lounge room was radically different. The thermometer on the indoor unit of the weather station showed 19.5°, but the (blue) infrared thermometer showed only 16°. And it did it with the other two thermometers in the house too. Yvonne thought that it was because I had been using it too much, but that makes no sense.
What about another thermometer? Found an old (orange) infrared thermometer and compared. Round 3° difference. Why? The blue thermometer showed low battery. OK, that shouldn't make a difference, but let's change the batteries anyway. It showed the correct (well, plausible) temperature again! It should be ashamed of itself showing incorrect values when the battery is low.
But it didn't stay that way. Once again it started showing incorrect temperatures. Is Yvonne right, maybe, that it's because it's being used so much? What could cause that? Overheating inside the device? I'm baffled.
|
A new display card?
|
Topic: technology, opinion | Link here |
tiwi.lemis.com, my current TV driver computer, has a display card with an nvidia GeForce GT 730 chip set. I get the feeling that it's sometimes not fast enough. What do I buy? How much do I want to pay? The sky's the limit.
I have a number of other nVidia cards lying around, two of them with the necessary HDMI output. One is a GeForce GT 210, the other a GeForce GT 710, both weaker. But I'm replacing the machine with a more powerful one. Would they work well enough? And then there's one in distress, a Microsoft box. What is it? How can I tell? With FreeBSD it's simple:
=== grog@tiwi (/dev/pts/5) ~ 13 -> pciconf -vl
vgapci0@pci0:1:0:0: class=0x030000 rev=0xa1 hdr=0x00 vendor=0x10de device=0x0f02 subvendor=0x10de subdevice=0x0000
vendor = 'NVIDIA Corporation'
device = 'GF108 [GeForce GT 730]'
class = display
subclass = VGA
Spent some time searching the system menus, but there was nothing there that mentioned display cards. Finally asked Gemini, which came up with this surprising answer: select “Performance” from the Task Manager and then “GPU 0” (no idea why it thinks there's only one). And that will show details of the display card. It's a GeForce GT 710, and I have another already, so there's no need to remove it from the machine.
Gemini does come up with another “solution” with the System Information pages, but it didn't work for me.
What a system! But I'm finding Gemini more and more useful.
|
Israel: Horror without end
|
Topic: politics, history, opinion | Link here |
I'm absolutely amazed and horrified that nobody has taken concrete steps to end the current Israeli genocide in Gaza. But when it came out today that they'll be on the United Nations list of violators of children's rights, Israel was horrified. It's not going to help much: today they liberated 4 of the hostages, at a cost of about 210 dead and 600 wounded, later updated to 274 dead and 700 wounded, a total of nearly 1000! That's a success! Who cares about the non-Jews?
But that's nothing new. While looking back through my diary, I saw this, from 17 November 1989:
A Texan finds oil in his back yard and goes to Dallas to find how to invest the money. They send him home and tell him to come back in a week. Then they told him: buy a jet bomber for $50,000,000, bombs for $4,000,000, hire a pilot for $1,000,000. When he asked why, he was told: "If the Jews do it, it must be a good investment". Even Grog understood after a while.
Somehow there's a disconnect. The Israelis really seem to believe that they're doing the right thing: Gilad Erdan posted “our army is the most moral in the world”. And somehow that's the general consensus in Israel: they seem to be living in a parallel universe.
Back to my question above: “Who cares about the non-Jews?”. I thought that might be a little harsh. But no, it really seems to be the Israeli view. The Washington Post article above includes the quote:
Gantz cited Israel’s first prime minister, David Ben-Gurion, as saying “it matter not what say the goyim, what is important is what do the Jews.”
| Sunday, 9 June 2024 | Dereel | Images for 9 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
teevee progress
|
Topic: technology | Link here |
Next step with teevee: put the disk in the spare machine on top of hydra. That was a non-starter, literally: it didn't make it through the power-on self test. I recall a machine being defective, and that must be the one. But it had 32 GB of memory in it, so I took half and increased the other machine to 22 GB, which should be more than enough. The other 16 GB can go into the new lagoon.
Then upgraded teevee from FreeBSD 13.1-STABLE to 14.1-STABLE, which went surprisingly smoothly once I found enough space for /usr/obj: it now takes up nearly 17 GB! Rebooting was another matter:
|
|||||||||||||||||||||||
What went wrong there? Tried rebooting, and it hung at this point:
|
|||||||||||||||||||||||
“module kernel exists but with wrong version”. What does that mean? Ah, it's nvidia's inimitable way of saying “the running kernel is not what I expect”. Yes, that's to be expected after an upgrade, but the module should just refuse to load. Fortunately I was able to boot single-user and remove the /boot/loader.conf file, after which things worked.
And I was even able to upgrade all ports, with nothing more than espeak being removed, though pkg showed itself from its amusing side:
=== root@teevee (/dev/pts/0) /usr/src 6 -> Log pkg bootstrap -f
===== Sun 9 Jun 2024 13:26:10 AEST on teevee.lemis.com: pkg bootstrap -f
pkg: Warning: Major OS version upgrade detected. Running "pkg bootstrap -f" recommended
pkg(8) is already installed. Forcing reinstallation through pkg(7).
The package management tool is not yet installed on your system.
Do you want to fetch and install it now? [y/N]: y
Installing pkg-1.21.3...
package pkg is already installed, forced install
Extracting pkg-1.21.3: .......... done
Bootstrapping pkg from pkg+http://pkg.FreeBSD.org/FreeBSD:14:amd64/quarterly, please wait...
Verifying signature with trusted certificate pkg.freebsd.org.2013102301... done
Only the display card was an issue: the computer chose it, and I didn't have anything connected. But on the whole things look quite positive.
|
Understanding Bluetooth
|
Topic: technology, opinion | Link here |
While reading through old diary entries, came across this article, which includes information about using my Bluetooth headphones. I had recently tried to use them, but I couldn't get them to pair, and I had assumed that they were defective. But no, not really, just Bluetooth. Tried the instructions there, and they worked. How I hate Bluetooth! If something goes wrong, there's almost no way to diagnose the problem. Time for a HOWTO page.
|
Hugin loses to one-liner
|
Topic: photography, technology, opinion | Link here |
Jari Kirma is playing with panorama software. Problem: Hugin doesn't run on his Apple machine. So he wrote what he claims is a one-liner in Mathematica that creates panoramas. The result isn't too bad:
|
|||||||||||||||||||||||
How much better could Hugin do? I asked him for the component images and stitched them. Oh:
|
||||||||||||||||||||||||
Even “Perfectly Clear“ didn't help much:
|
||||||||||||||||||||||||
What went wrong there? Another problem with Enblend 4.2? Tried it with version 4.1.4 on eureka, but the results were the same.
What can it be? The photos were taken with a mobile phone (Google Pixel 8 Pro), and presumably the exposures differed from one image to the next. My best bet is that my standard settings don't account for exposure differences. Do I want to investigate further?
|
Lena stumbles
|
Topic: animals | Link here |
While walking the dogs, Elena tripped and fell on the ground and appeared unable to get up. We've seen that before, and at the time we were relatively sure that it was a snake bite. But here there was no snake to be seen. She got up and walked on as if nothing had happened. Is it maybe just an issue that Borzois have more issues with balance than other dogs?
|
Acclimatizing Mona
|
Topic: animals | Link here |
Mona still hasn't come to terms with the dogs. She's no longer as afraid, but she's far from seeing them as harmless, let alone friends.
We've been discussing what to do. We had a cage in Yvonne's bedroom, in which Bruno spent the nights until about a week ago. Now he and Mona have the run of the house. Yvonne wanted to put Mona in the cage in the bedroom, but that's the domain of the dogs. So: bring the cage into the lounge room and let her get used to it. Then we can let the dogs in and see what happens. Here's the current situation:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
hydra: Watchdog timeouts
|
Topic: technology, opinion | Link here |
Seen after some work with despise, the Microsoft VM running on hydra:
Jun 8 09:00:16 hydra kernel: re0: watchdog timeout
Jun 8 09:00:16 hydra syslogd: sendto: No route to host
Jun 8 09:00:16 hydra kernel: re0: link state changed to DOWN
Jun 8 09:00:16 hydra syslogd: sendto: No route to host
Jun 8 09:00:21 hydra kernel: re0: link state changed to UP
That's presumably associated with the hangs that I have been seeing.
| Monday, 10 June 2024 | Dereel | Images for 10 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Beijing sauce noodles again
|
Topic: food and drink, opinion | Link here |
Two weeks ago I cooked a first attempt at Beijing “fried sauce” noodles (炸酱面). The results were interesting, but there were issues.
In principle they're surprisingly similar to a noodle dish from another national capital, Kuala Lumpur. They're both noodles in a dark sauce. The difference that I saw were: the Beijing noodles use different aromatics, notably star anise, sichuan pepper and cinnamon. They also use beans instead of soya sauce and caramel, and the noodles are different. Also, the Kuala Lumpur noodles use squid and prawns, while the Beijing ones (at least my base recipe) don't.
Today I thought of making a hybrid, but in the end I came out with a modified version of the Beijing noodles. Add squid, prawns and choi sam, use the same aromatics, double the sauce ingredients. The result looked pretty much the same, but it was much more substantial:
|
|||||||||||||||||||||||
And how was it? OK, but there's still a way to go. I forgot the raw garlic. And I think my choice of Shanxi planed noodles (not in the original recipe) was suboptimal: they taste too floury. Possibly things would be better with more sauce.
|
A new mobile phone
|
Topic: technology, language, opinion | Link here |
After considerable investigation, I have decided on a new mobile phone, a Xiaomi Redmi Note 13 with 5G coverage to maximize its useful life, though the 5G versions cost significantly more . Why that one? Fatigue, mostly. There are so many phones to choose from, and there's little to distinguish them. This one was one of the relatively few on sale from JB Hi-Fi at a price that I was prepared to pay ($479, discounted from $549). Why JB Hi-Fi? They're the only shop in town, they have relatively few models, and they have presumably chosen the most usable models. And a comparison with my current phone shows it to be better in all respects.
OK, what do they really cost? On eBay they start at $464.59, not that much cheaper than JB Hi-Fi. But there are discounts (“extra 70% off with coupon”, “extra 12% off with coupon”) from the same seller, depending on colour! But they lie: the 70% off is limited to $25, which is less than 12%. And the 12% applies only to white phones.
So: do I pay $408.84 for a white phone, or $439.59 for a black or green one that is otherwise identical? I'm not a dedicated follower of fashion. The white one it is, though I can understand why it's cheaper.
So what do I call it? Apart from other considerations, hirse.lemis.com is still going to be useful for everything except phone calls, so I need a new name. albo.lemis.com? Asked Gemini, which made a number of suggestions, including (in draft 2) alba, clearly an indication of lack of understanding of grammatical gender. So until proof of the contrary, albo it is.
|
Hugin fail
|
Topic: photography, technology, opinion | Link here |
I was really unhappy with yesterday's results with Jari Kirma's panorama. Spent some more time looking at it today. There are 30 component images, and they have an exposure range of EV 11.4 to 9.8 (at 15/12.8° ISO!), which seems to be appropriate. Here the unretouched first and last image:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
DxO PhotoLab has no trouble with the images. Here the thumbnails:
|
|||||||||||||||||||||||
But Hugin makes a complete mess of the exposures:
|
|||||||||||||||||||||||
Things were so bad that the out-of-box Hugin wasn't even able to align the images. What's wrong here? The source images are available at Day/20240610/tapiola-panorama-photos.zip.
|
Tapiola sinfonietta?
|
Topic: music | Link here |
One of the relatively well-known Finnish musical ensembles is the Tapiola Sinfonietta. I had never known where Tapiola is, and so the title to Jari Kirma's panorama came as a surprise. It proves to be part of Espoo, where Jari lives, and in fact it seems that he, too, is in Tapiola, only 260 m from the Tapiola Hall, officially Espoo Cultural Centre.
| Tuesday, 11 June 2024 | Dereel | Images for 11 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More teevee progress
|
Topic: technology, multimedia, opinion | Link here |
The only thing missing with teevee was the display card, and I put that in the “too hard” basket on Sunday. Today I got round to doing it. Yes, connect it to the monitor via HDMI and it comes up normally, without the dead display issues that I have had so often on tiwi.
Starting X was a different matter, of course. I need drivers, and that's the usual horror with the Nvidia site. But now we have artificial intelligence. What does Gemini say? Once again something helpful: install the nvidia-driver-470 package. And yes, that worked! Gemini seems to be more and more useful, though I'm sure I'll soon run into limitations.
But I still need a mouse, and while trying to find a combination of mouse and dongle (which ultimately failed) I ran into a panic!
panic: ffs_blkfree_cg: freeing free block
cpuid = 1
time = 1718080794
KDB: stack backtrace:
#0 0xffffffff80b7fa6d at kdb_backtrace+0x5d
#1 0xffffffff80b32691 at vpanic+0x131
#2 0xffffffff80b32553 at panic+0x43
#3 0xffffffff80e4d5c0 at ffs_freefile+0
#4 0xffffffff80e4a08a at ffs_blkfree+0x9a
#5 0xffffffff80e7454c at freework_freeblock+0x5fc
#6 0xffffffff80e67178 at handle_workitem_freeblocks+0x168
#7 0xffffffff80e602cd at process_worklist_item+0x24d
#8 0xffffffff80e5aa3d at softdep_process_worklist+0xed
#9 0xffffffff80e5e81f at softdep_flush+0x11f
#10 0xffffffff80aeca7f at fork_exit+0x7f
What's that? The file system was clean. Or at least fsck said so. Reboot and run fsck -f, bypassing soft updates. And how about that, it found a whole lot of inconsistencies. Moral: after copying a file system with dd, don't rely on soft updates to still be consistent.
|
Mona accepts cage
|
Topic: animals | Link here |
We've had the cage in the lounge room for a couple of days now, but so far Mona hasn't shown much interest. But today she did:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
There are also some indications that she's finally coming to terms with the dogs. Today Larissa came within 40 cm of her while she (Mona) was sitting on the rocking chair, and Mona didn't run away. She wasn't very happy either, but there is some progress.
|
More dead elephants
|
Topic: general, animals | Link here |
Big crash from the lounge room later. What I found was:
|
|||||||||||||||||||||||
That must have been Bruno's work. Fortunately nothing was damaged. A good thing that he's finally calming down a bit.
|
Understanding infrared thermometers
|
Topic: general, technology, opinion | Link here |
Found yet another old infrared thermometer in my office today. Put in some batteries and tried it out. It worked, sort of, but showed completely random temperatures up to 10° apart. I recall problems with it 7 years ago, and at the time I thought it was due to moisture. But clearly there's more to the accuracy of infrared thermometers than meets the eye.
| Wednesday, 12 June 2024 | Dereel | Images for 12 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Installing another air conditioner
|
Topic: Stones Road house, general, technology, opinion | Link here |
Nathan Barnes and his apprentice Justin along this morning at 8:00 to install a Mitsubishi MSZ-AP-60VG2 air conditioner. As if the early hour wasn't bad enough, the remote controlled front gate had got jammed, not for the first time, and I had to go out into the freezing cold and dark with only a dressing gown to un-jam it.
They got stuck into it pretty quickly. About the only issue was that we had to briefly take house power down, so nearly all the computers had to be powered down. Once again lagoon survived on the UPS, but the other machines didn't have one. I took the opportunity to put them in circuit, only for the main machines.
By 15:30 they were done and gone, and the air conditioner was working:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Nathan also continued his investigation of the Actron Air unit. The inside part of the circuit had lost more than half its pressure over the last week, and he found that the coil of the inside unit was once again clogged with gunk, but also had a massive leak. It'll need to be replaced, and for that I'm looking at round $3,000, roughly what we paid to have the new unit installed.
Why? Like Tony Nesci before him, he had never seen anything like it before, and he, too, thought that the material was dog hair until he looked at it more carefully. It's almost certainly spider's webs. Are the two issues related? On the face of it, no: one is dirt, the other is damage to the piping. But we should really look at stopping either happening again. Mice chewing up the duct between return air and inside unit? Maybe, but the clogging doesn't look like mice. In any case, he can't do it this month: he let a lot of other stuff lie to do our installation, and now it's his other customers' turn. July? August? At least we have a stop-gap, even if it's not exactly a replacement for the Actron unit.
|
Google: You have violated our policy
|
Topic: technology, general, opinion | Link here |
Sent a message about the Tapiola panoramas to the Hugin mailing list today. The response was swift and unexpected:
We're letting you know that we've permanently removed the content at https://groups.google.com/d/msg/hugin-ptx/6cIfYhbkz74/yG2lQnBmBgAJ
Why did this happen?
An external report flagged the content for illegal content or policy violations. As a result, our legal content and policy standards team removed the content for the following reason: unwanted content.
Learn more
You can learn more about our content policies and enforcement at our help center.
What you can do next
You may have the option to pursue your claims in court. If you have legal questions or wish to examine legal options that may be available to you, you may want to consult with your own legal counsel.
What an arrogant, inaccurate and stupid response! What “external report”? What inappropriate content? Why a URL when they have deleted it? And my only recourse is proving my claims in court. Google, I have found you obnoxious for some time, but this is extremely insulting. “Don't be evil”, indeed!
What was their bot objecting to? I guessed, without much evidence, that it was the number of image URLs that I put in the message. Rewrote it without those URLs and it was accepted. What a pain!
|
New photo toys
|
Topic: photography, opinion | Link here |
I've been taking photos now for 60 years. This unremarkable photo was taken at the Bath Festival on 11 June 1964, 60 years ago yesterday:
|
|||||||||||||||||||||||
The camera was a Voss Diaxette, a particularly primitive camera with no light meter. How did I get the exposure right? It wasn't just this image: all the photos I took looked relatively well exposed.
But recently I was reminded: I had a Zeiss Icon Ikophot exposure meter, and recently I found one on eBay for a price I could afford. It arrived today:
|
||||||||||||||||||||||||
It's really quite a pretty thing. Interesting to note the shutter speeds in the old series 1/25 to 1/125, and that when I received it the thing was set to 40 ASA. Or was it? It also shows 18° DIN, which is now 50 ASA. On the other hand, the sensitivity range is from 3 (!) to 800 ASA or 30° DIN.
|
||||||||||||||||||||||||
So is the dial slightly off-centre? You wouldn't think so for a device of this quality.
It's surprising how little reliable information there is on the device. This Camerapedia page gives a feeling:
The first Ikonophot [sic] 1328/1 light meter was manufactured in Dresden by Zeiss-Ikon A. G. in 1339.
But wait, there's more! An Ernst Leitz Wetzlar GmbH [German, no English version available] Hektor 13.5 cm f/4.5 and a 20 cm f/4.5 Telyt:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
According to this serial number list, the Hektor was made in 1951, and the Telyt was made in 1956. That makes the Hektor the oldest lens that I have, now 73 years old. Last week I found another site which gives me the same years.
What can I do them? Both of them fit my FED 1, but that won't help with the Telyt, which, as I only discovered later, is designed for a Visoflex and thus requires 62.5 mm distance from the camera flange.
One of the reasons for this purchase is that my FED has a film in it, and potentially there are interesting images on it. The lens for the FED is 50 mm f/3.5 Fed lens with no apparent serial number. It appears to be an Elmar copy, and the diaphragm is jammed at f/4. So if there are any exposures left on the film, I'll take them with the Hektor.
What do I do with the Telyt? I could theoretically mount it on my (Pentax) bellows, which compress to 45 mm, and use it on my Olympus cameras. I have an adapter from M42 to Micro Four Thirds, but I'd also need an adapter from M39 to M42. Is it worth it? Surprisingly it appears to be the widest aperture lens that I have at that focal length, though the Zuiko Digital ED 35-100 mm f/2.0 would beat it with a 2x teleconverter.
|
Mini hamburger bun
|
Topic: food and drink, opinion | Link here |
One of these silly things that we can't find in Australia are sensible hamburger buns. The ones we do find are far too soft. Today Yvonne brought some “mini sourgdough” buns, so we ate hamburgers again. Mini? We couldn't get our teeth round them, and Yvonne left half of hers for breakfast:
|
|||||||||||||||||||||||
| Thursday, 13 June 2024 | Dereel | Images for 13 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Warmth!
|
Topic: Stones Road house, general, animals, opinion | Link here |
Finally we have air conditioning again, after nearly 2 weeks:
|
|||||||||||||||||||||||
Wall-mounted air conditioners are never as uniform in their heat distribution as ducted systems, and there were temperature differences of up to 3°. I soon improved on that, but in the meantime the cats had found the warmest place:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
After adjusting the position of the vents and moving the cabinet beneath to one side, things worked better:
|
||||||||||||||||||||||||
|
Leica to M43 adapter
|
Topic: photography, opinion | Link here |
Four days after buying the Leitz lenses on 3 June 2024, I ordered a Leica to Micro Four Thirds lens adapter. The lenses came from Perth and took 9 days to arrive. The adapter came from Shenzhen and arrived today, after 7 days. What's wrong with Australia Post?
So of course I tried it out:
|
||||||||||||||||||||||||
Yes, it does what it says. But there are a couple of issues. First, the screw thread doesn't register correctly, so the tripod mount of the Hektor was off-centre:
|
||||||||||||||||||||||||
And I couldn't focus with the fake Elmar. Like many older lenses, it has an infinity lock that needs to be depressed to move from the infinity position:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
And the body of the adapter is too wide, getting in the way of the release pin:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
You'd think that they would have noticed that and fixed it.
|
Front gate issues
|
Topic: Stones Road house, general | Link here |
Our electric front gate has jammed a couple of times recently. Before we connected the opener, there was a conventional chain, and for some reason the wind has been blowing the gate into a more “closed” position on the wrong side of the chain hook:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Presumably a bigger hammer will fix that.
|
Quora: regurgitating?
|
Topic: technology, general, opinion | Link here |
It's pretty clear that Quora is on its last legs, and it has been for some time. Recently I have been receiving upvotes for old replies, notably “Should I buy the Olympus OM-D E-M5 or the EM1?”, which I answered in November 2016. Why am I getting upvotes now? My guess is that Quora picked the reply out of a hat and presented it to numerous people. Until last week I had had one upvote; now it's 14.
| Friday, 14 June 2024 | Dereel | Images for 14 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Peace!
|
Topic: general, photography | Link here |
Somehow the last couple of weeks have once again been really tiring. Today I did nothing much of importance apart from play around with my new photo toys.
|
Special offer on mobile phones
|
Topic: technology, opinion | Link here |
Seen on hirse (my current mobile phone, soon to be retired):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
“Here's your chance to get this item at a great price!”
What's wrong with this advertisement? It's from the seller who sold me exactly this phone. Why should I want another one? And why hasn't he sent the one I bought? And why should I pay over $40 more for this “special offer” than I paid for the one I bought? And why the option of paying either $450.65 or “Approx.” $276.97?
And where is the one I bought? Somewhere, maybe. Australia Post gives me a choice:
|
|
|||||||||
|
|||||||||||||||||||||||
The first is from the web site, the second from the Android app that I had vainly hoped would allow me to scan the tracking numbers. And what does the ¹ mean after the time? How do I select it? My guess is that it indicates that the app uses a 24 hour clock, just to confuse people, and the ¹ is for additional confusion.
| Saturday, 15 June 2024 | Dereel | Images for 15 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Comparing lenses
|
Topic: photography | Link here |
Spent much more time than expected today comparing my “Leica” lenses. Well, two of them: I couldn't do much with the 20 cm f/4.5 Telyt, which is designed for a Visoflex, and it needs an extension of 62.5 mm. I tried it with my extension tubes (26 mm) and hoped that the focus adjustment would add enough to allow me to focus at infinity. It didn't. So until I find a different way of mounting it, it's out of the game.
So: that leaves the Hektor (13,5 cm or 135 mm) and the ФЕД (FED) 50 mm. What should I compare them with? My modern lenses, of course, say the Olympus M.Zuiko Digital ED 12-200 mm f/3.5-6.3 set at 135 mm and the Zuiko Digital ED 50 mm f/2.0 Macro. And for an intermediate choice the 135 mm f/2.8 Exaktar and 50 mm f/1.4 Super-Takumar. Each of these lenses should be set to maximum aperture, f/4.5 (maybe f/4) and f/11.
Things were much more complicated than I expected. It took me an hour! First, it proved that the maximum aperture of the 12-200 M.Zuiko at 137 mm (the closest I could set) was f/6.3, and I wasn't able to set the aperture on the Exaktar. So change lenses: Leica DG Vario-Elmar 100-400 mm f/4.0-6.3. And for the 135 mm? I don't have anything else. So I chose the Nikkor 105 mm f/2.5, after which it made sense to use the Nikkor 50 mm f/1.4 instead of the Super-Takumar.
Then there's the question of motive. In principle I had chosen the front of the house:
|
|||||||||||||||||||||||
That's taken with the Hektor at f/4.5. But I couldn't choose that view for the 50 mm FED: the issue with the infinity lock on the adapter precluded focusing at distance. I was able to unlock it before mounting on the adapter, but then I couldn't focus to any distance. So I had to choose a different motive:
|
|||||||||||||||||||||||
That's the FED with the only aperture it knows, f/4.
The adapter also caused other problems, which I was able to work around:
|
|||||||||||||||||||||||
And then there's just the messing around with the lenses and adapters. But I got it done, leaving a considerable mess behind.
|
|||||||||||||||||||||||
And the results? The Hektor results look better than I had been led to expect. Only the FED stands out as being fuzzy, but then, that was the full aperture. I'll look at the nitty-gritty Some Other Time.
|
Another hydra crash
|
Topic: technology, photography | Link here |
Lots of photos to process today: not just the lens comparisons, but the weekly house photos. And once again hydra hung while processing them with the Microsoft VM. Only this time no watchdog timeout occurred, and I had to reboot it.
OK, another chance to try the net/realtek-re-kmod driver. Two months ago it failed to load, probably because the machine is up and running. Today I tried it in single-user mode, and it worked. But the on-board interface came up as re1, requiring some reconfiguration.
And then my virtual machines didn't work! It proved that they, too, were looking for re0. Fixed that, and despise came up and limped. Something's strange there.
|
Relocating video files
|
Topic: multimedia, technology, opinion | Link here |
tiwi, my TV computer, is continually running out of space. Currently I have three disks connected, with 8 TB, 6 TB and 4 TB. There's still space available, but the biggest consumers are films (166 GB) and series (6.4 TB). Clearly the series alone are almost enough to fill the first disk. I can put the others elsewhere, notably on the third disk (/VB3) which has only about 1 TB.
But how do I maintain the illusion of a single file system? VB3 is mounted on hydra, and I access it via NFS. Deep links!
=== grog@tiwi (/dev/pts/13) /spool/Series 16 -> df
Filesystem 1048576-blocks Used Avail Capacity Mounted on
/dev/ada0p4 7,567,870 7,459,980 32,211 100% /spool
hydra:/VB3 3,815,019 3,696,084 80,784 98% /VB3
/dev/da0p1 5,722,572 5,659,737 5,609 100% /VB2
hydra:/VB3/spool/Videos 3,815,019 3,696,084 80,784 98% /spool/Videos
hydra:/VB3/spool/Images 3,815,019 3,696,084 80,784 98% /spool/Images
It's not elegant, but it might be better than the alternatives. Oh for Vinum! Does ZFS allow file systems to span physical media? Surely it does, but just finding out is difficult.
| Sunday, 16 June 2024 | Dereel | Images for 16 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Cat photos?
|
Topic: animals, photography, opinion | Link here |
Woke up round 1:00 this morning and found Mona and Bruno lying on an armchair together. That's only the second time, and worth a photo.
But I failed. My new Olympus OM-D E-M1 Mark I didn't do anything like what I wanted. It wouldn't confirm focus, and when I pressed the shutter nothing seemed to happen, though the flash units went off seemingly at random. And of course by the time I changed cameras, Bruno had got up to investigate what I was doing.
What happened? When I got up, I checked. Yes, all settings correct. Normal test photos:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Well, almost. That first photo was out of focus. As in the middle of the night, I couldn't get it to focus. And it repeated 8 times over a few seconds: this was the last of the sequence. Checking again I found that the camera was set to 7 exposure bracketed images. How did that happen? I'm sure that it was correct the first time I checked. But yes, the photos of the cats were also repeated.
Hopefully that doesn't happen again.
|
Fake fake teriyaki
|
Topic: food and drink, opinion | Link here |
A while back I bought some “Teriyaki” spice pastes from ALDI. They're from the same series as the fake Pad Thai that I use for my even faker Pad Thai recipe.
|
|||||||||||||||||||||||
So why haven't I tried the Teriyaki? I don't really know what it is, and I don't normally have the ingredients (capsicum and broccoli). Finally I got hold of some, and only then did I check out what Teriyaki really is. I don't see much relationship with the ALDI paste.
Still, tried it:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Nothing to write home about (except that's what I'm doing). Not enough, and the vegetables have nothing to do with real teriyaki.
|
More hydra pain
|
Topic: technology | Link here |
Somehow the change of the network configuration on hydra has caused problems for VirtualBox. I had a lot of trouble connecting to my Microsoft VMs. Sometimes it worked, sometimes it didn't, and on a couple of occasions I ended up with rdesktop killing the session, apparently because I had started two, and the one left out in the cold gave up and somehow terminated the session.
In addition, there are issues with the audio, background noise that seem to be related to what the CPUs are doing. They're particularly evident when the Microsoft VMs are working.
I've had hydra for nearly 9 months now. Somehow things should have been easier.
| Monday, 17 June 2024 | Dereel → Ballarat → Dereel | Images for 17 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Chili sauce again
|
Topic: food and drink, opinion | Link here |
Huevos rancheros for breakfast today, and again I had no chili sauce. Made another batch of the Gochujang-based chili sauce that I made on 28 April. Somehow the amount of sesame oil (50 g for 100 g gochujang) seemed excessive, so I reduced it to 10 g. And I forgot the vinegar, which didn't seem to do any harm.
|
Glasses and vaccinations
|
Topic: health | Link here |
My new phone arrived today, so planned a journey into town, incorporating the vaccinations (COVID-19 and flu) that Yvonne had been nagging me about for a while.
Oh. Health First, my GP, don't do walk-in vaccinations. I could get an appointment at 16:00, but that didn't suit me. Mañana.
So into town, first to the Ballarat Indian grocery in Dana Street. All I wanted was a particular kind of beans. Do they have them? Probably. But to find out I would have to park, and that costs money. The beans aren't that urgent, so I put it off until I really needed them.
Then to Specsavers to have my glasses adjusted, which was remarkably simple. On to UFS to see if they could vaccinate me. No, COVID is only Tuesdays and Thursday. Oh, alright, you'll have to wait 12 minutes. “Appointment” at 15:00. Just enough time to go across the road to ALDI to buy a SIM card for my old phone (the old one goes into the new phone). Back just in time, to receive an SMS:
Hi GREGORY LEHEY, Booking confirmed for 17/06/2024 03:50 PM COVID-19 Pfizer (UN7ZU7) @ Sebastopol UFS Pharmacy. Save time by answering questions before appt https://www.medadvisor.com.au/booking/prescreen/U...
15:50? Ha ha, only joking. I was seen to almost immediately, and I was out by 15:07.
|
New mobile phone
|
Topic: technology, general, opinion | Link here |
The real reason for today's journey was to pick up my new Xiaomi Redmi Note 13. Back home with it and set it up with relatively little difficulty, though it took quite a while.
First observation: it was cheaper because it was white (and thus the name albo.lemis.com). But the protective cover is opaque and anthracite black. Only a little bit round the lenses shows the natural colour.
And though it copied a lot of stuff across, it didn't copy everything. A number of apps weren't copied, and random things like my ring time didn't make it either. And the home screen is a complete mess. I'm reminded of my eternal migration from eureka to hydra. Still, it wasn't too painful—yet.
And then registering the new SIM card for the old phone. Why? I won't be able to make calls after the end of August. But the phone can do LTE, so hopefully mobile data will work. ALDI says
And it's as simple as....
I have had different experiences, and once again I was right. Please supply your driving license numbers, both of them, along with a diagram showing that there was only one number. Put the same number in both fields, which satisfied the entry criteria, but
We’ve been unable to verify your identity electronically.
So I had to provide my credit card number, after which they were satisfied. And within only hours, the number was activated. And making calls from the phone now no longer claims that it will stop working at the end of August!
| Tuesday, 18 June 2024 | Dereel | Images for 18 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Under the weather
|
Topic: health | Link here |
I don't react strongly, but on the day after a COVID-19 vaccination I somehow feel less than active, and today I did very little.
|
Flowers in late winter
|
Topic: gardening | Link here |
It hasn't been warm lately—the weather forecasters speak of the coldest snap in years—but various flowers are still coming, notably this Dutch Iris:
|
|||||||||||||||||||||||
The Abutilons next to it are also preparing for a large show of flowers:
|
|||||||||||||||||||||||
|
More fun with albo
|
Topic: technology, general, opinion | Link here |
Spent a lot of time playing around with albo.lemis.com, my new Xiaomi Redmi Note 13. Copying things from hirse was only partial: yes, it seems that most apps got installed, and some of them also retained the configuration information. But only some of them; others required reconfiguration.
One of the apps that I knew I would have to install myself was the Mendhak GPS logger, which also require installing F-Droid. That was strange: F-Droid spent a considerable amount of time building up a repository, not saying that it hadn't completed, and until it was, I couldn't find the GPS logger. But when it was installed, it included all its configuration information! So there's something very strange going on there when normal apps lose their configuration and non-sanctioned apps keep theirs.
That was also the case with the home screen. It was completely different from the one on hirse, reminding me of the issues I had changing the window managers from eureka to hydra. I gradually managed to get it to look similar, though it seems to be too polite to fill the screen. And setting the icon size shows the stupidity of primarily graphic displays: on hydra it's in Settings/Home Screen. But that's too obvious: on albo it's Under Settings/Wallpaper & personalization/Icons. Obvious, right?
|
Google pain
|
Topic: technology, general, opinion | Link here |
Somehow Yvonne managed to get herself locked out of her Google account today. How do we get it back? Send a code to her mobile phone.
Where's her phone? Out in the paddocks:
|
|
|||||||||
Out to take a look, calling from my phone in the process. Nothing.
Is it really there? Or maybe in the stables? Back into the office. ping fossil. Normal ping responses. That's not possible that far from the access points. Tried again. It was in Yvonne's office! And that in plain view. But Google maps continued updating the false location for hours, until the next morning!
How can that work?
|
Fray Bentos steak and kidney
|
Topic: food and drink, opinion | Link here |
What does a real Steak and kidney pie taste like? Last week ALDI had one on sale, from Fray Bentos, a company I know but don't overly respect from the UK.
Still, it didn't look bad, so it was worth trying. It's in a metal can, and it's designed to be baked. The result, after 24 minutes at 180° in the “hair dryer” “air fryer”, didn't look bad:
|
|||||||||||||||||||||||
But there were tell-tale signs on the edge:
|
|||||||||||||||||||||||
And getting the pie out of the metal container was a catastrophe:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It didn't taste that bad, though Yvonne thought that there was too much pastry and not enough filling. We won't buy it again, but it was interesting. Considerably moister than our own.
| Wednesday, 19 June 2024 | Dereel | Images for 19 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Cold night
|
Topic: general, opinion | Link here |
Very cold night last night, with a lowest temperature of -2.3°. It showed in the air conditioner's attempt to heat:
It's not just that it took 6 hours to come up to normal room temperature: the room temperature graph (blue) shows clear steps, indicating pauses for de-icing. And the view outside when I got up was corresponding:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Mona at two months
|
Topic: animals, general | Link here |
We've had Mona for two months now. She still hasn't settled in completely, mainly because of her fear of the dogs. That's subsiding, but it's still there.
And with Bruno? Nothing like what we expected. Part of the problem was that Bruno had never seen another cat from age of 14 weeks until Mona came along, and he was used to playing with dogs. That's a lot rougher than Mona likes, and it has taken Bruno some time to become a cat again. But with few exceptions they're still not lying together. Will they ever? It might just end up being a different relationship. I think the next breakthrough will be when Mona comes to terms with the dogs.
|
More fun with albo
|
Topic: technology, general, opinion | Link here |
More playing around today with my albo.lemis.com, my new Xiaomi Redmi Note 13. It proved that the Mendhak GPS logger was not logging. Why? It seems that the configuration was copied, but the write permissions weren't.
And then there's this irritating message “At the end of August, it's likely that this phone will not be able to make calls”. That's from the old phone, but the SIM is now in the new phone. I had called ALDIMobile support yesterday (2435 from subscribing phones), and Vince had told me that he would raise a ticket to have the message cleared.
But today it was still there, and it's irritating. It takes about 15 seconds, and it delays every phone call. Called up again and spoke with Melanie, who had difficulty understanding the concepts, but told me that I would have to put up with it until the end of August. I told her that that was not acceptable, but she said there was no alternative. Complaint? Can't take a complaint. Supervisor? Hold on. Two minutes of complete silence, then I was disconnected.
Another call, answered by Austin. Yes, that's normal for the first 48 hours. But I've had this phone (specifically Xiaomi Note 13 5G) for 5 days! Oh, then check that VoLTE is enabled. Xiaomi support will help.
Three different answers! It wasn't until I hung up that I realized that I had had the phone for less than 48 hours. So let's wait a bit first. Also checked VoLTE. Lots of videos showing how to do it, like this one:
That includes this view:
|
|
|||||||||
But that's not what I saw:
|
|||||||||||||||||||||||
Why do these things always look different? Do I have different firmware? At some point I do: when I got the phone, I upgraded apparently from MIUI to HyoerOS. And making calls with Wi-Fi may be an advantage. But why should I lose VoLTE in the process?
A little later it occurred to me that my phone is a Redmi Note 13 5G, and of course it's different in this area. Is that the reason?
The obvious thing to do now is: nothing. Wait a day or two and see if this irritating message goes away, as Austin suggests.
|
More app pain
|
Topic: technology, opinion | Link here |
Another app that didn't get transferred to my new phone was “WiFi File Transfer”. Why not? It's gone! Well, there's another with the same name. It's marvellous!
Transfer files and folders from phone's storage media to other platforms such as PC is an awesome feat. Just wifi, not needed a cable.
What are we coming to when a simple file transfer is “awesome”? And this sentence no verb?
OK, install that. Not what I expected:
|
|||||||||||||||||||||||
And somehow it managed to install without providing an icon. Certainly not what I expected. What are we coming to?
|
Redmi Note 13 camera
|
Topic: photography, technology, opinion | Link here |
The new Xiaomi Redmi Note 13 boasts a 108 MP camera! That's more than any of my real cameras, even in Hi-Res mode. How good is it? While playing in the lounge room, took this forgettable picture, along with a natural-sized crop:
|
||||||||||||||||||||||||
|
|
||||||||||
It's not bad under the circumstances, but the most interesting part is the resolution: 12 MP. How do I get 108 MP? Do I get 108 MP? Do I care?
| Thursday, 20 June 2024 | Dereel | Images for 20 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Consequences of frost
|
Topic: gardening, general, opinion | Link here |
Yesterday's frost left its mark:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The Cannas aren't important, but the Hibiscus rosa-sinensis “Uncle Max” looks very unhappy, and doubtless the buds won't flower. Hopefully it will be better than a couple of years ago.
|
Garage door fail
|
Topic: Stones Road house, general, opinion | Link here |
Yvonne had an appointment in town today—she thought. It proved to be for tomorrow, but she didn't find that out until she was there.
To get there, of course, she first had to get the car out of the garage. And the garage door opener didn't work! We had to use the emergency pull, which looks underdimensioned for the effort required:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Try as I might, I couldn't get the door open completely. So Yvonne drove out, while I held on to it. When I let go, it returned to this position:
|
|||||||||||||||||||||||
It wasn't until much later that I discovered that this is the normal “open” position.
So what caused the problem? Tripped circuit breaker, again from the septic tank pump (thanks, Jim Lannen, I'll never be able to forget you). Reengage the door into the automatic mechanism. That's described on the pull cord:
|
|||||||||||||||||||||||
But it didn't engage. Things really look like this, first disengaged, then engaged.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The important thing is the orange bead. But it didn't engage in the chain. It took me a while to understand that it has to engage in this part of the chain, requiring the machine to bring it into position:
|
|||||||||||||||||||||||
After that, all was clear. But it took me 20 minutes to find out.
|
Pump problems
|
Topic: Stones Road house, general, opinion | Link here |
One of the external units on the circuit breaker that tripped was the bore water pump. It had no pressure, and pressing the prime lever had no effect. Taking off the top of the contacts was a surprise:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
How do I clean that?
|
More albo fun
|
Topic: technology, general, opinion | Link here |
So gradually albo.lemis.com is working roughly the way I want it, or as close as you could expect from an Android device. As Austin predicted, the irritating “this phone will not work after August 31” message is now gone. Tried once again to get something useful from “WiFi File Transfer”. Now it can see the device IP address (awesome!), but it can't see itself. Start another app and it no longer shows in the list of active apps, and it hides itself by calling itself “FTP File Transfer”. If I try to restart it, it just notes that its “HTTP Port” 1234 is in use, but not that it has it open:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But the change in name is an indication: this isn't an HTTP connection at all:
|
|
|||||||||
So it's completely useless.
While searching for other alternatives, came across an old acquaintance: AirDroid. I used it decades ago. But it required login, something that I didn't like, and some time later I ran into other problems and replaced it with the now-defunct functional version of WiFi File Transfer.
OK, try again. I don't understand what they're talking about! I need a client to access it! And of course there is no client for FreeBSD. What happened to the Internet Protocols? I can't even set a port number. But looking back, I saw an example:
|
||||||||||||||||||||||
Just access flachmann.lemis.com:8888. OK, flachmann is gone, but what about albo? Yes! It works! And it seems to be completely undocumented, something that I commented about at the time. And in passing, that's a good vindication of the detail that I keep in this diary: I can't find anything in the current documentation that points to this (most logical) means of access.
Of course, it's still all graphical, and that's why I didn't use WiFi File Transfer much: that's why I've been using WiFi FTP Server for the past 4 years. But it proves that AirDroid does have some redeeming features: in particular, it can display SMSs! But somehow all these apps seem to live in a parallel universe. I still don't understand. I created a new account, since it had forgotten the old one, but all the welcome message (“Welcome to AirDroid?”; question mark indeed) offered was a downloadable client for Microsoft or Apple, or a web client. But to sign in I needed to fill in a CAPTCHA. Sorry, AirDroid, I don't accept that kind of insult.
One other thing that it offers was an opportunity to shoot myself in the foot. I can no longer find it. Possibly it's this selection with the unexpected icon:
|
|
|||||||||
It caused a YouTube display to take over the monitor (the TV in the lounge room), and nothing I could do would get me back to my X display. After some frantic searching, starting another (second) X server and so on, I discovered that it had caused a display on hires.lemis.com (the TV, and not hisen.lemis.com, my old phone), and to get back I needed to select the correct TV input. But it's still not clear what that has to do with YouTube.
|
Emergency location!
|
Topic: technology, opinion | Link here |
One of my favourite gripes with Android phones is how difficult it is to find where they are. That's particularly silly in case of emergency. But now, scarcely more than 15 years after the introduction, it seems that somebody has done something about it:
|
|||||||||||||||||||||||
How does it work? Do I want to know? For no good reason it appears to be limited to emergency numbers (112, 000).
| Friday, 21 June 2024 | Dereel | Images for 21 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
albo: complete?
|
Topic: technology, general, opinion | Link here |
Some minor playing around with my new Xiaomi Redmi Note 13. It's not the way I want it—it is Android, after all—and it's not even the same as hirse, my old phone. But it will have to do.
I don't know whether it's just the increased attention to the phone, or something else, but it seems that there are far more advertisements than on hirse. I came across this page, which gave me some suggestions that might work.
|
“Wise” money transfers
|
Topic: technology, opinion | Link here |
I had to transfer a small amount of money to Germany today. “Wise”, the vaguely named financial service, seems the cheapest. But once again they did everything to annoy me. To complete the transaction, I first needed to install their app on my phone, with all its attendant security implications. And then they used it for “security” verification:
|
|||||||||||||||||||||||
So to perform an action on the phone, they send me a “security number” to the phone, requiring me to jump between two apps. After the event it occurred to me that this was a perfect application for AirDroid. As it was, it was just particularly inconvenient, especially since all their messages for the past year were joined in one big message; AirDroid separates them. But what kind of security is that?
My response to their “how easy was this transaction?” (1 out of 5, the minimum):
This APPALLING app! 2 mail messages, three password entries, and I have violated the security of my phone by installing sensitive information. I'm half inclined to go elsewhere.
|
Draining the oceans
|
Topic: general, opinion | Link here |
Seen today:
| Saturday, 22 June 2024 | Dereel | Images for 22 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Garden flowers in early winter
|
Topic: gardening, general | Link here |
Yesterday was the June solstice, time for the monthly garden flower photos.
Just before the June solstice we had a relatively heavy frost (-2.4°). That shows:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And our long-suffering Hibiscus rosa-sinensis “Uncle Max” has suffered just as it wanted to produce more flowers:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But at least the rear of the bush has escaped the worst of the frost:
|
|||||||||||||||||||||||
But it's also the time where autumn meets spring. This year it's not very pronounced. There aren't too many roses left:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And there aren't too many spring flowers yet, though this Iris is already there:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Our second lime bush is gradually bearing fruit:
|
|||||||||||||||||||||||
And the Abutilon over Piccola's grave is gradually flowering better, though the big one near the laundry door hasn't quite come out yet:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Ubuntu again
|
Topic: technology, opinion | Link here |
Installed the latest version Ubuntu Linux on a virtual machine today.
Why? A couple of reasons: to see whether VirtualBox still works after the change in networking infrastructure that caused disaster and despise, my two Microsoft VMs, to lose connectivity. And to see how well it does out of the box with rendering non-Roman fonts, thus a user-oriented desktop version.
It didn't start well. Configured it with all the defaults, and it just hung on boot. After a while I decided that it wasn't going to do anything, and went searching the web, coming up with Ubuntu 18.04 installation on a VMWare stucks [sic]. The default memory size that VBox selects (only 1 GB) is too small, and Ubuntu is too polite to complain. With a sane amount of memory (8 GB) it ran presumably as intended, still leaving a blank screen long enough to be concerning. And then it discovered that it needed an updated installer. OK, no problems, in the process proving that networking was working.
After that, a pretty screen that told me nothing. No indication what to do next. What do beginners do? I clicked on the “install” icon (twice, please, in case one fails) and it started the installation all over again.
Finally it was done. Blank screen. Ha ha, only joking, that's the screen saver. Log in, and another incomprehensible display:
|
|||||||||||||||||||||||
Where's xterm? How do I talk to the machine? Yes, there's a terminal displayed on the left, but it wasn't when I first connected. And it's some strange xterm substitute that just confused me. It also came up in reverse video, of course. Somehow Ubuntu seems to like black. OK, never mind, what does the network look like? Ah, no ifconfig, that's an old, worn-out magic word. But you can install it.
As expected, it used DHCP. A bit of searching allowed me to find the network settings, and they worked.
But the real point of the exercise was to see whether Ubuntu could display unusual fonts out of the box. Yes! So that's an area where FreeBSD is still lagging behind. In the short term, I can just copy the fonts.
So: I can't see myself actually using Ubuntu—it scares me almost more than Android—but the effort was well worth while. Now to find out what network problems disaster and despise have.
|
Sushi?
|
Topic: food and drink, opinion | Link here |
A while back we bought a “sushi kit“, mainly rice, nori and sushi vinegar. I lost the instructions, so we didn't do anything.
But Yvonne forced my hand. She wanted to eat sushi, so yesterday she bought some in town. Somehow the prices for sushi are horrendous, and it seems that this place has changed owners recently. Yvonne was not impressed, and somehow the sushi itself didn't taste as good as I recall, though I have difficulty deciding what the issue was, apart from the fact that the nori was so tough that I couldn't bite through it. I suppose it matches that their stated web site https://sushinoodleonmair.com.au/ is currently not accessible.
So: time to learn how to do this myself. There are plenty of videos on YouTube. Watched this one, which was instructive:
| Sunday, 23 June 2024 | Dereel | Images for 23 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Frijoles refritos?
|
Topic: food and drink, opinion | Link here |
Cooked another batch of frijoles de la olla (boiled beans) today. They're intended to make frijoles refritos. In principle I'm happy with my recipe, but can it be improved? One thing that occurs to me: it contains lard, something that makes the beans unsuitable for vegetarians, Jews and Muslims. But why? Frying the beans also takes (significantly more) lard. Why not add it all then?
Off on the web looking for Spanish language recipes to avoid the US American bean paste. And how about that, I found this page, which goes into significant detail. It agrees with me that the beans should be crushed, not puréeed, so it must be right. But there are a number of details that I didn't know, including adding onion and garlic. And it doesn't mention anything specific about boiling the beans. I'll look at it again next time I fry.
| Monday, 24 June 2024 | Dereel | Images for 24 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Huevos a la tigre: omelette?
|
Topic: food and drink, opinion | Link here |
Huevos a la tigre for breakfast today. It's really a kind of omelette. Or is it? Not after my once-non-stick frying pan had its way:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Mobile phone power consumption
|
Topic: technology, general, opinion | Link here |
I've been playing more than usual with albo.lemis.com, my new Xiaomi Redmi Note 13 mobile phone. And the power consumption differs markedly from hirse.lemis.com, my old Redmi 9T. And it shows in the power consumption. Here first albo, then hirse (run the cursor over an image to compare it with its neighbour):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Is the Note 13 really that much hungrier than the 9T? Or is it just that I have been using it more?
|
Understanding Microsoft issues with VirtualBox
|
Topic: technology, opinion | Link here |
Saturday's installation of Ubuntu confirmed that there wasn't much wrong with the VirtualBox network configuration on hydra. OK, what's the issue with Microsoft? Spent some time investigating.
First, can it be that the VMs didn't like the host configuration changing under their running image? Rebooted despise. No improvement. OK, for the fun of it, try the “troubleshooter”, which has never shot any trouble for me. But this time it came up and said “done!”, specifically “restored default gateway”. That's nonsense, of course: I was trying (and failing) to ping the default gateway. But of course I tried again anyway. And it worked! Fixed, but with the wrong explanation.
OK, on to disaster, the other clone, almost exactly the same configuration bar the memory size. This time I didn't reboot. And though it claimed to have fixed the problem, this time it didn't. OK, reboot. And yes, it tried to fix it again, but set up DHCP. Maybe that's what it did before, and I didn't find it because I was looking for disaster, not dhcp-242. But after configuring the interface on the Microsoft side to the correct IP address, I wasn't able to ping eureka. Run the troubleshooter again. “Ethernet 2” (there is no “Ethernet 1”) didn't have a valid IP configuration, whatever that means. It fixed it. At some asynchronous point I had connectivity again.
|
|
|||||||||
While pondering that, discovered that I could no longer contact despise. Another fix, then ran a ping from eureka for both VMs. That was instructive. After about 3 minutes, I got:
64 bytes from 192.109.197.168: icmp_seq=177 ttl=128 time=0.313 ms
64 bytes from 192.109.197.168: icmp_seq=178 ttl=128 time=0.302 ms
64 bytes from 192.109.197.168: icmp_seq=179 ttl=128 time=0.258 ms
64 bytes from 192.109.197.168: icmp_seq=196 ttl=128 time=13397.802 ms
64 bytes from 192.109.197.168: icmp_seq=197 ttl=128 time=12396.873 ms
64 bytes from 192.109.197.168: icmp_seq=198 ttl=128 time=11393.059 ms
64 bytes from 192.109.197.168: icmp_seq=199 ttl=128 time=10389.815 ms
64 bytes from 192.109.197.168: icmp_seq=200 ttl=128 time=9388.793 ms
64 bytes from 192.109.197.168: icmp_seq=201 ttl=128 time=8387.910 ms
64 bytes from 192.109.197.168: icmp_seq=202 ttl=128 time=7386.827 ms
64 bytes from 192.109.197.168: icmp_seq=203 ttl=128 time=6385.940 ms
64 bytes from 192.109.197.168: icmp_seq=204 ttl=128 time=5384.064 ms
64 bytes from 192.109.197.168: icmp_seq=205 ttl=128 time=4382.936 ms
64 bytes from 192.109.197.168: icmp_seq=206 ttl=128 time=3380.054 ms
64 bytes from 192.109.197.168: icmp_seq=207 ttl=128 time=2377.515 ms
64 bytes from 192.109.197.168: icmp_seq=208 ttl=128 time=1375.960 ms
64 bytes from 192.109.197.168: icmp_seq=209 ttl=128 time=373.828 ms
And then nothing. The ping times are interesting: they're at one second intervals, so basically the response to ICMP sequences 196 to 209 all happened at the same time (24 ms between all of them). And it was preceded by 17 dropped packets.
Does this happen with despise too? At first I thought that it didn't, but after a good 2 hours I had:
64 bytes from 192.109.197.170: icmp_seq=8018 ttl=128 time=0.603 ms
64 bytes from 192.109.197.170: icmp_seq=8031 ttl=128 time=18192.978 ms
64 bytes from 192.109.197.170: icmp_seq=8032 ttl=128 time=17193.275 ms
64 bytes from 192.109.197.170: icmp_seq=8033 ttl=128 time=16193.338 ms
64 bytes from 192.109.197.170: icmp_seq=8034 ttl=128 time=15186.715 ms
64 bytes from 192.109.197.170: icmp_seq=8035 ttl=128 time=14128.738 ms
64 bytes from 192.109.197.170: icmp_seq=8036 ttl=128 time=13121.802 ms
64 bytes from 192.109.197.170: icmp_seq=8037 ttl=128 time=12120.614 ms
64 bytes from 192.109.197.170: icmp_seq=8038 ttl=128 time=11118.807 ms
64 bytes from 192.109.197.170: icmp_seq=8039 ttl=128 time=10117.802 ms
64 bytes from 192.109.197.170: icmp_seq=8040 ttl=128 time=9113.875 ms
64 bytes from 192.109.197.170: icmp_seq=8041 ttl=128 time=8050.493 ms
64 bytes from 192.109.197.170: icmp_seq=8042 ttl=128 time=7047.142 ms
64 bytes from 192.109.197.170: icmp_seq=8043 ttl=128 time=6046.126 ms
64 bytes from 192.109.197.170: icmp_seq=8044 ttl=128 time=5044.124 ms
64 bytes from 192.109.197.170: icmp_seq=8045 ttl=128 time=4030.953 ms
64 bytes from 192.109.197.170: icmp_seq=8046 ttl=128 time=3029.581 ms
64 bytes from 192.109.197.170: icmp_seq=8047 ttl=128 time=2023.347 ms
64 bytes from 192.109.197.170: icmp_seq=8048 ttl=128 time=1014.190 ms
Here the response times are not quite as close, but clearly it's the same syndrome.
So: my best bet now is that there's really something wrong with Microsoft on VirtualBox, and it's probably related to the hangs that I had with the old network card, which never had any issues except with this combination. Whose fault is it? Microsoft? VirtualBox? FreeBSD (the interface with VirtualBox)? Time to try other alternatives.
| Tuesday, 25 June 2024 | Dereel | Images for 25 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Preserved lemons
|
Topic: food and drink | Link here |
We occasionally eat Moroccan preserved lemons, but they've been out of fashion for a number of years, and we're running out.
But why buy them? It can't be that difficult to make them ourselves. Spent some time reading recipes, noting that all US-based recipes want sugar, and no other ones do. And some want various spices. This one suggests freezing the lemons after preparation, and then thawing them again, to speed up the maturation process. But How to Make Moroccan Preserved Lemons looked most authentic. No sugar, of course. No spices. Just lemons and salt:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Now to wait a month or so.
|
More Microsoft on VirtualBox investigation
|
Topic: technology, opinion | Link here |
What's wrong with the networking on disaster.lemis.com and despise.lemis.com, my Microsoft 10 virtual machines running under VirtualBox. Is it the hardware? Is it FreeBSD? Is it VirtualBox? Is it Microsoft?
For the fun of it, asked Bard, which came up with this answer, referring to a forum post that didn't quite address my issues, but it gave me some ideas to follow on. This post is interesting because it refers to the same network hardware that I have. This incomplete post seems to indicate that a different network adapter might solve the problem. But if so, I have two problems. Certainly we have different symptoms now. Previously the entire system would hang, and only under load, while now it seems to be only the Microsoft guest, and it does it at random. This post, in the same thread, suggests setting “Paravirtualization Interface” to “Hyper-V”. Tried that, with the convenient result that it hung very quickly, saving me time.
What is it doing in this time? Tried a tcpdump, which showed me continuous
12:15:37.799085 ARP, Request who-has eureka.lemis.com tell despise.lemis.com, length 46
12:15:37.799087 ARP, Request who-has eureka.lemis.com tell despise.lemis.com, length 28
12:15:37.799213 ARP, Reply eureka.lemis.com is-at bc:5f:f4:c9:9b:bf (oui Unknown), length 46
For some reason there are always two ARP requests, a pair per second. So it seems that the response was not getting through.
And then there's a number of these in the /var/log/messages:
Jun 25 12:40:47 hydra kernel: vboxdrv: XXXXXXXXXXXXXXXX VMMR0.r0
Jun 25 12:40:47 hydra syslogd: sendto: Network is down
Jun 25 12:40:47 hydra kernel: vboxdrv: XXXXXXXXXXXXXXXX VBoxDDR0.r0
Jun 25 12:40:47 hydra syslogd: sendto: Network is down
Jun 25 12:40:47 hydra kernel: VMMR0InitVM: eflags=40246 fKernelFeatures=0x2 (SUPKERNELFEATURES_SMAP=1)
Jun 25 12:40:47 hydra syslogd: sendto: Network is down
Jun 25 12:40:47 hydra syslogd: last message repeated 2 times
Somehow messages with XXX in them (for FreeBSD, 3 Xs are enough) suggests debug output. And the “network is down” messages from syslogd suggest that there's something wrong. Are the XXXs related? The interspersed promiscuous mode messages suggest that it's happening while I'm looking at the network. I'll keep an eye on /var/log/messages from now on.
And then there's another issue that may or may not be related: noise in the audio output, another thing that I should investigate. It seems to be related to the virtual machines, though it may just be related to the load on the host machine.
|
fvwm3: progress
|
Topic: technology, opinion | Link here |
Somehow I still can't make much sense of the required syntax for the fvwm3 configuration file. There seems to be a disconnect between the (relatively copious) documentation and my understanding. So time for some experimentation. Made these changes (inter alia) in the configuration file:
+Colorset 21 fg black, bg yellow
-Style "xterm-r" Icon xterm.xpm,Color yellow/black,IconBox -40 300 -1 -140
+Style "xterm-r" Icon xterm.xpm,Colorset 21,IconBox -40 300 -1 -140
Success! Well, sort of. Now these windows have icons that have a yellow background and a black foreground, which proves to be the opposite of what I wanted. That's easy enough to fix, but I still need to find out how to highlight the windows themselves. Still, the first step of progress.
| Wednesday, 26 June 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Beijing “fried sauce” noodles again
|
Topic: food and drink, opinion | Link here |
Another attempt at Beijing “Fried Sauce” Noodles today. Somehow I'm still not happy with the results. Today I tried, basing on the recipe of 10 June 2024:
The result? Probably too much star anise. The noodles were much better after being fried. And for some reason, the quantity was too much. Somehow I still have issues.
|
More disaster
|
Topic: technology, opinion | Link here |
So why are my Microsoft virtual machines getting network hangs? An obvious thing to try would be to upgrade VirtualBox. That went smoothly, but it caused problems with Qt that were more difficult to resolve.
On restarting (this time disaster.lemis.com) I had to recover the net configuration again, after which things ran smoothly. How long? Set up a ping to record the time and stop when there was a failure. It was still running when I went to bed. Solved? I don't trust it.
|
Linsensuppe: quantities?
|
Topic: food and drink, opinion | Link here |
Yesterday Yvonne cooked a variation on Linsensuppe, enough to freeze multiple servings. But how big should they be? I thought 500 to 600 g, Yvonne thought 600 to 700. So I prepared 700 g portions for freezing and kept one for tonight.
Surprise: 700 g isn't enough. 800 g next time?
| Thursday, 27 June 2024 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More Chilliz mi udang
|
Topic: food and drink, opinion | Link here |
Somehow I have ended up with a second packet of Chilliz mi udang paste. I tried it last year and found it boring. But it's there, so I made it up. Yes, boring. Not bad, but not to be repeated
|
More VirtualBox fun
|
Topic: technology, opinion | Link here |
When I went to bed last night, disaster.lemis.com was still running fine. Fixed?
No. When I got up this morning, I discovered that it had stopped responding at 23:32:20, with the same symptoms as I have been having before.
What next? Look at the log files. Nothing, neither in /var/log/messages nor in the VirtualBox logs. Where do I go from here? Changing hypervisors doesn't seem to be an option: Bhyve seems to have issues with Microsoft, VMware doesn't run on FreeBSD, and what I've seen of Xen suggests that it needs to run on the bare hardware.
Change machines? That currently looks like the best option. Use one of my ThinkCentres. But that will require preparation.
|
fvwm3: success?
|
Topic: technology, opinion | Link here |
So I moved on to my other issue, fvwm3 configuration. The obvious thing was to change all of the specifications that fvwm3 complained about. That wasn't as simple as it looked. For example, it complained about this line:
HilightColor black palegreen
What does the man page say? What man page? I have fvwm3(1), but it doesn't mention anything like that. On the web I found this page, but though the content is different and more complete, it also didn't help. But somewhere that I forgot to write down told me:
HilightColor textcolor backgroundcolor
This command is obsoleted by the Style options HilightFore and HilightBack. Please use
Style * HilightFore textcolor, HilightBack backgroundcolor
instead.
Only much later did I discover that there are multiple man pages, though fvwm3(1) doesn't have a SEE ALSO section:
/usr/local/share/man/man1/fvwm3.1.gz
/usr/local/share/man/man1/fvwm3all.1.gz
/usr/local/share/man/man1/fvwm3commands.1.gz
/usr/local/share/man/man1/fvwm3menus.1.gz
/usr/local/share/man/man1/fvwm3styles.1.gz
I don't know how much it would have helped. fvwm3 still didn't want to know:
[1719458711.919614] parse_and_set_window_style: Bad style option: HilightFore black
[1719458711.919618] parse_and_set_window_style: Bad style option: HilightBack palegreen
But it seems that I don't really need that. After setting all my Colorsets, it still didn't want to do what I want. Clearly I'm missing something important, but it seems that this part is most important:
# 3: Colorsets
#
# 0 - Default
# 1 - Inactive Windows
# 2 - Active Window
# 3 - Inactive Windows Borders
# 4 - Active Windows Borders
# 5 - Menu - Inactive Item
# 6 - Menu - Active Item
# 7 - Menu - Grayed Item
# 8 - Menu - Title
Played around with that, and gradually I got to the point where some of the colours were correct. I could even live with that—the colours I use were chosen more at random than out of any logic. But I wish I understood this stuff.
Next step: the xterms files. A lot of what's there would fit better into an .Xdefaults file. How much work should I do?
| Friday, 28 June 2024 | Dereel | Images for 28 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More VirtualBox fun
|
Topic: technology, opinion | Link here |
As planned, set up one of my ThinkCentres for VirtualBox testing. I was going to call it monorchid in memory of my old test box for SMPng, but then it occurred to me that even the oldest machines now have multiple processors. This one has four, so I called it quartet.lemis.com.
Things didn't go easily. I ran into the old issue with that I had in March. But even after setting that, things weren't smooth. Did I have the kernel modules?
=== root@quartet (/dev/pts/0) /src/VirtualBox 30 -> kldstat|grep vbox=== root@quartet (/dev/pts/0) /src/VirtualBox 40 -> kldload vboxnetadp
kldload: can't load vboxnetadp: module already loaded or in kernel=== root@quartet (/dev/pts/0) /src/VirtualBox 41 -> kldload vboxdrv
kldload: can't load vboxdrv: module already loaded or in kernel
Huh? This is a GENERIC kernel, which doens't have VirtualBox modules. Another Schrödinger's cat? Do we have some mismatch somewhere? To be on the safe side, started bringing the system up to date. That's slow. Another day gone by.
|
Why quartet?
|
Topic: history, technology | Link here |
Why did I call the new machine quartet.lemis.com? That goes back decades. At Tandem our first Unix machine was an LXN, a rebadged Altos machine. Our first real Tandem Unix box was the Integrity S2. So when it came to naming the machines, the LXN was called solo.euts.tandem.com, and the S2 was called trio.euts.tandem.com.
After I left Tandem, they had a new machine, which they called quattro.euts.tandem.com—that would never have happened if I had stayed. In English, we have solo, duo, trio, quartet and more. Since my “new” machine has 4 CPUs, quartet is the obvious choice.
|
SSD strangenesses
|
Topic: technology, opinion | Link here |
The first step in installing quartet was copying a system image. That's simple: I have a number of identical 128 GB SSDs, one of which has the system for teevee. So all I need is dd.
=== root@teevee (/dev/pts/0) /home/grog 3 -> time dd if=/dev/ada0 of=/dev/ada1 bs=1m
122104+1 records in
122104+1 records out
128035676160 bytes transferred in 1944.455866 secs (65846532 bytes/sec)
65 MB/s? That's not exactly blindingly fast for an SSD. I had been watching “progress” from another xterm, which showed me:
=== grog@teevee (/dev/pts/1) ~ 3 -> iostat 1
tty ada0 ada1 pass0 cpu
tin tout KB/t tps MB/s KB/t tps MB/s KB/t tps MB/s us ni sy in id
...
0 84 1024 19 18.85 1024 19 18.85 0.0 0 0.00 0 0 0 0 100
0 80 1024 18 17.96 1024 19 18.96 0.0 0 0.00 0 0 0 0 100
0 80 1024 195 194.80 1024 195 194.80 0.0 0 0.00 0 0 0 0 100
0 80 1024 50 50.04 1024 49 49.06 0.0 0 0.00 0 0 0 0 100
0 81 1024 18 18.15 1024 18 18.15 0.0 0 0.00 0 0 0 0 100
0 77 1024 19 19.27 1024 19 19.27 0.0 0 0.00 0 0 0 0 100
0 84 1024 19 18.80 1024 19 18.80 0.0 0 0.00 0 0 0 0 100
0 77 1024 19 19.17 1024 19 19.17 0.0 0 0.00 0 0 0 0 100
0 80 1024 19 18.99 1024 19 18.99 0.0 0 0.00 0 0 0 0 100
0 81 1024 50 49.53 1024 51 50.54 0.0 0 0.00 0 0 0 0 100
0 79 1024 192 191.85 1024 191 190.86 0.0 0 0.00 0 0 0 0 100
What happened there? It reminds me of Shingled Magnetic Recording disks, but these are both SSDs. It seemed to be related either to copying or to writing: I could read at round 600 MB/s.
|
Fixing the bore pump
|
Topic: gardening, general, opinion | Link here |
Finally got my act together to remove the cobwebs that I found in the power box for the bore pump last week. Adrian from UPI had suggested using compressed air to blow away the cobwebs.
That was easier said than done. After an attempt, it looked like this:
|
|||||||||||||||||||||||
Is that any better? Tried with a brush from the kitchen and then compressed air, and things looked better:
|
|||||||||||||||||||||||
But it still wouldn't start. With a multimeter I confirmed that there was power on the first of these boxes (lower on the first image), but nothing on the second.
|
|||||||||||||||||||||||
But the cable in between looks like this:
|
|||||||||||||||||||||||
That's probably the real cause of the problem. My days of electrical wiring are over, so Adrian can come and fix it.
| Saturday, 29 June 2024 | Dereel | Images for 29 June 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Still more VirtualBox fun
|
Topic: technology, opinion | Link here |
Finally my world build was done. Rebooted quartet and ran make installworld. It failed!
install -o root -g wheel -m 444 ftime.3.gz /usr/share/man/man3/
install: ftime.3.gz: No such file or directory
How did that happen? My best bet is that the culprit was the nightly cron job that updates the source trees. Installed as best I could, and then on.
OK, update the VirtualBox packages to be on the safe side. But no, I was up to date.
=== root@quartet (/dev/pts/0) /src/VirtualBox 40 -> kldload vboxnetadp
kldload: can't load vboxnetadp: module already loaded or in kernel
We've seen that before. OK, what does ktrace tell me?
72203 kldload CALL kldload(0x36febbf8778)
72203 kldload NAMI "/boot/kernel/linker.hints"
72203 kldload NAMI "/boot/kernel/vboxnetadp"
72203 kldload NAMI "/boot/kernel/vboxnetadp.ko"
72203 kldload NAMI "/boot/modules/linker.hints"
72203 kldload NAMI "/boot/modules/vboxnetadp.ko"
72203 kldload NAMI "/boot/modules/vboxnetadp.ko"
72203 kldload NAMI "/boot/kernel/linker.hints"
72203 kldload NAMI "/boot/kernel/vboxdrv"
72203 kldload NAMI "/boot/kernel/vboxdrv.ko"
72203 kldload NAMI "/boot/modules/linker.hints"
72203 kldload NAMI "/boot/modules/vboxdrv.ko"
72203 kldload NAMI "/boot/modules/vboxdrv.ko"
72203 kldload NAMI "/boot/kernel/linker.hints"
72203 kldload NAMI "/boot/kernel/kernel"
72203 kldload RET kldload -1 errno 17 File exists
72203 kldload CALL write(0x2,0x36febbf7560,0x9)
72203 kldload GIO fd 2 wrote 9 bytes
"kldload: "
72203 kldload RET write 9
72203 kldload CALL write(0x2,0x36febbf7650,0x39)
72203 kldload GIO fd 2 wrote 57 bytes
"can't load vboxnetadp: module already loaded or in kernel"
What's it looking at there? Clearly it's finding dependencies. To be sure, I'd have to have a look at the source code. But while I was scratching my head about that, I saw this on the console:
Jun 29 10:01:52 quartet kernel: KLD vboxdrv.ko: depends on kernel - not available or version mismatch
Jun 29 10:01:52 quartet kernel: KLD vboxnetadp.ko: depends on vboxdrv - not available or version mismatch
That's a completely different message. I really need to fix that.
OK, it looks like the packages are too stupid to understand kernel dependencies. Rebuild the ports virtualbox-ose and virtualbox-ose-kmod. virtualbox-ose was an absolute pain. First it decided to build gcc 13, and then it tripped over irrelevant python dependencies, which I worked around by removing the old version and trying again. But then this one came up:
===> py311-setuptools-rust-1.9.0 depends on executable: cargo - not found
*** Error code 1
What's that? But thinking about it, do I really need virtualbox-ose? virtualbox-ose-kmod is the kernel module. Built that, which was relatively fast and painless, and finally, after something like 24 hours, I was able to load the modules.
Next, start disaster. Can't do that: “VT-x is disabled in the BIOS for all CPU modes (VERR_VMX_MSR_ALL_VMX_DISABLED)”.
OK, check the BIOS. How do I enter? I thought it was either F2 or DEL for all BIOSes, but that didn't work. Finally a web search told me: for ThinkCentres, F1 is your friend. And there, under Advanced/CPU setup, I found that I had to enable Intel (R) Virtualization Technology, which then enabled me to enable VT-d, hopefully the same thing as VT-x, and TxT.
And that worked. Finally I was able to start disaster. Got an rdesktop session running. And then disaster struck: I got the same network hang that I had had under hydra. Only this time quartet had grabbed the mouse and maybe the keyboard, so I couldn't respond. There's a solution to that, the Host key, normally bound to Control_R. But my key map doesn't have a Control_R!
Shot down the VirtualBox instance, which unstuck things, and of course updated my key map to add a Control-R key. I chose the crescent moon symbol at top right of the Sun Type 7 keyboard, which, as the old keymap told me, is normally bound to XF86PowerOff—talk about intuitive!
But somehow I'm stuck, again. The good news seems to be that it's not related to hydra. But what can I do? Try bhyve again? All the instructions relate to installing new copies of Microsoft on bhyve. Off looking again, and for the fun of it, asked Gemini, which indeed came up with instructions that don't look overly encouraging. In particular, they require additional drivers on the guest side, something that seems to contradict the idea of a virtual machine. But maybe I'll take a look. In the meantime, I have real machines that can do the job.
So: for the moment, back to my other issue, fvwm3. That shows some promise of becoming useful, even if I still don't understand it.
| Sunday, 30 June 2024 | Dereel | Images for 30 June 2024 |
| Top of page | ||
| previous day |
|
Moulding bread dough
|
Topic: food and drink, opinion | Link here |
Baked bread again today. As everything was ready for proofing, I found a little more dough in the pot. Should I mix it in again? That's a lot of work. What happens if I just push it on top of the dough?
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, the splodges are ugly, but will they stay that way? No, it seems. Here before putting into the oven and afterwards:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
So I don't need to be overly careful with the surface.
|
Donate somebody else's organs!
|
Topic: health, technology, opinion | Link here |
I've been planning to apply for an organ donor pass for years now, and today I found I can do it on the web. Only one minute! Well, 3. First I need to find my Medicare card and enter the details, and then enter my name in a form that the form likes (to match what's on the card).
And that's it! I'm a certified organ donor! All I needed was my name, address and Medicare card number. I could do that for lots of people without them knowing. In an age where people are paranoid about security, it's amazing that I can get a spare kidney or two without the donor knowing (even before his death). I didn't even get a confirmation email message.
|
Understanding Xdefaults
|
Topic: technology, opinion | Link here |
Now I have an almost functional fvwm3 configuration. How do I polish it? An obvious thing would be to move much of the specifications to the .Xdefaults file. I already have some stuff in there. For example, for xterm:
xterm*Geometry: 100x50
xterm*faceName: DejaVu Sans Mono
xterm*faceSize: 9
So I don't need to put those details in the fvwm3 config. But what about the others? Currently I have:
+ "hydra" Exec LC_ALL=en_AU.UTF-8 /usr/local/bin/uxterm -name "xterm" -s -sl 2048 -sb -ls -j -rw -display :0.3 -geometry 90x50+53+0 -e /usr/local/bin/bash &
+ "hydra" Exec LC_ALL=en_AU.UTF-8 /usr/local/bin/uxterm -name "xterm-r" -s -sl 2048 -sb -ls -j -rw -display :0.3 -geometry 90x50-53+0 -e /usr/local/bin/bash &
That defines the xterm parameters for left and right xterms. The big difference is the geometry (one 53 pixels from the left margin, one 53 pixels from the right margin). For that, the -name parameter is different. But how do I tell .Xdefaults how to set the geometry? Off looking in the application defaults (in this case, /usr/local/lib/X11/app-defaults/XTerm), and to my surprise only found one of the parameters that I have in .Xdefaults, faceSize. And also nothing that started with xterm*.
So where's the list? And do the names correspond to the names on the xterm invocation? Where are they defined? A web search just told me that .Xdefaults is an old, worn-out magic word, and that .Xresources is better, but also no longer needed.
It's been over a third of a century since I started using X. It seems that I have forgotten more than I have learnt.
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 |
