|
| Greg's diary |
| October 2020 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Thursday, 1 October 2020 | Dereel | Images for 1 October 2020 |
| Top of page | ||
| next day | ||
| last day |
|
Collectable cameras
|
Topic: photography | Link here |
I have now bought at least 3 old SLR cameras for collection purposes—4 if you count the second Asahi Pentax SV. Time to start documenting them.
First, though, a couple of questions about the Nikon equipment. Why is the diaphragm on the Nikkor 35 mm f/1.4 sticking? What can I do about it? And how easy would it be to make the Nikon F presentable? Surely there must be a Nikon collectors' group on Facebook.
Yes! In fact there are two, one much more active than the other. Signed up and was accepted almost immediately, on condition that I show a photo of my kit:
That got a lot of “likes”, and I also got some useful information about the sticking diaphragm and the covering for the F. That inspired me to take a closer look at it, something that I didn't get when I bought it. Finally found out how to remove the viewfinder, which stuck a little. Inside it looks quite good:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The scratching on the battery cover in the last photo is new: I had difficulty getting out the (spent) batteries. Before that it looked fine.
Also noticed that the damage to the leatherette is only on the front of the camera. The back and the finder are in perfect condition. Left outside in the sun, maybe on the passenger seat of a car? And one person offered me a replacement, and more generally that's what CustomCameraCovers are there for. And then there's also cameraleather, who show an ancient Exakta on their home page, but whose web site appears to be a place holder.
And the lens with the sticking diaphragm? Found this video showing partial dismantling of a Nikkor 50 mm f/1.4, the kind that I also have, and which is potentially very similar inside. I could at least remove the mount and see if the lever mechanics are part of the problem.
I really need to write up a page.
|
Little garden work, again
|
Topic: gardening | Link here |
Finally got round to putting up the garden lattice screen that has now blown down twice. In each case, it was an east wind that blew it down, so I mounted a couple more cross-members on the other side of the posts and screwed it to that. The result looks a little strange:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But it should work, and it has the advantage that the Clematis “Edo Murasaki” (which is already producing buds) will get more sun:
|
|||||||||||||||||||||||
Also planted two chili plants, a Serrano and a chile de árbol:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
| Friday, 2 October 2020 | Dereel | Images for 2 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Bloody “smart” phones, ipHone edition
|
Topic: technology, general, opinion | Link here |
Tried to call Petra Gietz on her mobile phone this morning. It diverted immediately to voice mail. OK, not a problem, she came by a few minutes later, so I told her about the diversion. Oh yes, she knew about that, but she didn't know what was wrong.
It's an iPhone 5, something I have had (and rejected) before. How hard can it be? Took a look, in the process also, at Petra's request, to disable the requirement for a pass code when powering on.
I couldn't manage it and had to enlist Daniel O'Connor's help. The real issue is the menu system (why does the icon for “Settings” look like the fan cover on a computer power supply?). The first time round I found a way to disable the “passcode” for phone access, and to tell the phone to answer when spoken to.
Neither worked. Tried again, in the process discovering that I had lost the track to the menu items Finally found more settings: “Only answer for favourites? ON”, and “Disable Passcode”. OK, allowed answering for any Tom, Dick and Harry, and disabled the “passcode”.
Success! And it only took half an hour.
Once Apple advertised with “The computer you can learn in a day”. Now, it seems, you can't even learn to use a phone. This is progress?
|
US Military: unreachable
|
Topic: technology, politics, general, opinion | Link here |
Not completely unexpected mail today:
Date: Fri, 2 Oct 2020 02:10:55 +0000 (UTC)
From: Mail Delivery System <MAILER-DAEMON@lax.lemis.com>
To: groggyhimself@lemis.com
Subject: Undelivered Mail Returned to Sender
This is the mail system at host lax.lemis.com.
I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.
<abuse@nic.mil>: connect to POSTAL.nic.mil[207.132.116.7]:25: Operation timed out
Somehow this all demonstrates a level of incompetence that I hadn't expected from the most powerful military in the world.
| Saturday, 3 October 2020 | Dereel | Images for 3 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Strange spam
|
Topic: general, technology, politics, opinion | Link here |
From time to time I see spam subject lines that make me sit back and think. Here are two:
29 N + 03-10-2020 Home Solar Installat ( 87) N + Home Solar Panel Incentives For Local Residents
30 N + 03-10-2020 AmericanPatriotCo Co ( 58) N + Honor Your Favorite President With A Trump Coin
Turn the first one around. Home Solar Panel Incentives For Non-Local Residents? What did they think when they used the word “local”? Geographically close to the spammer? People who live near their houses? The mind boggles.
And HonOr Your FavOrite president? Clearly they're jumping to multiple conclusions, but considering that Donald Trump is currently in hospital trying not to die from a disease he doesn't believe in, the timing seems more than inappropriate.
|
More Nikon F insights
|
Topic: photography, opinion | Link here |
As I've said before, when I was young I lusted after a Nikon F, the main reason I bought my tatty camera a few months back.
Since joining the Nikon Collectors group on Facebook a couple of days ago, there has been a remarkable amount of discussion about the camera and why the Photomic head doesn't fit. Today I posted some details. Here a paraphrase:
The real issue is getting the Photomic head mounted correctly on my camera.
First: this is a Nikon F, serial number serial number 7075271, fitted with a Photomic FTn (I think) viewfinder, serial number 755418. From https://www.destoutz.ch/typ_production_data_f.html it's reasonable to believe that the camera was made in the first 3 months of 1970.
Bruce Thomas posted photos of two older cameras with different viewfinder mounts. If I understand it correctly, there's a “wall” at the back on the older camera. That's not present on mine, I think. Here more detail photos:
Image title: Nikon F focus screen 2 Dimensions: 5194 x 3888, 2429 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 3 October 2020, thumbnails All images taken on Saturday, 3 October 2020, small Diary entry for Saturday, 3 October 2020 Complete exposure details
Image title: Nikon F focus screen 2 detail Dimensions: 1451 x 1227, 282 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 3 October 2020, thumbnails All images taken on Saturday, 3 October 2020, small Diary entry for Saturday, 3 October 2020 Complete exposure details
The back of the viewfinder well doesn't seem to have any such wall, but it does have a layer of foam about 1 mm thick.
Image title: Nikon F focus screen 4 Dimensions: 5191 x 3888, 3329 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Saturday, 3 October 2020, thumbnails All images taken on Saturday, 3 October 2020, small Diary entry for Saturday, 3 October 2020 Complete exposure details
What's that for? Should I remove it?
I also attached some photos of the head, in case that gives any clues:
Image title: Nikon F Photomic 7 Dimensions: 5184 x 3888, 1691 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Friday, 2 October 2020, thumbnails All images taken on Friday, 2 October 2020, small Diary entry for Friday, 2 October 2020 Complete exposure details
And how about that, I got two useful answers. One noted that a camera with this serial number shouldn't have the “Nippon Kogaku” emblem on the top plate:
|
|
|||||||||
That apparently went out of style some time earlier. So how did my camera come to have it? There's a suggestion that all is not as it seems.
But I've been there before. 54 years ago I damaged the top plate of my Pentax Spotmatic, and had it replaced. And with the top plate I got a new serial number, 1211890 instead of 1182896. At the time the camera was only a little over a year old, but by 1970 the Nikon F had been in production for over 10 years.
Also, of course, requests for more photos. Mañana.
|
Jubiläumstropfen
|
Topic: history, food and drink, opinion | Link here |
By 30 years ago I had lived most of my adult life in Germany, and I felt relatively comfortable there. There were strangenesses, of course, one of which was the “Deutsche Demokratische Republik”, the German democratic republic, a Soviet puppet state nearly 40 years old and looking to go on for ever.
And then, suddenly, it was gone. 30 years later I'm still trying to get my head around it: from the first demonstrations in Leipzig in September 1989 it took less than 10 months for the first step of reunification, the Währungs-, Wirtschafts- und Sozialunion, translated into English as Economic merger. And 3 months later, less than 13 months after the first demonstrations, and 30 years ago today, Germany was reunified. It's amazing how quickly it happened.
On a completely different historical tack, about 14 years ago Daniel O'Connor gave me a bottle of beer that he had brewed. It must have been something special, because I put it in the fridge and left it for a special occasion. And how about that, it was brewed on 3 October 2005, 15 years ago today, and exactly halfway between the reunification and today.
What would it taste like? I was prepared for it to be completely ruined.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
How about that, it still had a head. In fact, quite a good head; the last photo of the three above was taken 10 minutes after the other two. Carbonation on the low side, also not very bitter. But it seems that the hops were supposed to be on the low side (only 20 IBU), and quite possibly the low carbonation was intentional.
|
|||||||||||||||||||||
Aaah!
|
Garden in spring
|
Topic: gardening | Link here |
It was quite warm today, a top of 28.5°. The Japanese cherry is flowering, if only just:
|
|||||||||||||||||||||||
And what else? The Carpobrotus on the island in the middle of the driveway is starting to flower:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Also took a look at the Paulownia kawakamii, which doesn't look as happy:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What happened there? My guess is frost, and that it will recover. But it's barely 1 m high, and at this age it should be fully grown at about 5 m. Will it ever make it?
| Sunday, 4 October 2020 | Dereel | Images for 4 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Summer time!
|
Topic: general, photography, opinion | Link here |
Daylight Saving Time began today, time to reset clocks. 8 conventional clocks and another 6 in the cameras.
This time setting irritates me not only because I have to do it, but because it's so imprecise. None of them allow you to set the seconds, and so I can be up to 30 seconds off the correct time. The oven doesn't even allow you to say “set the time now”; it just times out on input and then sets the time. For three of the cameras, things were much the same, but for the OM-D models I can use this wonderful OI.Share “smart” phone app. Yes, it's a pain, but they've changed it. Now, before I can use it at all, I had to fill out a questionnaire about what I use my camera for, even entering my date of birth (but it guessed that, 1 January 1990, so I just needed to confirm). What on earth are they thinking?
They weren't the only ones. As on Friday, a stupid “help” app forced its annoyance on me:
|
|
|||||||||
It wouldn't go away, and appeared not to understand the expletives I directed at it.
Why did that happen? I forget how I finally got rid of it. But then, of course, I was back with the old pain of OI.Share:
|
|
|||||||||
Damn, how do you set things up? At the bottom of the screen there's a “Camera: Off” symbol, which is Olympus' way of saying “Press here to reestablish contact”. Did that, had to scan again, and got:
|
|
|||||||||
OK, that's because of the Android "there can only be one” mantra, so I have to find “network” setup and select the camera. But what do I see?
|
|
|||||||||
“Check password and try again”. Why? I had just scanned the bloody thing. And why didn't OI.Share tell me that?
Finally I connected. That's all I need to do: it sets the time on connection.
But on another attempt I got:
|
|
|||||||||
And that was the OM-D E-M1 Mark II. I still had the E-M1 Mark I and the E-M5 Mark III to handle. They're similar cameras, so things must be similar, right?
Wrong. The E-M1 Mark II has a “Wi-Fi” icon showing on the setup screen, interestingly touch sensitive even when the touch screen is turned off. The others don't. Spent about 5 minutes RTFMing the E-M5 Mark III to discover that I need to enable Wi-Fi from the Playback menu. Now isn't that obvious? The further instructions seemed to be incorrect, something I'll need to look at again. But finally got it done, in the process discovering that the camera had advanced fully 4 minutes since the last setting.
And the E-M1 Mark I? I gave up. I really need to investigate this stuff in even more detail. But why did I have so much trouble today? This happens twice a year.
|
Baking beans
|
Topic: food and drink, opinion | Link here |
Bacon and eggs for breakfast today, including baked beans. Recently I've been really baking them in the “hair dryer” Air fryer. But that's far bigger than needed; it should also work in the smaller “coffee machine”.
It didn't. The coffee machine has a top temperature of 200°, and after 14 minutes (instead of the 10 I had in the hair dryer) there was still no suggestion of browning. “Well don't to that then”.
Coincidentally that was the last of my baked beans, so I needed to make some more. The recipe is clear: bake in the oven at 120°; more makes it boil too vigorously.
But that was too cool. I went up to 140° before any kind of bubbling was evident. What am I missing?
|
More Nikon insights
|
Topic: photography, opinion | Link here |
More photos of my Nikon F and viewfinder today. It seems that the release lever for the viewfinder has been snapped off:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
There should be a black lever projecting up to the left from the stump in the middle of the first photo, but it appears to have been broken off.
Also the front plate appears to be the old style:
|
|||||||||||||||||||||||
So it seems that the two halves of the top plate are of different vintage:
|
|||||||||||||||||||||||
Sent off a post to the Facebook group, but got no reply of substance. At the very least it has been an interesting investigation.
|
More wildflowers
|
Topic: gardening, opinion | Link here |
I've seen these flowers before, but if I found out what they are, I've forgotten:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Are they some kind of orchid?
|
US military attack decoded
|
Topic: technology, opinion | Link here |
Juha Kupiainen came up with an insight on the presumed attack from the US military last week. The URL was:
http://www.lemis.com/accepted-SubZero?aHR0cDovL3d3dy5sZW1pcy5jb20vZ3JvZy9Eb2N1bWVudGF0aW9uL0xpb25zLw==;cGPM5qAnLsbMaV7KZOmgdpyLReLuEGTO7yFyD9dPYd0=
He put that through https://www.base64decode.org/ and came up with:
http://www.lemis.com/grog/Documentation/Lions/^G^F<j^BrlƕN^Giȴ^.^FL^W t^]
Now why was that used? And what's the trailing junk? It does, however, suggest that it wasn't as malicious as I might have expected.
| Monday, 5 October 2020 | Dereel | Images for 5 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
No new Nikon information
|
Topic: photography, opinion | Link here |
Over the last few days I've had a lot of information from the Nikon Collectors group on Facebook. But yesterday's post fell flat: one person claimed, without explanation, that repairing the camera wasn't worth the trouble. And that was all.
So what do I need? The camera really doesn't look good:
|
|||||||||||||||||||||||
The most obvious problem is the peeling leatherette, which looks as if it has been stuck back with some inappropriate adhesive. But that can be replaced relatively easily and cheaply. What's worse is the fact that the viewfinder isn't mounted correctly:
|
|||||||||||||||||||||||
I've already established that the release lever on the left is broken, but it still works with some pressure. The real issue is that the name plate on the camera should be completely covered by the viewfinder. Some passing remarks on the collectors group suggest that the problem is that the name plate needs modification so that the viewfinder will fit in place.
OK, what does it look like?
|
|||||||||||||||||||||||
Took it off, only later noticing that I don't seem to have been the first to do so: there's a scratch by the right-hand (seen from the viewer) screw which suggests a slipped screwdriver (aren't photos good? This was taken the day before and confirms that I didn't do it myself). Without the plate, the viewfinder fitted properly. The only issue was that it wasn't locked in place, and so it couldn't operate the shutter dial reliably.
So: new plate? Where do I get one? Modify this one? Another passing remark on the collectors group suggested that it needed filing, potentially at top left and right. Off looking for photos, but surprisingly I couldn't find anything useful. Time to start a Nikon resources page.
|
Nikon FM2 exposure metering
|
Topic: photography, opinion | Link here |
The batteries for my Nikon FM2 arrived today. Does the metering system work? Is it accurate? Tried it with the 85 mm f/1.4, comparing it with the Olympus OM-D E-M1 Mark II and the M.Zuiko Digital 45 mm f/1.8, roughly the same field of view.
Firstly, it's really clear how much more precise modern cameras are. The Nikon, like the Pentax Spotmatic, allows you to set the film speed in ⅓ EV steps (since that's the way film speeds have “always” been allocated), but the diaphragm is only settable in whole stops, possibly because of the way the meter works—the Pentax has half stop clicks. And in both cases the shutter speed is settable only in 1 EV steps. By contrast, on the Olympus all three are settable in ⅓ stop increments (though this is configurable). The Olympus is so much easier to set.
And the metering? As good as correct; it set probably about ⅓ stop more exposure, but that's in the range of lens transmissions and fields of view. Interestingly, the Nikon was set to ASA (ISO) 25 when I received it, a speed that was very slow by the time it was made. It was used almost only by Kodachrome II and Kodachrome 25. The latter was discontinued in 2002. Had the camera not been used since then?
|
Azaleas in bloom
|
Topic: gardening, opinion | Link here |
After a couple of days of warm weather, the Azaleas in the south bed are flowering:
|
|||||||||||||||||||||||
High time to spread some fertilizer, also for the Camellia japonica, which is still looking unhappy. And somehow I hadn't expected the Snowdrops to still have such foliage, four months after the first flower, and when the daffodils have nearly finished flowering. I'll have to rearrange things.
| Tuesday, 6 October 2020 | Dereel | Images for 6 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Making tortillas
|
Topic: food and drink, opinion | Link here |
Tuesdays have become our huevos rancheros day, a dish that has evolved considerably since I first started cooking them, at least nearly 12 years ago. The recipe has evolved considerably in that time, but I haven't ever really addressed the issue of making the tortillas, which initially caused me untold grief. Time to step back and rewrite the recipe.
You can find many videos online showing how to make tortillas. They're typically made by experienced Mexican housewives who can make the things blindfolded. They show how easy it is—for them. What they don't show is how things can go wrong. That's what this page is for.
Tortillas are made from masa and water. Where do you get fresh masa? I don't know any source in Australia. Instead I use second best, “masa harina”, dried masa flour. For reasons I don't understand, the makes keep changing. As of October 2020, the source I have is called “El Maizal”:
|
|||||||||||||||||||||||
That's what I'm using here. They're so well-known that they don't even have a web site.
The short version is:
Simple, isn't it? There are plenty of videos showing how to do it.
The two biggest problems are: how do I keep the tortilla together until it's in the comal? And how do I cook it correctly?
The problem with masa is that, well, it's maize. It has no gluten to hold it together. The process of making the masa helps a little, but it's still relatively crumbly. Even the experts cheat here: this video and this one both add wheat flour, and the second one even adds oil:
|
|||||||||||||||||||||||
But that's not correct, and it's not necessary.
The other problem is the cooking. It seems that if you do it right, they will inflate like a balloon, but that has never worked for me. More to the point, how do you cook them so that they stay soft and don't brown, but are cooked through? This is still an issue for me.
So: here's my detailed procedure. In this example I'm making two tortillas, because that's what I normally do.
The ratio of water to flour depends on the brand of masa harina and also the kind. Currently I take 25 g of masa per tortilla, and 1.5 times as much water for blue masa, and 1.8 times as much for yellow masa. I have no idea how this compares with the instructions on the package, which use volumetric measure: “Two cups of masa and two cups of water”. They don't differentiate between the ratios for blue and yellow masa, which doesn't work for me.
Here I'm using yellow masa, so I have 50 g of masa and 90 g of water:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Leave to stand for about 10 minutes. I don't know if this is crucial—the instructions don't mention it—but in my experience it seems to make subsequent processing easier.
Mix the dough into a ball and weigh again. Hey! 4 g are gone!
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I don't know why this is, but it happens every time with different scales. Yes, a little will probably get lost during mixing, but 4 g seems excessive.
Tare the scales and break off the weight of one tortilla, in this case 68 g. Roll into a ball and place on baking paper in a tortillador:
|
|||||||||||||||||||||||
Place another sheet of paper and press until the tortilla is the size of the top plate:
|
|||||||||||||||||||||||
Gently peel the top sheet of paper, pulling straight back parallel to the tortilla:
|
|||||||||||||||||||||||
If you pull upwards, there's a very good chance of pulling chunks out of the tortilla.
Turn over the tortilla and place it on a third sheet of paper. Peel the bottom (now top) sheet off in the same fashion. The third sheet is necessary because the first one will be moist, and the tortilla will stick to it.
I'm cheating here. Instead of a comal, I use an electric roti maker. I still can't make up my mind if that's a good idea or not. Preheat the roti maker, turn the tortilla, still with its third sheet of paper, and place it on the plate:
|
|||||||||||||||||||||||
The paper, being dry, should peel off quickly and easily.
Leave the tortilla with the roti maker open for about 30 seconds before closing it. The tortilla is initially very soft, and if you close it earlier, you can tear the tortilla:
|
|||||||||||||||||||||||
When the tortillas are cooked, put them into a tortilla holder:
|
|||||||||||||||||||||||
Close it and let them rest for at least 10 minutes. Maybe due to the way I cook them, they can be somewhat hard when they come off the roti maker, but after a while they become softer and more flexible. I then heat them in a microwave oven before serving (10 seconds at 1200 W per tortilla).
|
More wildflowers
|
Topic: gardening | Link here |
We've been living in Stones Road for over 5 years now, long enough for me to find most local wildflowers. And maybe I have, but there are some that seem new. Have I seen these before?
|
||||||||||||||||||||||||
Possibly I have, but I'm not sure.
On the other hand, I think I have seen these before, but not identified them:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
What I certainly haven't seen before are these:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
They appear to be some kind of grass. I should investigate more carefully.
| Wednesday, 7 October 2020 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another documentation day
|
Topic: food and drink, opinion | Link here |
I write a lot, arguably too much. Today I spent most of the day writing up my tortilla page, and didn't even get finished. But after all, what's life for? I have my fun with it.
|
Race condition, caught
|
Topic: technology, multimedia, opinion | Link here |
I have a script for watching TV. It runs mpv to display the file and does some bookkeeping, including updating a list of files that we have seen. That's not completely transparent: it writes a temporary file foo in the current directory, arguably something that I should change. Today we watched the final episode in a directory. Can I delete the directory? Not yet: foo is still there. Let's check:
=== grog@teevee (/dev/pts/3) /spool/Series/90-min/Nord-bei-Nordwest 9 -> ls -l
total 0
ls: ./foo: No such file or directory
-rw-r--r-- 1 grog home 0 7 Oct 21:18 foo
Somewhere between the second and first line (in that sequence), foo was removed. I've never seen that before, and it's not clear where the error message comes from.
| Thursday, 8 October 2020 | Dereel | Images for 8 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Bloody Powercor again!
|
Topic: Stones Road house, general, opinion | Link here |
Yet Another electricity interruption notice from Powercor today. That's the fifth planned grid power outage since the beginning of July! And if to add injury to insult, it's in the middle of the night (2:00 to 5:30), when there's no hope of any photovoltaic power to make up for it.
About the only light in the darkness is that, if they keep to those times, I could use up to 1.7 kW during that time without running out of power. In fact, we only use a fairly constant 900 W during the night, so we should be able to weather this one too, if I remember to charge the battery before we go to bed.
|
Nikon F issues revisited
|
Topic: photography, opinion | Link here |
When I signed up to the Nikon collectors group on Facebook, I got a remarkable amount of helpful information very quickly. But then it dried up again, and I'm left with a number of questions. The biggest is: how do I modify the name plate to mount the Photomic head?
That's a question for Google, and it, too, came up with a number of hits, notably this one by Richard de Stoutz, whose other pages have already been very useful. The page shows a bewildering number of minor differences, but in principle there are only three kinds:
Type 1 has a separate thin metal plate behind the visible plate. It's also usually hollowed out from behind. It seems that this version was very short-lived. The highest serial number that de Stoutz records was 6410742, which corresponds to August 1960 at the latest in this list.
Type 2, the one my camera has, looks otherwise similar, but isn't hollowed out:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
According to de Stoutz' page, this version was fitted to cameras with serial numbers between 6401164 and 6899999, corresponding to manufacture dates up to May 1969.
Type 3 is the version for the Photomic heads. Here an image from the de Stoutz page:
Yes, it's rounded, but the details aren't clear enough for me to set to with a file. First I need (still) more information.
| Friday, 9 October 2020 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More Nikon restoration
|
Topic: photography | Link here |
Mail from Michael James this morning, pointing to this video clip about the Photomic FTN:
That's quite useful. There are no new insights, but some illustrations and clarifications about how the name plate and the viewfinder well are structured. I need to look at the exact structure more carefully. In particular, what is the purpose of the blue padding at the back of the well on mine?
In any case, it's beginning to look as if I will be able to fit the viewfinder with appropriate modifications, so I took up an offer of replacement leatherette by Bruce Thomas. He's sending it express (from New South Wales), and has the expectation that it will arrive here by Wednesday or Thursday next week. That sounds optimistic to me, but then I'm not in a hurry. Now I just need to find out what kind of adhesive I need.
|
Robotic lawn mowers?
|
Topic: gardening, opinion | Link here |
Our lawn is overgrowing, and we can't find anybody to mow it. Now why did we sell our lawn mower? Maybe we should buy a new one.
Off looking, and came up with the surprising discovery that there are robotic mowers on the market, rather like robotic vacuum cleaners. Our experience with the latter was not very positive, but that could be because of the model. If we can get one to work in our garden, it should be easier than relying on outside help, and it might even save money. The real issue is: what if we buy one and it doesn't come up to expectations? More investigation needed.
|
Osteopathy
|
Topic: health, opinion | Link here |
SARS2 isn't the only zoonotic disease causing concern at the moment. Round here strains of avian influenza are doing the rounds (or, hopefully, not doing them). What are they? This official page suggests H7N7, H7N6 and H5N2.
How dangerous are they? Not very, people agree, but all farms have been instructed not to sell any avian products, including eggs. That includes Michelle de Groot and MJ Wisbey, suppliers of hen and duck eggs respectively.
MJ has chosen a different occupation: osteopathy. It seems that she's a qualified medical practitioner, and with Paul Smith's approval Yvonne went to see her today. I still haven't understood exactly what MJ did, but Yvonne came home noticeably taller (5 cm?). That's quite impressive, especially since Yvonne also feels much better.
| Saturday, 10 October 2020 | Dereel | Images for 10 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Preparing the Nikon F
|
Topic: photography | Link here |
So I have new skins (“leather kit”) on their way for my Nikon F. How do I attach them? How were the old ones attached? Badly, of course, but it's worth taking a more detailed look:
|
|||||||||||||||||||||||
My guess is still that the camera was left in a car in direct sunshine. The rear looks fine:
|
|||||||||||||||||||||||
And the adhesive? I removed the remains of the leather on the left side of the camera. It shows some thick yellowish adhesive in the area that was visible, but almost nothing where the skin had not peeled off:
|
|||||||||||||||||||||||
So clearly the yellowish stuff isn't what I'm looking for. There must be some online resource that can explain it to me.
| Sunday, 11 October 2020 | Dereel | Images for 11 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Catching up with life
|
Topic: general | Link here |
I have a number of indicators of how busy I am. One is the time when I complete my diary for the previous day, which can vary from mid-morning to late afternoon. Another is the number of messages in my email inbox.
Based on those indications, I'm not very busy at the moment. For the first time in years I have less than 20 messages in my inbox, and the commit message for yesterday was:
$Id: diary-oct2020.php,v 1.11 2020/10/11 01:54:00 grog Exp $
That's UTC, of course, really 12:54, but still relatively early. So I had time to do other things, and squandered it trying to keep a port alive, while Yvonne hoped in vain that I would finally tidy up the shed.
|
More ports pain
|
Topic: language, technology, opinion | Link here |
One of the things I've been meaning to do for some time is to learn about the Proto-Indo-European language. OK, after looking at the TV news, tried a document I have:
=== grog@teevee (/dev/pts/10) ~/Documentation/Language 172 -> xpdf An_outline_of_Proto_Indo_European.pdf
Bus error
Dammit, what's wrong now? It happened with every document. Update? OK, that worked, requiring a surprising number of Qt dependencies (aren't there a surprising number?). Try again.
=== grog@teevee (/dev/pts/10) ~/Documentation/Language 172 -> xpdf An_outline_of_Proto_Indo_European.pdf
Config Error: No paper information available - using defaults
Cannot mix incompatible Qt library (5.13.2) with this library (5.15.0)
Abort trap
Dammit! How do I update 40 different Qt ports? OK, let's build from source.
=== root@teevee (/dev/pts/33) /usr/ports/graphics/xpdf 19 -> Make all
===> Applying FreeBSD patches for xpdf-4.02,1 from /eureka/home/src/FreeBSD/svn/ports/graphics/xpdf/../xpdf4/files
1 out of 1 hunks failed--saving rejects to xpdf-qt/CMakeLists.txt.rej
===> FAILED Applying FreeBSD patch-xpdf-qt_CMakeLists.txt
===> Cleanly applied FreeBSD patch(es) patch-cmake-config.txt
===> FAILED to apply cleanly FreeBSD patch(es) patch-xpdf-qt_CMakeLists.txt
*** Error code 1
Dammit, doesn't anything work any more? Looking back through my diary, I've been complaining about this for 20 years. Only 20 years? Yes, that's as long as I have been keeping this diary since having a computer at home. I've been porting software since 1992, and in “Porting UNIX Software” I described the pain—with the hope that things would get better. But that was 27 years ago, and my diary is peppered with entries like this. I don't suppose I can really hope for any improvement.
| Monday, 12 October 2020 | Dereel | Images for 12 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Investigating xpdf port
|
Topic: technology, opinion | Link here |
So why did my xpdf build fail yesterday? One of the patch files didn't apply cleanly. OK, take a look.
The patch file's not big:
--- xpdf-qt/CMakeLists.txt.orig 2017-08-08 23:22:50 UTC ***
+++ xpdf-qt/CMakeLists.txt
@@ -86,6 +86,6 @@ if ((QT4_FOUND OR Qt5Widgets_FOUND)
${CMAKE_THREAD_LIBS_INIT})
set_property(TARGET xpdf PROPERTY WIN32_EXECUTABLE 1)
- install(TARGETS xpdf RUNTIME DESTINATION bin)
- install(FILES ${PROJECT_SOURCE_DIR}/doc/xpdf.1 DESTINATION man/man1)
+ install(TARGETS xpdf RUNTIME DESTINATION libexec/xpdf)
+ install(FILES ${PROJECT_SOURCE_DIR}/doc/xpdf.1 DESTINATION share/xpdf/man/man1)
endif ()
And the file to be patched? The patch shows that the section in question starts at line 86:
target_link_libraries(xpdf ${QT_LIBRARIES} ${EXTRA_QT_LIBRARIES}
${PAPER_LIBRARY}
${FREETYPE_LIBRARY} ${FREETYPE_OTHER_LIBS}
${DTYPE_LIBRARY}
${LCMS_LIBRARY}
${FONTCONFIG_LIBRARY}
${CMAKE_THREAD_LIBS_INIT})
set_property(TARGET xpdf PROPERTY WIN32_EXECUTABLE 1)
if (QT4_FOUND)
set_property(TARGET xpdf PROPERTY QT4_NO_LINK_QTMAIN ON)
else ()
set_property(TARGET xpdf PROPERTY Qt5_NO_LINK_QTMAIN ON)
endif ()
install(TARGETS xpdf RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
install(FILES ${PROJECT_SOURCE_DIR}/doc/xpdf.1 DESTINATION ${CMAKE_INSTALL_MANDIR}/man1)
That actually starts at line 85 and goes on to the end. Clearly the file has changed since the patch was generated, as the modification timestamps confirm:
=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4 106 -> TZ=UTC ls -lT ../xpdf/work/xpdf-4.02/xpdf-qt/CMakeLists.txt.orig
-rw-r--r-- 1 root wheel 3337 25 Sep 19:54:33 2019 ../xpdf/work/xpdf-4.02/xpdf-qt/CMakeLists.txt.orig
The patch says that the file was last modified at 2017-08-08 23:22:50 UTC, but the file I unpacked was modified at 25 Sep 19:54:33 2019. And clearly it's different.
But how can that happen? I've had issues with updating the source tree in the past, and they're probably still there (one of the few silver linings on the proposed FreeBSD conversion to git), but that doesn't seem to be the case here. And the distinfo file, which contains tarball checksums, should have caught any modification. It was modified on 6 October 2019, so presumably it includes the new tarball.
But how could this build? There's a package out there. Is the patch file obsolete? It looks like it, but it doesn't seem to have been removed.
Dammit, I don't want to get involved in fixing ports with distasteful things like Qt and cmake. Enter a bug report? First I need to know if svn is doing the right thing.
I found a workaround, though: install xpdf3, the previous version of xpdf. I can't see any difference in functionality. But Yet Again the Ports Collection is making life difficult for me.
|
New gardener?
|
Topic: gardening | Link here |
Paul Gallagher from Jim's Mowing (an ISP, to judge by the Web URL) along today to take a look at the garden. Yes, he can do the job, and he's not even expensive, contrary to my expectations. He'll be along a week today.
|
More Nikon modification tales
|
Topic: photography, opinion | Link here |
The Nikon Collector's group on Facebook seems to have dried up, but there must be other resources. Off to search for modify nikon f name plate and came up with this forum thread, which shows further insights, but still doesn't show the details I'm looking for. So started another thread. Hopefully something will come of that.
| Tuesday, 13 October 2020 | Dereel | Images for 13 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Life by fifths
|
Topic: gardening, history, opinion | Link here |
From my daily calendar cron job today:
80% life: <a href="/grog/diary-may2006.php#17">Wednesday, 17 May 2006</a>
May 17 Greg last flew in an aeroplane, 2006
20% of my life ago? Or, alternatively, I'm now 25% older than the last time I got off a plane. I don't miss it.
But of course, there are more, maybe time to reflect:
20% life: Sunday, 24 February 1963
40% life: Saturday, 23 July 1977
60% life: Friday, 20 December 1991
80% life: Wednesday, 17 May 2006
May 17 Greg last flew in an aeroplane, 2006
7/8 life: Wednesday, 12 October 2011
Oct 12 Dennis MacAlistair Ritchie died at the age of 70, 2011
The 40% and 60% dates roughly cover the time I was working with Tandem equipment. And just by coincidence, it's both ⅛ of my life since dmr died, and also 9 years.
|
xpdf issues: blame svn!
|
Topic: technology, opinion | Link here |
I still couldn't understand why my xpdf build failed. Off to look at the source tree in more detail. Ran
=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4/files 110 -> svn log patch-xpdf-qt_CMakeLists.txt
svn: E155010: The node '/home/src/FreeBSD/svn/ports/graphics/xpdf4/files/patch-xpdf-qt_CMakeLists.txt' was not found.
Oh. svn update failed?
=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4/files 111 -> svn up
Updating '.':
At revision 552295.
No. So why wasn't it deleted? svn is too modern to have a real man page, of course, and the suggestion in the excuse man page sounds like an insult:
Documentation for Subversion and its tools, including detailed usage
explanations of the svn, svnadmin, svnserve and svnlook programs, his-
torical background, philosophical approaches and reasonings, etc., can
be found at .
That's 450 pages! And I don't care about history or philosophy. I just want the bloody thing to work!
But why just svn up? Shouldn't there be options there? I had a recollection of a -Pd, but that's an old, worn-out magic option from CVS. svn help shows no obvious ways to influence its update behaviour.
So: what happens if I move the files subdirectory out of the way and update?
=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4 120 -> mv files filed=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4 121 -> svn up
Updating '.':
Restored 'files'
Restored 'files/xpdf-man.conf.in'
Restored 'files/patch-xpdf-qt_XpdfWidgetPrint.cc'
Restored 'files/xpdf.desktop'
Restored 'files/patch-cmake-config.txt'
At revision 552295.=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4 122 -> l file*
filed:
total 1
-rw-r--r-- 1 grog lemis 1,986 7 Oct 2019 patch-cmake-config.txt
-rw-r--r-- 1 grog lemis 520 14 Sep 2017 patch-xpdf-qt_CMakeLists.txt
-rw-r--r-- 1 grog lemis 714 26 Dec 2018 patch-xpdf-qt_XpdfWidgetPrint.cc
-rw-r--r-- 1 grog lemis 5,793 14 Sep 2017 patch-xpdf_CMakeLists.txt
-rw-r--r-- 1 grog lemis 35 14 Sep 2017 xpdf-man.conf.in
-rw-r--r-- 1 grog lemis 210 21 Apr 2018 xpdf.desktop
files:
total 1
-rw-r--r-- 1 grog lemis 1,986 13 Oct 12:52 patch-cmake-config.txt
-rw-r--r-- 1 grog lemis 714 13 Oct 12:52 patch-xpdf-qt_XpdfWidgetPrint.cc
-rw-r--r-- 1 grog lemis 35 13 Oct 12:52 xpdf-man.conf.in
-rw-r--r-- 1 grog lemis 210 13 Oct 12:52 xpdf.desktop=== grog@eureka (/dev/pts/42) /usr/ports/graphics/xpdf4 123 ->
So there was another dead file there as well. How is this supposed to work? I've had issues with consistency almost since I first started using it, though atypically I didn't grumble about them. It's also irritating that svn doesn't restore the modification timestamp. From the file sizes, it's reasonable to assume that the four remaining patches are the same, but the timestamps differ.
Roll on git! Hopefully (dare I hope?) it won't have as many consistency issues.
|
More garden plants
|
Topic: gardening | Link here |
One of the more successful plants in the “garden” is this Leucadendron_salignum which Lorraine Carranza gave us two years ago. We planted it in a pot in front of the house, and this year it's really showing its colours:
|
|||||||||||||||||||||||
It's interesting to note some of the flowers seem to have joined together:
|
|||||||||||||||||||||||
Also finally wired up our Pandorea pandorana (“Wonga wonga”), which has been growing unchecked along the ground, and is now flowering:
|
|||||||||||||||||||||||
Hopefully it will get the message and climb upwards.
And then there's this strange plant that I've been watching for a while:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What is it?
|
NBN maintenance: Becoming less obtrusive?
|
Topic: technology, opinion | Link here |
Another National Broadband Network scheduled outage notice today. There's nothing unusual about that, but this time there was a difference: during a 12 hour window from 23:00 on 5 November to 11:00 on 6 November, I may experience up to 5 one minute outages! I may, since I monitor the connectivity, but I doubt anybody else will. Hopefully this is an indication that they're becoming less obtrusive.
| Wednesday, 14 October 2020 | Dereel | Images for 14 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Scanning old photos again
|
Topic: history, photography, general | Link here |
One of the persistent attributes of this diary is the number of photos I include. That's been the case almost since I started the online diary twenty years ago.
Clearly my paper diary of the 1960s didn't include inline photos. But there are references, and I still have the negatives, so I could add the images. I even scanned the negatives over 10 years ago. But at the time I had little connection with the paper diary (which I only typed in some time later), I also didn't have the exposure record that I kept at the time, and the scanning parameters were suboptimal.
Since then I have optimized the scanning procedure, and I have found a method of entering Exif data for the negatives. So today I started again. It's slow business, and by evening I didn't have any results worth showing. In the process, it's becoming clear that I need much more documentation.
| Thursday, 15 October 2020 | Dereel | Images for 15 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Steak and kidney pie, revisited yet again
|
Topic: food and drink, photography | Link here |
Lucy Osmond, our local district nurse and feel-good person, is compiling a Dereel recipe book with recipes, well, from Dereel. She's planning to include two of our recipes: steak and kidney pie and chicken salad. And of course she had questions, some unexpected.
By chance we had planned to eat steak and kidney pie today anyway, so time for lots of photos, and also reconsidering some aspects.
One of the bases of my approach is to make individual pies per person using pre-cooked filling, which I typically freeze in portions.
In the end took over 40 photos, most of which I'll never need. And of course we reconsidered some details, in particular whether the pies should be brushed with egg before baking. This time we decided “yes”, which proved to be an issue. As they stood, the instructions were to bake for 20 to 30 minutes. But after 5 minutes, we had:
|
|||||||||||||||||||||||
That would have completely burnt them. So aluminium foil it was.
The other thing that we changed was to eat a χωριάτικη σαλάτα, which for some reason most people transcribe as “horiatiki salad”, though it should be “khoriatiki salad”:
|
|||||||||||||||||||||||
That was in place of peas and carrots, and it proved to not be enough.
|
Another mandolin
|
Topic: food and drink, opinion | Link here |
One of the more interesting kitchen tools that I have discovered in recent years is the mandolin. Or is that mandoline? It seems that cooks have chosen on the latter non-standard spelling, and of course Wikipedia has to agree. But then, why not call it a guitare?
I now have two mandolins, and they both have their uses. The newer one can cut julienne strips, or effectively chop onions. But both have the disadvantage that they can only cut certain thicknesses.
A while back ALDI had a “professional” mandolin on special, with adjustable thickness. But it was so flimsy that it was useless. Yesterday they had another one, not “professional” and thus much more robust. And it had a (one) thicker setting. For the $10 they wanted, it might have been useful.
But it was so blunt that I just couldn't cut onions At All. Back it goes. What frustration!
| Friday, 16 October 2020 | Dereel | Images for 16 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Power fail!
|
Topic: Stones Road house, general, opinion | Link here |
This morning Powercor had announced a grid power failure between 2:00 and 5:30. Clearly no hope of photovoltaic power then, but with fully charged batteries I should be able to weather the outage.
I didn't sleep well, for some reason, and round 5:40 I was up, confirming that the computers were still running, with the potential exception of distress, which seemed to have spontaneously rebooted some time yesterday, and probably unrelated to the grid maintenance.
So how long was the outage?
|
|||||||||||||||||||||||
1:02 to 1:45! Ah, that's the new Ingeteam software, which knows about daylight saving time, but is too polite to report it. The times were really 2:02 to 2:45, and the battery charge dropped to only 89%. If the outage had lasted the full advertised 3½ hours, it would presumably have dropped to round 42%, still acceptable. So it seems that my approach works.
|
Where's my failure report?
|
Topic: technology, Stones Road house, general, opinion | Link here |
I have a daily cron job that reports grid power outages. But today there was none. Why?
Off to look at my Heath Robinson power monitoring scripts. There are two main parts, apart from the MySQL database. ~/solar/suckdb fakes a web page to the inverter and creates a MySQL command with the help of a sed script:
curl -N 'http://inverter.lemis.com/system/events/sse/stream' -H 'Accept: text/event-stream' -H 'Referer: http://inverter.lemis.com/dashboard/main.html' -H 'Cookie: __utma=51177012.766199467.1486957619.1542435223.1545448630.4; __utmc=51177012; __utmz=51177012.1542145939.2.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none); adblk=adblk_no' -H 'Connection: keep-alive' -H 'Cache-Control: no-cache' 2>/dev/null | sed -l -f frobdb.sed
And ~/solar/insertdb runs suckdb and inserts the records into the database, politely ignoring errors:
while :; do suckdb | grep --line-buffered ^Insert | mysql household; done
So what went wrong? Off to take a look. The table hadn't been updated in nearly a month!
=== root@eureka (/dev/pts/11) ~ 1422 -> ls -l, /var/db/mysql/household/powerstats.*
-rw-rw---- 1 mysql mysql 9,070 24 Apr 15:20 /var/db/mysql/household/powerstats.frm
-rw-rw---- 1 mysql mysql 5,708,447,744 24 Sep 11:58 /var/db/mysql/household/powerstats.ibd
Running insertdb gave:
=== grog@eureka (/dev/pts/31) ~/solar 297 -> ./insertdb
ERROR 1054 (42S22) at line 1: Unknown column 'StatusBat' in 'field list'
ERROR 1054 (42S22) at line 1: Unknown column 'StatusBat' in 'field list'
Huh? Where did that come from?
=== grog@eureka (/dev/pts/31) ~/solar 298 -> suckdb
event: /ems/sse/stream
Insert into powerstats set Id = "127.0.0.1:502:1", Pac = 1759, SetPoint = 0, Alarms = "", Status = "On-grid", StatusCode = 3, StatusBat = 1, SOC = 100, VBat = 273, PacBat = 2, PacPV = 1757, FromPV = false, Codes = "", W1 = -137, W2 = 0, W3 = 0;
But that has always run. Why did it fail? Looking back at the last update, that was the day I upgraded the inverter firmware. It seems that in the process they added a column to their database table (if that's the way they view things). That makes sense, and I can't exactly blame Ingeteam for not telling me. A simple case of:
mysql> alter table powerstats add StatusBat int after StatusCode;
Query OK, 0 rows affected (24 min 13.40 sec)
And the insertdb script was still running, apparently since September 2019, and presumably related to issues I had at the time.
So it's no wonder that I hadn't reported any transient grid power failures in the last few weeks. I'll probably never know if there were any, unless Powercor keeps the information accessible.
|
More firefox pain
|
Topic: technology, general, opinion | Link here |
It's been over 6 months since I started a project to replace eureka with two machines, the other called dereel. It was slow going, and without real reasons I postponed things.
One of the biggest reasons is the horror that is firefox. More and more web sites are refusing to talk to my old version, or placing severe restrictions on its usage. Today I discovered that I can no longer log in to eBay, or at least not under certain circumstances.
OK, I have a more modern (and thus harder to use) firefox on dereel. Just get it to run on a eureka server. How hard can it be? So hard that I not only failed, but managed to be no longer able to use eureka's firefox on eureka.
The first issue, of course, was getting usable key bindings for text input, now that they have broken the interface that allowed add-ons to do that. I had a recollection of a file ~/.gtkrc-2.0, but that didn't work. Finally found this page, which tells me that it's now ~/.config/gtk-3.0/settings.ini, a Microsoft like name, and it requires Microsoft-like syntax:
[Settings]
gtk-key-theme-name = Emacs
OK, that works, and starting it from an xterm gave me the chance to marvel at the amazing amount of vomit it produces on its home tty. Now use urlview to get firefox to display a mail attachment.
[-- Autoview stderr of /home/local/bin/dobrowser '/home/grog/mutt-eureka-KER6pyHl' --]
Maximum number of clients reachedUnable to init server: Could not connect: Abstract UNIX domain socket addresses not supported on this system
Error: cannot open display: :0.2
Huh? Tried it from an xterm. Can't find /home/grog/mutt-eureka-KER6pyHl. Of course! It's too polite to use things like complete file names. It should be /home/grog/mutt-eureka-KER6pyHl.html.
So why doesn't it work with urlview? Added lots of debugging output to my dobrowser script, all of which showed that it was working normally (and not mutilating the file name).
While pondering this, tried accessing eBay.
|
|||||||||||||||||||||||
“Firefox has just been updated in the background. Click Restart Firefox to complete the update.”
Huh? firefox is too stupid to “update in the background”. But I couldn't do anything, so I had to restart. And of course the problem didn't go away: I had to close the tab.
Spent some time trying to work out what was going on. Sometimes I could access a site, sometimes I got this message. But clearly things weren't working. Stopped the firefox instance on dereel and went back to the old version on eureka. It still didn't work! Somehow I'm reminded of this cartoon:
| Saturday, 17 October 2020 | Dereel | Images for 17 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Analysing X problems
|
Topic: technology, opinion | Link here |
After yesterday's catastrophe with X communication, my first thought was how to recover the situation to a point where I could continue working. What happens on server 1? Things work! In principle, I could stay there, but the rest of my environment is on server 0, and switching servers takes time.
When I returned to server 0, without thinking I tried to view an email attachment. It worked! But not for long. After a while I ran into the old problem again.
So I did some more investigation. The more I look, the clearer is that it doesn't directly relate to firefox. This article clarifies that it's a general resource issue (as the message implies), and suggested using xlsclients, xrestop and xwininfo -root -children.
I know xwininfo, though not the options suggested, but what are the other two? xlsclients proved to be installed, probably part of X, but I couldn't find xrestop. The description makes it clear that it's to be understood as x-res-top, a top-like utility.
xlsclients doesn't seem to be very interesting. It shows one line per client, though there's an -l option that gives more information:
Window 0x15a00010:
Machine: lagoon.lemis.com
Name: grog (7) lagoon:grog
Icon Name: grog@lagoon:grog
Command: /usr/local/bin/xterm -name xterm-rx -bg BlanchedAlmond -s -sl 2048 -sb -ls -j -rw -display eureka:0.2 -geometry 100x65-53+0 -fn 10x20 -e /usr/local/bin/bash
Instance/Class: xterm-rx/XTerm
That's still not very interesting, and it doesn't mention the display on which it is active. What about xwininfo? It produced much more output, a total of over 420 entries, like:
0x10a00051 "Xpdf: Print": ("printDialog_popup" "Xpdf") 5x5+0+0 +0+0
0x10a00050 "Xpdf: Find": ("findDialog_popup" "Xpdf") 5x5+0+0 +0+0
0x10a00042 "Xpdf: About": ("aboutDialog_popup" "Xpdf") 450x300+0+0 +0+0
0x10a0002e (has no name): () 5x5+0+0 +0+0
0x10a0002d "Xpdf: Password": ("passwordDialog_popup" "Xpdf") 5x5+0+0 +0+0
0x10a00005 (has no name): () 1600x1400+0+0 +0+0
0x10a00004 "xpdf": ("xpdf" "Xpdf") 1600x1400+0+0 +0+0
Are these all from one process? There were several repeats of this kind of information. But those seem to be multiple windows for a single process (client?). And as I was told in an X course over 30 years ago, “windows are cheap”.
What about xrestop? The article doesn't raise much hope that it will help. What the author did write was:
From what I can gather, the problem might not be too many clients at all, but rather resources kept around in the X server for clients who have long-since detached. But it would also appear that you cannot (easily?) relate X resources back to their client.
OK, run ktrace against firefox and see what happens. The thing that caught my eye in the dump started like this:
18767 firefox GIO fd 9 read 3128 bytes
0x0000 0000 0004 c06d c59c 0002 3130 0012 4d49 542d 4d41 4749 |.....m....10..MIT-MAGI|
0x0016 432d 434f 4f4b 4945 2d31 0010 57bd a0fb b03e 1002 978e |C-COOKIE-1..W....>....|
0x002c eb76 e6ce 2dfb 0100 0010 6461 656d 6f6e 2e6c 656d 6973 |.v..-.....daemon.lemis|
0x0042 2e63 6f6d 0002 3130 0012 4d49 542d 4d41 4749 432d 434f |.com..10..MIT-MAGIC-CO|
0x0058 4f4b 4945 2d31 0010 57bd a0fb b03e 1002 978e eb76 e6ce |OKIE-1..W....>.....v..|
0x006e 2dfb 0000 0004 c06d c589 0002 3131 0012 4d49 542d 4d41 |-......m....11..MIT-MA|
0x0084 4749 432d 434f 4f4b 4945 2d31 0010 61bf 5185 4553 49d6 |GIC-COOKIE-1..a.Q.ESI.|
0x009a 8e36 7ff0 f120 ce8f 0100 0011 6672 6565 6269 652e 6c65 |.6... ......freebie.le|
0x00b0 6d69 732e 636f 6d00 0231 3100 124d 4954 2d4d 4147 4943 |mis.com..11..MIT-MAGIC|
It's full of names of old, worn-out systems. A bit of checking showed that it was the .Xauthority file. What are those names doing in there? Take a look
=== grog@eureka (/dev/pts/48) ~ 14 -> strings ~/.Xauthority | sort -u
...
battunga.lemis.com
daemon.lemis.com
deeveear.lemis.com
defake.lemis.com
euboea.lemis.com
freebie.lemis.com
monorchid.lemis.com
quartet.lemis.com
sydney.worldwide.lemis.com
tv2.lemis.com
wantadilla.lemis.com
zaphod.lemis.com
I don't think I have used any of those names since we moved to Dereel 13 years ago! And sydney was a Dell Inspiron 7500 laptop (internal project name “Sydney”, thus the machine name) that I had retired when I joined IBM in 2001. Why are these names still there?
Elsewhere I found a reference to the .Xauthority file, including this article, really intended for Citrix users, but pointing to the xauth program. And it, of course, confirmed the existence of these entries.
But what do I do about it? The articles I found had discovered more information, but they, too, didn't know how to handle the problem. Should I create a new .Xauthority file? Or is it because I have been running the servers so long that they can't handle things any more?
=== grog@eureka (/dev/pts/36) ~ 240 -> ps aux | grep Xorg
USER PID %CPU %MEM VSZ RSS TT STAT STARTED TIME COMMAND
root 4570 1.1 0.4 3521824 132468 v0 S 11Sep19 12751:19.13 /usr/local/bin/X :0 -config xorg-0.conf -logverbose 6 -listen tcp -auth /home/grog/.serverauth.4548 (Xorg)
root 39256 0.0 2.7 4254968 912068 v4 S 13Sep19 2889:08.35 /usr/local/bin/X :1 -config xorg-1.conf -listen tcp -auth /home/grog/.serverauth.39236 (Xorg)
Hmm, even for my circumstances they're been running a long time. Probably the problem would go away if I restarted server 0. But I'd like to understand the problem, not work around it. How can I tell how many clients I'm allowed? How can I change it? Can I do it on the fly?
| Sunday, 18 October 2020 | Dereel | Images for 18 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Still more X issues
|
Topic: technology, opinion | Link here |
Yesterday I established that the error message I was receiving when trying to display email attachments on firefox was genuine: the X server had too many clients. How many?
=== grog@eureka (/dev/pts/27) ~ 144 -> xlsclients |wc
45 661 5495=== grog@eureka (/dev/pts/27) ~ 146 -> xwininfo -root -children | wc -l
424
45 clients? That doesn't seem very likely. Clearly there are many more windows. Looking at the window list on my displays I found:
| Display | Number | |
| of clients | ||
| :0.0 | 19 | |
| :0.1 | 51 | |
| :0.2 | 69 | |
| :0.3 | 24 | |
| Total | 163 |
That's rather more than 45. If my guess is correct, I should have difficulty starting simpler clients, like xterms. Do I?
=== grog@teevee (/dev/pts/4) ~ 859 -> xterm &
[1] 55239=== grog@teevee (/dev/pts/4) ~ 860 -> xterm &
[2] 55254=== grog@teevee (/dev/pts/4) ~ 861 -> Maximum number of clients reachedxterm: Xt error: Can't open display: eureka:0.1
OK, I had noted that xpdf had had a surprising number of windows. And there were many open, mainly camera manuals. I can close them. How many clients do I have now?
=== grog@eureka (/dev/pts/27) ~ 148 -> xlsclients | wc
44 659 5475=== grog@eureka (/dev/pts/27) ~ 149 -> xwininfo -root -children | wc -l
219
OK, only half the windows, and it seems that xlsclients doesn't help much. But now I could start at least 10 xterms from teevee, so it had freed up a number of client slots.
So it wasn't a firefox issue after all. Started a new firefox from dereel and went about my work. My today's diary through the years has a number of links that I wanted to open in separate tabs. The first worked, but then the others crashed (“Gah!”). Repeatedly. I had to open one, handle it, close it, and then I could open another. firefox bug?
While pondering that, ran into the dreaded “max clients” error again. Looking on dereel, I discovered 25 firefox processes. Stopped them, and things worked for a while again.
So: it was a firefox problem after all. My guess is that the corpses lying around on dereel were the result of firefox crashing in some form or another. This is clearly a firefox bug, one that almost certainly will never be fixed.
So: give up. The original intention was to set up X running on two displays, and this is apparently the only way that's going to work. Reinstated the old firefox on eureka:0.2, and started a new firefox from dereel to eureka:0.1, where it won't be called on to do things that it doesn't understand. Now to fight X configuration.
In passing, it's interesting to note that my experimentation broke firemacs (the Emacs emulation) on the old firefox instances as well. How did it manage that? But there's no reason to try to reinstate it: I'm going to have to live without it in the future, along with this stupid search function that can be invoked by c-f, but requires mouse interaction to continue.
|
Coming up tomatoes
|
Topic: gardening, opinion | Link here |
A month ago I planted out the biggest of my tomato seedlings. It's earlier than recommended (danger of frost), but I thought that I could take that risk: I have more seedlings. If they survive, they'll have a head start and bear fruit earlier.
I was wrong. No, there has been no frost, but it's been cool (as it always is at this time of year), and the seedlings grew much more slowly (if at all) than the remainder left in the lounge room window. Here are the seedlings I planted last month, then and now (left to right: Beefsteak, Grosse Lisse and Tommy Toe):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's the Tommy Toe hanging off the edge of the planting box to the right. I didn't take a photo of it when I planted it a few days later.
The ones growing inside are now too big for their tubes, so I repotted them in larger pots:
|
|||||||||||||||||||||||
When should I plant them outside? Beginning of November?
In passing, the chilis (in front and much smaller in the indoor photo) are also not doing well outside. The árbol has disappeared (eaten?), and the serrano is also not doing well:
|
|||||||||||||||||||||||
| Monday, 19 October 2020 | Dereel | Images for 19 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
New gardener
|
Topic: gardening, opinion | Link here |
Paul Gallagher, the gardener, along today to catch up with a sorely neglected garden. I was a little concerned about him missing things like overgrown shrubs and cutting them down, but in fact he went through quickly and efficiently. In three hours he had mowed all the lawn and whippersnipped nearly all the edges; that would have taken Mick at least twice that time. I'm quite impressed. But why? Is it just the equipment? It certainly means that we're paying less now than we were before.
He didn't get the weeding done, of course, but he made a good attack. He'll be back in two weeks for the next round of mowing, and hopefully he'll get the weeds more under control.
|
More flowers
|
Topic: gardening, opinion | Link here |
Walking the dogs this afternoon, saw this plant:
|
||||||||||||||||||||||||
What is it? It's almost certainly a Thelymitra rubra, like this one that we saw in the same place this time last year:
|
||||||||||||||||||||||||
And back home, the pile of topsoil in the east garden is hosting a number of plants, including this:
|
||||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What's that? I've never seen one before. It appears to be a creeper, and my guess is that it arrived with the topsoil.
| Tuesday, 20 October 2020 | Dereel | Images for 20 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Keeping FreeBSD up to date
|
Topic: technology, opinion | Link here |
FreeBSD has been around for nearly 30 years now, and most of its files base on 4.4BSD. The important ones have been updated, of course, but there were also a lot of historical files, like the calendar information for the calendar program. I have it set up so that it sends me email with all calendar entries for the day.
A couple of days ago I got such an entry:
Oct 18 Boston Shoemakers form first US labor org., 1648
Huh? The USA wasn't founded until 1776. Even non-US Americans know that. Off to check, and how about that, it did happen. But clearly it wasn't the USA, and the date was 19 October.
OK, what's subversion for? Updated the entry, first committing in the wrong place. And that seems to have stirred up some form of discontent. We had already had a discussion that the files were anachronisms in this day of the World Wide Web, and that they should be maintained outside the tree. But how?
For some reason, 26 years after the files were imported from 4.4BSD, some people have what appears to be undue haste to perform a switch, which will almost certainly require moving from subversion to Git, something that just plain scares me.
| Wednesday, 21 October 2020 | Dereel | Images for 21 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Penang laksa “recipe”, updated
|
Topic: food and drink, opinion | Link here |
I've been making Penang laksa from various pastes for some time now, and they don't taste at all bad. Gradually I've come to a method that's worth recording.
Start with a packet of Richmond Assam Laksa paste:
|
|||||||||||||||||||||||
Take a good quantity of chicken broth (about 800 g), add 40 g petis udang and 25 g tamarind paste, and mix up well. Traditionally the petis udang is provided on a teaspoon stuck into the soup, presumably so that people who don't like it can just pull it out again. But it's very difficult to mix, so it makes sense to mix it first.
Then add the paste and dilute to a total of 1600 g. That's enough for four portions. Typically, like today, I use one portion and freeze the other three individually.
In the one portion, add a frozen fish ball and a pork ball. Bring to the boil and leave for a while for the balls to thaw and warm up. Take 60 to 80 g of meat or prawns. Warm up again, add the meat or fish, and just before serving add the prawns. Pour over laksa noodles and garnish with daun kesum, sliced raw onions and cucumber.
And I still don't have a recipe page for it!
I finally got round to writing up the recipe in June 2021.
|
More calendar(1) investigations
|
Topic: technology, opinion | Link here |
What do we do with calendar(1)? What do other systems do? Took a look at Linux, Mac OS, NetBSD, OpenBSD, and to my surprise all of them had calendar(1) and a similar set of calendar files, in each case in the hierarchy /usr/share/calendar/. The data files themselves are clearly closely related—every single one had the incorrect entry for the Boston Shoemakers—but the Linux one also had additional files, like calendar.kazakhstan. To my surprise, the Linux file for Australia included the line:
* $FreeBSD: head/usr.bin/calendar/calendars/calendar.australia 345251 2019-03-17 22:26:50Z grog $
And that's the current version on FreeBSD. So somebody in Linux-world is copying our data files, if not very frequently: there have been 17 further commits since then.
And the other systems? NetBSD doesn't even have calendar.australia, and the relatively few files don't include revision information. It does, of course, have a file calendar.netbsd. OpenBSD is similar: no calendar.australia, but a calendar.openbsd, with version information:
* $OpenBSD: calendar.openbsd,v 1.36 2014/08/17 09:00:44 jsg Exp $
And the files on Mac OS also have (much older) $FreeBSD$ version information.
* $FreeBSD: src/usr.bin/calendar/calendars/calendar.australia,v 1.7 2006/10/06 23:20:01 flz Exp $ */
That's interesting for a number of reasons. First, it means that these files are by no means as uninteresting as people might think. Also, Apple is (not for the first time) way out of date. And lastly, at least Linux and Mac OS are actively drawing their files from the FreeBSD repositories. There's nothing wrong with that, of course, but it raises interesting questions about any migration from the current structure.
While trawling the logs, also found this interesting pair of commits:
r358562 | cem | 2020-03-03 11:20:08 +1100 (Tue, 03 Mar 2020) | 9 lines
Add extremely useful calendar(1) application to FreeBSD
It does extremely useful things like execute sendmail and spew dubiously
accurate factoids.
From the feedback, it seems like it is an essential utility in a modern unix
and not at all a useless bikeshed. How do those Linux people live without it?
Reverts r358561.
------------------------------------------------------------------------
r358561 | cem | 2020-03-03 10:37:47 +1100 (Tue, 03 Mar 2020) | 4 lines
Fix typo in r278616
FreeBSD isn't an encyclopedia.
It would seem that in the 43 minutes between the two commits, somebody had persuaded cem of the utility of the files. The comment about Linux is amusing, given that they do use it, and the adaptation to the current Australian time zone seems just plain irritating.
So: more discussion? Not much. Stefan Esser came up with a suggested patch that, at first sight, looks appropriate, but doesn't address the fact that other systems draw their updates from our repository. I'll take a look at it tomorrow.
|
Flushing inactive memory
|
Topic: technology, opinion | Link here |
I've been grumbling relatively silently for some time that my system has so much memory in “inactive” state. Here a typical view:
last pid: 42544; load averages: 1.53, 1.28, 1.20 up 407+02:00:35 12:00:29
1677 processes:1 running, 1663 sleeping, 4 stopped, 9 zombie
CPU: 21.8% user, 0.6% nice, 3.0% system, 0.3% interrupt, 74.4% idle
Mem: 3477M Active, 23G Inact, 3197M Wired, 176M Cache, 1656M Buf, 1750M Free
Swap: 20G Total, 9799M Used, 10G Free, 47% Inuse
On the one hand I have 23 GB of inactive memory, over 70% of total memory, but I'm also using nearly 10 GB of swap. Why couldn't the memory requests be satisfied from the inactive pool? When I come into the office in the morning, the system spends several minutes swapping before it can respond adequately. I had thought that this could be related to the old version of FreeBSD that I'm running on eureka, or the length of time it has been running (currently 407 days), but I see similar issues on teevee and dereel, which haven't been running nearly as long.
Today I started a number of virtual machines to check the calendar files. They need a lot of memory, and after firing up four of them I had:
last pid: 43123; load averages: 1.98, 1.52, 1.31 up 407+02:04:53 12:04:47
1685 processes:1 running, 1671 sleeping, 4 stopped, 9 zombie
CPU: 4.8% user, 0.0% nice, 1.9% system, 0.2% interrupt, 93.0% idle
Mem: 3734M Active, 1981M Inact, 12G Wired, 11G Cache, 908M Buf, 2907M Free
Swap: 20G Total, 9795M Used, 10G Free, 47% Inuse, 76K In
Clearly the swap usage wouldn't drop, but suddenly I have 21 GB more memory. And after closing the VMs again, I had:
last pid: 43776; load averages: 1.37, 1.45, 1.31 up 407+02:06:54 12:06:48
1681 processes:2 running, 1666 sleeping, 4 stopped, 9 zombie
CPU: 4.7% user, 0.0% nice, 1.7% system, 0.2% interrupt, 93.4% idle
Mem: 3512M Active, 5508M Inact, 3621M Wired, 10G Cache, 1655M Buf, 8549M Free
Swap: 20G Total, 9794M Used, 10G Free, 47% Inuse
Maybe I should fire up a big VM first thing every morning.
|
Looking at Leonid's limp
|
Topic: animals | Link here |
Leonid is still limping, over four months after his indiscretion. Set off to the vet in Ballarat this afternoon, but only got as far as the gate of the property before it occurred to me that we'd just have to wait in the car while the vet did her thing. So I stayed home and Yvonne went by herself.
Result? She thinks it's his elbow. And I thought they were going to take X-rays. I'm reminded of a similar incident 6 years ago. Now they won't be able to look at him until next week.
| Thursday, 22 October 2020 | Dereel | Images for 22 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Garden flowers in mid-spring
|
Topic: gardening, general, opinion | Link here |
It's a month after the September equinox, and gradually spring is progressing, time for the monthly garden photos.
For the middle of spring, a number of plants are looking quite unhappy, despite my attempts. The Hibiscus rosa-sinensis “Uncle Max” isn't dead, but it's certainly not thriving:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
At the very least I need to protect it from the elements next winter.
The Hibiscus syriacus are coming ahead as expected, so it's not just the time of year:
|
|||||||||||||||||||||||
And my five-year-old rosemary bush is still barely alive:
|
|||||||||||||||||||||||
In fact, the whole bed seems to be suffering. The Syringia (Lilac) that I planted some time back started growing a while back, and once again it's withering:
|
|||||||||||||||||||||||
And the Buddleja × weyeriana that I planted last month also doesn't seem to be doing overly well. Here it was when I planted it last month:
|
|||||||||||||||||||||||
And now it looks like this:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, it's producing new growth, but it did that in the old place, while the shrub died. By contrast, the second plant, still in its pot, looks much better:
|
|||||||||||||||||||||||
The first one isn't dead yet, but it seems to be in a bed where everything dies. There are also a couple of Salvia microphylla that seem to have died, and even the Banksia integrifolia, now 4 m tall, doesn't look very happy:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
A number of the cones look strange:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
What's wrong in that bed?
Elsewhere things aren't looking overly good either. The Paulownia kawakamii, which should now be about 5 m tall, has survived the burnt buds of a month ago, and it's looking better:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But it's nothing like what it should be. And the Schinus molle that we planted last summer is looking no better than when we planted it:
|
|||||||||||||||||||||||
Somehow I'm completely discouraged.
There are a few positive signs. The Iris are now in bloom:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And the Strelitzia nicolai is gradually flowering again:
|
|||||||||||||||||||||||
And the Alyogyne huegelii is flowering effectively non-stop:
|
|||||||||||||||||||||||
That's really one of the few plants that looks good and grows well.
|
More calendar(1) wrangling
|
Topic: technology, opinion | Link here |
More discussion on the FreeBSD-arch mailing list today. Somehow we're running around in circles. It seems that unknown people want this “fixed” immediately, after 24 years. And though in principle people agreed with my argument about unifying with the other projects, there's no time to wait. Why? Who are the unknown people? Why now? And why should I have to do the work? Why not take a little time to do it right?
|
Chicken salad
|
Topic: food and drink, opinion | Link here |
One of the recipes that Lucy Osmond chose for the Dereel cookbook was a chicken salad that we made years ago. It seems that we haven't made it for 9 years, so today we caught up, at least to get a photo:
|
|||||||||||||||||||||||
It doesn't seem to be very interesting. But if she likes it...
| Friday, 23 October 2020 | Dereel | Images for 23 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Still more calendar(1) issues
|
Topic: technology, opinion | Link here |
I've had a complaint, copied to the FreeBSD core team and from a member of the team, about demands I made and that I was disrespectful and unproductive.
I was? I tried hard to avoid any such suggestion. So I answered and asked what I had done wrong and what demands I had made, and got a response saying “This is too hard. You're nit-picking and demanding”.
Once upon a time the core team took things seriously. But it really seems that there are unpleasant undertones in the current core team, something that I thought we had eliminated 20 years ago.
|
Goodbye Internode
|
Topic: technology, opinion | Link here |
I've had relations with Internode for well over 20 years, mainly because it was in Adelaide and people I knew worked there, but it wasn't until 7 June 2005 that I finally opened an account. But somehow the friendly connection with Internode wasn't reflected when I was a customer. And it didn't last long, of course: two years later I moved to Dereel and had to find alternative Internet connections.
I returned to Internode 10 years ago and stayed with them for another 3 years until I got an NBN link.
Why didn't I stay with Internode? I wasn't happy with the treatment I had had. It had deteriorated further in the previous 3 years, and after a couple of other attempts I ended up with Aussie Broadband. Somehow I found that sad.
But then today I read this article by Simon Hackett, the founder of Internode. He had long left when I signed up with Internode in 2010, but even before that the writing was on the wall. And now Simon, like me on NBN Fixed Wireless, had his own bad customer experience with Internode. As he titled it, “How to ignore a customer without even trying”. Internode doesn't offer the latest 75/10 speed for Fixed Wireless (there, he wrote it in terms that I can understand), so he suggested they do something about it. The suggestion took a roundabout way via iiNet and didn't address his concerns in the slightest. They didn't even show that they understood who he was.
What did he do? Switched to Aussie Broadband, only 7 years after me.
| Saturday, 24 October 2020 | Dereel | Images for 24 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
New DxO PhotoLab
|
Topic: photography, technology, opinion | Link here |
DxO have released a new version of PhotoLab, version 4. As a long-standing customer and user of PhotoLab, I found out, of course—from DPReview. As I've seen before, it seems that DxO didn't see any reason to inform its customers.
OK, find the thing, not made easier by their web site breakage. Finally downloaded and installed it, which involved restarting my Microsoft box, distress. And how about that, it looks pretty much like the old one. No obvious bugs fixed: navigating the preview strip still is touch-and-go (it worked in previous releases), and the cursor breakage I have seen is still there. But they have changed the side bars with the tools. Now you can (apparently) only see some tools at any one time, presumably an inappropriate attempt to make up for the size of the tools.
And most important of all, highlighting overexposed and underexposed areas is now turned on by default. And I can't turn the bloody thing off! Off to look for the manual. I've been there before, so went looking for it in the links I save at the time. But that's nearly 4 years ago! In the meantime they had discarded the links and even one of the hosts.
More searching the site. How about using their search function to look for "photolab manual". 13 hits, only one including the word photolab, and that was not relevant.
OK, ask Google. Aaah! It's not called a manual, fool, it's a “user guide”, and it's here in PDF form, as you'd expect, at least until they find time to move it elsewhere.
On the positive side, its 232 pages long and apparently quite detailed. OK, search for “highlight”. One hit:
DxO PhotoLab subtly and automatically retains highlights in RAW images while preventing them from being clipped, and even recovers certain details in areas that are already clipped. (We advise you to work in RAW whenever possible to take advantage of the greater latitude in processing and correcting your images.)
I suppose it's typical that they shout, and the functionality looks useful, but how do I turn this bloody highlighting off? Simple: stop the program and return to PhotoLab 3. They haven't got off to a good start.
|
More riding arena measurements
|
Topic: Stones Road house, general | Link here |
Call from David Rowe this afternoon, wanting to come around “in about 2 minutes” to do some more measurements on the site of the shed to cover the riding arena. He did that, but I'm left wondering why this is all so complicated.
|
Ossobuco revisited
|
Topic: food and drink, opinion | Link here |
Ossobuco for dinner tonight. Time to take some photos of the process:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
Mandolins: useful?
|
Topic: food and drink, opinion | Link here |
A big part of making ossobuco is the battuto for the soffrito. Finely chop onions, carrots and celery:
|
|||||||||||||||||||||||
What better to do that than with a mandolin (which for some reason is frequently spelt “mandoline”, apparently without change in pronunciation)?
Well, it's not that simple. Getting the onion right was difficult. The first attempt looked like this:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
It requires just the right amount of pressure to get things cut properly, and even then the end of the onion gets left over and has to be done by hand:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The carrots were more problems. Firstly, how do you slide them at the beginning? They just turn. And when I finally got them started (leaving a flat surface to stop them turning) I couldn't get them done all the way. This stupid mandolin pressure pad doesn't go below the line of the edges of the device, so it can't engage on this carrot:
|
|||||||||||||||||||||||
So at the end I had a lot left over:
|
|||||||||||||||||||||||
Unexpectedly, the celery went through with almost no problems:
|
|||||||||||||||||||||||
The only problem there was that the pressure pad doesn't work on them, so I could only slice them as far as I could hold them with my fingers.
So: was it worth it? I still don't know. My guess is yes: it probably saved time, and potentially I can find ways to improve the proportion that I can cut with the mandolin.
| Sunday, 25 October 2020 | Dereel | Images for 25 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Scanning old photos, continued
|
Topic: photography, technology, opinion | Link here |
It's been well over a week since I scanned in some photos I took 55 years ago. The real issue is to put them on a directory corresponding to the day on which they were taken, and fake Exif data for them.
That's not as difficult as it seems: I kept an exposure notebook at the time, and a couple of years later I transcribed it to an A5 ring notebook:
And I already have a PHP script to add the Exif data. How hard can it be?
Bearable, I suppose. But there were a number of surprises:
I had to enter the exposure details, of course. For that I use an extension of the Makejpeg format that I started decades ago. Here the first two entries:
I309_DxO Gaston-1 d c 1 l 9 F 11 f 4.8 s 250 i 500 a Paul_Hallett
I310_DxO Gaston-2 a Paul_Hallett
Those are the first two entries on the handwritten sheet above. They give the file name, the name for the photo, and Exif information: d specifies “use directory date”, c 1 is camera 1 (Pentax SV), l 9 is the lens (55 mm f/1.8 Super Takumar), F 11 is film 11 (Mark V), f 4.8 is the aperture (f/4.8), s 250 is the shutter speed (1/250 s), i 500 is the linear ISO speed (500) and a Paul_Hallett is the author.
Most of this information gets carried over from one image to the next, just not the author information. Since both images were exposed the same, that's all that needed to be carried over.
And here are the corresponding photos. Run the cursor over an image to see the Exif data:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
That's done now. But it took a lot of work. Hopefully it will be easier from now on.
|
DxO PhotoLab surprise
|
Topic: photography, technology, opinion | Link here |
Yesterday I grumbled that the new DxO PhotoLab 4 highlighted overexposed parts of images, and I couldn't turn it off. Now it doesn't highlight them, and I can't turn it on.
I really need to read that manual.
|
Statements returning to bite me
|
Topic: history, technology, opinion | Link here |
I'm not the only person looking back over the years. Peter Jeremy came up with this quote from IRC:
<PJ-}> 2010-10-25 19:35:20 <groggyhimself> What good is dump(8)?
<PJ-}> 2010-10-25 19:36:41 <groggyhimself> If I find a dump in 10 years' time, I can put it on the compost.
<PJ-}> 2010-10-25 19:36:47 <groggyhimself> tar will prevail.
And the funny thing is, 10 years later, I am still using dump(8). I wonder why. I think I was right 10 years ago.
| Monday, 26 October 2020 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
calendar(1) hurry
|
Topic: technology, opinion | Link here |
I was quite upset about the accusations I received from a core team member last Friday, especially as he didn't respond to my questions (“too hard”). And he went and authorized commits that I thought were overly hasty. I didn't react.
The result? Half the project up in arms, some because it didn't go far enough, others because it went too far, still others because it meant that the core system (here the remaining part of calendar(1)) had to know about /usr/local, while there were people trying to move /usr/local elsewhere. Up to the core team intervention we had exchanged 15 mail messages, involving 4 people. Now we have a total of 80 messages and more participants than I care to count. And things still haven't settled down.
And my objection? I wanted to unify the data files across the projects that use them. My concern was that this approach would effectively kill that idea. So far, it seems, I'm right. That's sad.
|
NBN reduces data rates
|
Topic: technology, general, opinion | Link here |
A while back the National Broadband Network introduced a new rate for fixed wireless network links, “best effort” up to 75 Mb/s down, 10 Mb/s up, and apparently (as I only now discover) called “Fixed Wireless Plus” to placate people who can't count. And that's all you can get from Aussie Broadband now.
But it's too much! This article describes the NBN's reaction to unfair users who actually use this bandwidth.
the usage profile of the network has changed, to the point where a small subset of ‘heavy’ users are routinely in breach of the fixed wireless fair use policy...
How dare they? But what fair use policy? If they have one, they have hidden it well, even from Google, which just comes up with this PDF, about an “Ethernet Product Module”, whatever that may be. Aussie also have a fair use policy (PDF), but it doesn't address this issue.
So: they've discovered a throughput problem and have blamed it on the users. Sorry, that's not acceptable. One of the impressions I have had in recent times is that traffic is cheap; that's the only reason I can find for Aussie having only two traffic categories, 100 GB/month or unlimited, at only marginally different prices. So if these people are using too much traffic, the NBN has miscalculated. In response, then should limit the traffic. But what do they want to do? Limit the speed! And that to lower than it used to be:
This is ridiculous for a number of reasons. First, how much work is it to restrict traffic based on type? And if they can't carry the traffic that they have sold, then they should change their traffic limits. And finally, if this is really necessary, then why not go back to the old 25/5 rates?
| Tuesday, 27 October 2020 | Dereel → Ballarat → Dereel → Ballarat → Dereel | Images for 27 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Leo to the vets again
|
Topic: animals, general, opinion | Link here |
Into town early this morning with Yvonne and Leonid to take Leo to the vets for X-rays. In principle it would have been enough for Yvonne to go, but she also had her car serviced—225,000 km already, compared to the 177,000 on my car. We bought the Commodore 7 years ago, and since then we've averaged 10,800 km per year. Not much, but the Hyundai Elantra that I bought two weeks earlier has only done 25,500 km in that time, an average of 3,600 km per year. A far cry from November 1968, where I did 10,000 km in a month.
Leo had to stay at the vets all day, and we didn't pick him up until 16:30. The price was lower than I expected, “only” $510. As I said, they weren't greedy.
In fact, I was wrong. The vet who examined him was Dr Stewart Greedy. He wasn't available when we picked Leo up, but he called later on, explaining also why they had to anaesthetize him: they wanted to move the joints and see what was wrong. Diagnosis: arthritis. It seems that that's not counterindicated by the sudden onset; apparently a trauma can cause it.
What do we do now? Stewart recommends 4Cyte, a “premium daily joint supplement used to protect and repair damaged joints, and actively maintain joint health”. Probably not a bad idea, but we have a potential alternative: MJ (now Melissa) Wisbey, the osteopath, also does dogs, and Yvonne is very happy with the results she has had on her person. So we'll probably take him along there soon.
|
Wine: trouble
|
Topic: food and drink, opinion | Link here |
A month or so ago I bought some Sauvignon Blanc wine from McLaren Vale Cellars. A rather silly name, “Troublemaker.” (note the full stop at the end). But it didn't taste bad, for the price ($120 for a case of 12 bottles). So when they came up with an offer of Troublemaker. for $99 per case, I jumped at it and bought 5 cases.
They arrived at the post office today, and we picked them up on the first time back from Ballarat. And then, while unpacking the boot of the car, I read: “Chardonnay”! I had ordered the wrong wine!
Catastrophe! $500 worth of wine that I didn't want. What do I do? First try a bottle (after cooling, of course). And in fact it didn't taste at all bad. So the positive side is that I have saved $105.
|
Interesting garden bushes
|
Topic: gardening | Link here |
While picking up Leonid from the vet's, looked across the road:
|
||||||||||||||||||||||||
What are they? On closer examination, it seems that the left-handed one is an Azalea or a Rhododendron.
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
And the other?
|
||||||||||||||||||||||||
I hadn't expected anything that big.
| Wednesday, 28 October 2020 | Dereel | Images for 28 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Sun orchids without sun
|
Topic: gardening | Link here |
This month has been cooler than normal for October. Since we've moved here, the average October temperature has been between 14.22° and 14.33°, but this year it has only been 13.05°, The maximum temperatures have been similar: 29.7° this year, between 29.8° and 34.8° in previous years. And today was neither overly warm nor overly sunny.
All the more surprise to find this when walking the dogs:
|
||||||||||||||||||||||||
That's a Thelymitra pauciflora, “sun orchid”, that only flowers in the sunshine. I had thought that it was a month too early, but on checking I find that I found some this time two years ago.
I should really go to more trouble to get really excellent photos of them.
Apart from that, it's Iris time:
|
|||||||||||||||||||||||
|
Rain!
|
Topic: general, opinion | Link here |
A sudden bout of heavy rain this afternoon, though surprisingly the rainfall radar showed nothing. Out later to take a look:
|
|||||||||||||||||||||||
Damn, water has entered into the entrance area. More problems with the pump? No, it was working fine, only the water took too long to drain into the sump. In about 20 minutes we had round 6.5 mm.
|
DxO PhotoLab 4: understood
|
Topic: photography, technology, opinion | Link here |
I still haven't got round to reading the DxO PhotoLab 4 manual, but I've been using it from time to time despite my issues.
What were they? The big one was highlighting underexposed and overexposed parts of the image. It used to be at top right, but it was gone. And then I found it, purely by chance, at top left. It's part of the histogram “palette”, and for some reason they moved that to the other side of the screen.
And the “only see some tools”? A feature, not a bug. The default display shows them all, but you can reduce them to specific categories, maybe not a bad idea. On the positive side, it has a history display that shows what corrections have been applied and, I think, allows returning to earlier settings. That's definitely an advantage, though they should probably tune it not to show an entry for every movement of the exposure bar, for example.
| Thursday, 29 October 2020 | Dereel | Images for 29 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another PV recalibration
|
Topic: general, Stones Road house | Link here |
Another PV recalibration this morning, from 1:50 to 8:21. Due to the time, it didn't have much effect on the power generation:
|
|||||||||||||||||||||||
The most interesting thing about it was the when. In the last 6 months the recalibrations have been regularly once a month, but the last recalibration was on 23 September, 36 days ago.
|
Nikon lens adapter
|
Topic: photography, opinion | Link here |
The adapter for Nikon lenses to Micro Four Thirds system bodies that I ordered last month arrived today after hardly more than 6 weeks. Off to Napoleons to pick it up.
People had warned me adapter about buying cheap adapters. The best are apparently the Metabones adapters, costing round $200. I had ignored them and bought the cheapest I could find, $12.93. Or was that $12.92? Or $11.95? eBay can't make up its mind. One way or another, it's only about 6% of the price of a Good Adapter.
And how does it work? Fine, no problems. Spent a bit of time playing around with my favourite Nikon lens, the Nikkor 85 mm f/1.4. It looks quite impressive on the camera, here contrasted with the M.Zuiko Digital ED 75 mm f/1.8 :
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The real advantage of this lens is the wide aperture, not in f-related form, but the absolute aperture, which is what influences the depth of field. My Voigtländer Nokton 25 mm f/0.95 has an aperture of 26.3 mm, the 75 mm Zuiko has 41.7 mm, and the Nikkor has 60.7 mm. How does that work in practice?
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
The rose at the right in the background makes it clearer:
|
|
||||||||||
|
|
||||||||||
|
|
||||||||||
Playing around again, I came up with this comparison between the 75 mm m.Zuiko and the Nikkor:
|
|||||||||||||||||||||||
|
||||||||||||||||||||||||
Not much in it, right? Well, not until I look at it more closely:
|
|
|||||||||
|
|
||||||||||
Now that's a big difference.
I had already noted—almost exactly 4 years ago, in fact—that the quality of the Pentax Super-Takumar 50 mm f/1.4 can't measure up to modern lenses. Is this just the 85 mm Nikkor, or are they all like that? Unlike the Super-Takumar, the Nikkor didn't improve when stopped down. It looks as if I'm going to take a whole set of comparison photos.
|
More wildflowers
|
Topic: gardening, opinion | Link here |
On the way home from Napoleons, went down Misery Creek Road for the first time in over a year. Nothing very new there; I'll have to compare to identify these flowers:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
The first is clearly a Diuris, but what species?
|
Watsonia in flower pot
|
Topic: gardening, general, opinion | Link here |
Watsonia are weeds that grow round here, and I have permission to remove them from the nature strip facing Stones Road. But some of them are pretty:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Yvonne decided that they were just what we needed for a flower pot:
|
|||||||||||||||||||||||
|
Making friends with DxO PhotoLab 4
|
Topic: technology, photography, opinion | Link here |
I seem to have resolved the issues that I have with DxO PhotoLab 4, so I bought the upgrade. Processing today's photos was strange: one of the Watsonias seemed to have developed a yellow throat:
|
|
|||||||||
Did it do that before? Compared with version 3 on the same image and with the same settings:
|
|
|||||||||
|
|
|
|||||||||
No! That's strange, particularly since the results of processing don't seem to be any different:
|
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Run the cursor over an image to compare it with its neighbour.
The screen shots above were accidentally run through Ashampoo photo optimizer, which changed the appearance but didn't influence the yellow colouring. But the effect wasn't repeatable.
| Friday, 30 October 2020 | Dereel | Images for 30 October 2020 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Comparing lens sharpness
|
Topic: photography, opinion | Link here |
After my surprising results with the Nikkor 85 mm f/1.4 yesterday, I ran a not-so-quick but dirty comparison of a total of 9 lenses. From longest to shortest (the way I shot),
I took each photo at three different apertures: wide open (between f/0.95 and f/2.8, depending on the lens), f/5.6 and f/16. I didn't apply any lens corrections to the 25 mm Summilux and the 75 mm m.Zuiko (the others didn't have any correction information). Here's the subject (this one taken with the m.Zuiko 75 mm at f/5.6):
|
|||||||||||||||||||||||
Sadly, the lighting kept changing, which didn't help the comparisons.
The results were interesting, though I still haven't finished analysing them. I took a series of crops from the bottom right-hand corner of the gate, round the gate opener actuator. I had expected that the 85 mm Nikkor wouldn't do very well at any aperture, but in fact, there was a significant improvement at f/5.6 (run the cursor over an image to compare it with its neighbour on all of the images below):
|
|
|||||||||
|
|
|||||||||
|
|
|||||||||
But unlike my experience 4 years ago, the Summilux didn't do that well at full aperture either:
|
|
|||||||||
|
|
|||||||||
|
|
|||||||||
And the Nokton? In these images, it's not so obviously worse than the Summilux (first image)
|
|
|
|||||||||
|
|
|||||||||
Here are the three images for the Nokton:
|
|
|
|||||||||
|
|
|||||||||
|
|
|||||||||
It's not really interesting to look at the other lenses. Clearly my subject wasn't good enough, and for real comparisons I need something more controllable.
The real winner was the m.Zuiko 75 mm, which seemed to do just as well at any aperture:
|
|
|||||||||
|
|
|||||||||
|
|
|||||||||
|
Leonid meets Melissa
|
Topic: health, animals, opinion | Link here |
As planned, off with Leonid to meet Melissa Wisbey, the osteopath, who took a good look at him, not only the arthritic elbow.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
In fact, it almost seems as if she was looking at everything else. She was particularly interested in the fact that he had had a claw removed 2½ years ago. But that was literally at the other end of his body. Exactly: we had noticed that he still chews on it, which we had attributed to some kind of habit. But it seems that it might still be sensitive, after all this time, and that to keep the load off his paw, he had put more pressure on the other front leg.
She did some magic that wasn't immediately apparent to me, but it must have been to Leo: at the beginning, understandably, he wasn't too happy about having to lie down and be prodded at, but at the end, after he had got up, he went to her and lay down in front of her. And when we got home, we went walking with him, and he walked without any irregularities for the first time in a long time.
He'll be going back in a couple of weeks, but so far, as with Yvonne, things are looking very positive.
|
More Google Translate masterpieces
|
Topic: language, opinion | Link here |
Seen online recently. Google translate translates this German sentence:
„Was ist der Unterschied zwischen warum wieso weshalb weswegen darum deshalb und deswegen“
as
What is the difference between why why why why why why and why
The original poster took it as an example of the difference in precision between German and English, but that's not the case at all. It's an example of the extreme limits of Google Translate. You just need to leave out the „und“ (recognizably “and”) and the “translation” changes to “What is the difference between why why why why why therefore why”. And of course the question should end in a ?. But if you do that, the translation drops one of the “why”s. And my guess is that it will be patched, as so many Google Translate translations have been patched before. The real issue is that it tries to translate without any understanding of syntax.
| Saturday, 31 October 2020 | Dereel | Images for 31 October 2020 |
| Top of page | ||
| previous day |
|
Another grid power outage
|
Topic: general, Stones Road house | Link here |
Another short grid power failure this morning at 7:16:56.
|
More garden flowers
|
Topic: gardening | Link here |
Our Strelitzia nicolai is flowering again:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
And the Clematis “General Sikorski” is also flowering.
|
|||||||||||||||||||||||
It would look normal if it weren't already 2 years old.
|
Reduced depth of field
|
Topic: photography, gardening, opinion | Link here |
One of the things that took me a very long time to understand was the relationship between focal length, aperture and depth of field. Yes, everybody knows that longer focal length lenses have less depth of field than ones with shorter focal length, and wider apertures also bring less depth of field with them. So my Nikkor 85 mm f/1.4 should have particularly shallow depth of field, just what I need for shots of individual plants in the garden, like this cherry tree:
|
|||||||||||||||||||||||
That was taken at a distance of about 4 m with the Voigtländer Nokton 25 mm f/0.95 25 mm at f/4, the maximum aperture of my M.Zuiko Digital ED 12-100 mm f/4.0 IS PRO lens. According to my depth of field program, the depth of field should have been 1.71 m (3.32 to 5.03 m). At f/0.95 the depth of field drops to 0.39 m (3.814 to 4.205 m):
|
|||||||||||||||||||||||
But we can do better. The Nikkor 85 mm at f/1.4 has only 0.05 m depth of field. Problem: at that distance, you can only get a fraction of the tree in the image:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Yes, the background is much more out of focus than with the Voigtländer (crop in the second image above), but to get the whole tree in the field, I would need to go back to 4 / 25 × 85 m, or 13.6 m. And at that distance, the depth of field is 0.574 m, exactly the same as the 25 mm at f/1.4.
And that's the relationship that it took me a while to understand: for a given object, the depth of field is independent of the focal length! As Marc Levoy explained in one of his lectures, it's basic triangulation.
The only way to get less depth of field is to effectively increase the sensor size by stitching. And that's what I tried to day, with great lack of success. I should have realized that I would need 16 images, and a bush like this is not an easy subject. After 9 images I gave up, and my stitching attempts showed that I had lost: I ended up with 4 different groups that Hugin couldn't resolve. Even if it had been able to, the motion of the leaves would have made it useless.
|
Skewers again
|
Topic: food and drink, language, opinion | Link here |
Dammit, what word should I used for meat (or other food) on a spit? Calling them spits is correct but both ambiguous and somewhat distasteful. People here call them “kebabs”, though a real kebab doesn't have to be on a skewer. “Skewer” is the most logical, but not very well known. In the past, it seems, I have bowed to usage and called them kebabs. Should I change to “skewers”?
Anyway, today we had skewered meat (there, another possibility) for dinner. Yvonne had acquiesced to my desire for lamb skewers—for me. For herself she had bought chicken. I grilled them in the “hair dryer” “air fryer” for the 18 minutes in my list, and the chicken came out fine—and the lamb overdone! That's going to be fun to fix. I see that I already have a comment “maybe not the way to go” against beef kebabs. I should have read that first.
|
Cropping multiple images uniformly
|
Topic: technology, photography, opinion | Link here |
Cropping yesterday's images was not easy. The photos were only approximately in the same place, something like this:
|
|||||||||||||||||||||||
What I wanted was 600x450 pixel crops of the bottom right-hand corner of the gate (run the cursor over an image to compare it with its neighbour):
|
|
|
|||||||||
|
|
|
|||||||||
There were 26 images. How should I do it?
GIMP can do that. But it's such a pain to use. If you invoke it with 26 file names, you get 26 windows (and then some). You can set a 600x450 crop rectangle fine, but you still have to navigate the dozens of windows.
xv can do this too, serially in the same window. But you can't set the crop rectangle size and have it stay.
Finally (for me) ImageMagick's convert program can do all sorts of things. But how to you tell it where to set the frame?
In the end, I made somewhat larger crops with xv, positioning the bottom right of the rectangle on the mounting bracket for the gate opener, and then cropped with convert. And once again I ran foul of ImageMagick's arcane syntax. In principle something like this should do it:
for i in *orig; do
echo convert -crop 600x450-0-0 $i ${i/-orig/}
echo exifcopy $i ${i/-orig/}
done | sh
But it didn't. I got the crop from top left (in X geometry syntax, 600x450+0+0). It seems that the -0-0 in the geometry string tells convert to offset that much from top left. What I needed was gravity, not with left, right, up and down, but north, south, east and west:
for i in *orig; do
echo convert -crop 600x450-0-0 -gravity southeast $i ${i/-orig/}
echo exifcopy $i ${i/-orig/}
done | sh
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 |
