|
| Greg's diary |
| April 2008 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Tuesday, 1 April 2008 | Dereel | Images for 1 April 2008 |
| Top of page | ||
| next day | ||
| last day |
The “Beautiful Architecture” book is due for final submissions today (well, yesterday in the USA), though I see that nearly half the authors haven't checked anything in yet. It's interesting to have access to the repository for this kind of book. Spent most of the day going through the review feedback from the editors.
Darah's Greasy Heel is still not cured, despite further treatment. It's a lot better, as I can see from a page I've been keeping. Here are photos from 21 March and today:
|
|||||||||||||||||||||||
But it's still there, and it should be gone. What a pain this Greasy Heel is!
| Wednesday, 2 April 2008 | Dereel | Images for 2 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Another day with little to show for it. We're still worried about Darah, though today's photo of her hoof proved to be misleading: it seemed to show new infection below the row of scabs:
|
|||||||||||||||||||||||
Fortunately, that proved to be just dirt blown in by the wind. The weather was really windy today, and we spent most of the day indoors as a result. Finally caught up on reading magazines. Why is it, now that I'm retired, I'm still weeks behind reading magazines? How did I ever manage when I was working?
| Thursday, 3 April 2008 | Dereel | Images for 3 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
The weather is getting cooler, and it's becoming increasingly clear that the temperature control of our Fujitsu air conditioning is not just bad, it's completely broken. It only has a scale between 16° and 30° (18° to 30° when cooling, barely enough if it were correct), but it can be off the correct mark by up to 10°. In the evening I had to set the temperature to 30°, and even then it cut out when the room temperature hit 20° to 21°. Time to get Phil in to take a look. The problem's pretty clear: it's measuring the temperature in the wrong place, somewhere inside the indoor unit, instead of in the room.
Spent some time tidying up, things that I've left lying for months. That progressed quite well until Yvonne returned from shopping, bringing a new digital camera for herself, a Supra DC 5900 on special at ALDI for $89. ALDI labelled it “Traveller”, but of course they didn't re-flash the firmware, so the EXIF data shows the real name.
To my surprise (because she's not really into photography), she liked it. For $89 you can't expect too much, but it has a “5 MP” resolution (really 4.9 million), 3-fold zoom and a 1 GB memory card. It's not significantly worse in specs that the Nikon “Coolpix” L1 I bought in Amalfi 2 years ago..
But how to process the photos? My own processing requires a lot of typing and moving things around. Did some discussion and found, of course, that there are two ways to handle the issue:
The GUI way. I've already dabbled with digikam, which wanted me to do things its way, so did a bit of searching and found lphoto, which I started installing. It's a Python application for KDE, so it took forever to install, and still wasn't finished by the evening.
The UNIX way, done with scripts. First, get devd to automatically mount the camera, then write a little script that goes into the DCIM subdirectory and copies the files to Yvonne's directory hierarchy. That worked quite nicely, but in the process discovered that it would be really nice to get ls(1) to output file modification timestamps in user-defined formats (with the help of strftime). Did a bit of playing around with that, and discovered that it's trivially simple, so I can now do things like this:
=== grog@lagoon (/dev/ttyp5) ~/Photos/20080403 7 -> ls -lD '%Y%m%d'total 41 -rwxrwxrwx 1 grog 502 1477075 20080101 BILD0001.JPG -rwxrwxrwx 1 grog 502 1626257 20080101 BILD0002.JPG=== grog@lagoon (/dev/ttyp5) ~/Photos/20080403 8 -> ls -lD '%e-%m-%Y'total 41 -rwxrwxrwx 1 grog 502 1477075 1-01-2008 BILD0001.JPG -rwxrwxrwx 1 grog 502 1626257 1-01-2008 BILD0002.JPG
That makes it much easier to create the directories, but it wasn't finished by the evening either.
| Friday, 4 April 2008 | Dereel | Images for 4 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Somehow spent most of the day working on getting Yvonne's camera stuff working. I think it's worthwhile, and once again I'm reminded of the difference between the traditional UNIX approach and today's “modern” approach.
50 years ago, the word was “contemporary”. At least that one was honest. How much of the style of the 1950s has survived?
What I'm doing now:
attach 5 {
device-name "umass0";
action "sleep 10 && mount /camera && xterm -display :0 -e /usr/local/bin/getphotos";
};
It's annoying that I have to look for umass0 and not da0, but it should do. The console messages I see are:
Apr 4 15:37:35 lagoon kernel: umass0: Supra DC 5900, rev 1.10/0.01, addr 3 Apr 4 15:37:36 lagoon kernel: da0 at umass-sim0 bus 0 target 0 lun 0 Apr 4 15:37:36 lagoon kernel: da0: <Supra DC 5900 1.00> Removable Direct Access SCSI-2 device Apr 4 15:37:36 lagoon kernel: da0: 1.000MB/s transfers Apr 4 15:37:36 lagoon kernel: da0: 982MB (2012160 512 byte sectors: 64H 32S/T 982C)
Looking at that, I suppose I could refine the entry to only perform these actions when the Supra DC 5900 appears, but at the moment that's all we have anyway.
The sleep is needed because of a race condition. As the console messages show, there's a delay between discovering umass0 and creating da0. If I try to mount da0 immediately, it fails. The 10 second delay could probably be reduced to 1 or 2 seconds.
getphotos creates destination directories based on the date (in EXIF for JPEG photos, the modification time stamp for AVI files) and copies any files that don't already exist. It then offers to delete the files on camera, and umounts the camera. It is very specific to the environment, down to the directory names both on disk and on camera; I should probably improve on that.
This means that Yvonne really needs to do nothing but connect the camera to the computer and turn it on to have the photos copied. To umount, unfortunately, she still needs to enter a single character to tell the script when she's finished.
So why is this better than the GUI crowd? For one thing, it's personalized. That really requires far fewer actions. On the other side, it's not as flexible. Many people seem to view flash cards on cameras as permanent storage. My script doesn't cater for that—rightly so, I think. Who keeps his photos in his camera and also makes backups? Somehow the idea of flash as permanent storage scares me.
Still, this just gets the photos onto the computer. She still has to process them. Currently we're using xv, with a set of rules I set up years ago. Some people would use the GIMP for this kind of work. That's fine for one or two photos, but it's so slow!. For my weekly house photos even the xv method is too painful. I wonder how much of the modern photo software would handle things with as little user intervention as my scripts with ImageMagick. Somehow people are still prepared to go to far more repetitive effort than I am.
| Saturday, 5 April 2008 | Dereel | Images for 5 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Another day with not much to show for itself. Where does the time go? Did find some time to play around with cooking: Yvonne had brought back some thinly sliced “sizzling steak”, so tried to make some fajitas, not overly successfully. The dish tasted fine, but the meat didn't taste the way I recall it. I suspect the meat is wrong; it's more suited to bulgogi. It's interesting, though, considering how popular fajitas are in the US, that only one of my many Mexican and US cookbooks actually included a recipe. The Wikipedia article bears this out.
| Sunday, 6 April 2008 | Dereel | Images for 6 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
David Yeardley came over today with his Ditch Witch, a trench digger, to dig the trenches for the garden irrigation system. That went well, despite a number of strange things we found in the ground, including pipes, mounting brackets, bolts and a horse shoe. It's amazing that this land has only been settled for about 150 years, and already there's so much human débris in the ground.
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
David also brought a post package that Chris had picked up from the Post Office on Friday: a remote control for my Olympus, and an adaptor for using M-42 lenses on it. That made more sense than it sounds, since I have five M-42 lenses and a lot of close-up gear. With the longest lens, a 300 mm f/5.5 Hanimex, and 2x and 3x teleconverters, I have a focal length of 1800mm, 72x more linear magnification than a “standard” focal length of 25mm—and a maximum aperture of f/33:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The first photo above was taken at f/96, not the sort of thing you could do hand-held; in fact, it's difficult enough on a tripod. I discovered a number of things:
The tripod itself just isn't steady enough, not helped by the fact that the camera clamp screw had worked loose. Even after steadying it, though, I found an improvement by using the “anti-shock” feature, which lifts the mirror and then waits a couple of seconds before taking the photo. The remote control helps, of course.
The small apertures mean slow shutter speeds—1/6 second in the one above. Not what you'd use for catching birds on the wing; even the wind in the grass made it unsharp.
Metering proves to be more difficult than I thought. One of the photos I took was with the 50mm f/1.4 standard lens from my Spotmatic, considered to be quite a good lens. It was badly overexposed, and I had to compensate by a whole EV:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Even then, the image was surprisingly yellow; how did that happen? Maybe the cement in the lens has yellowed with age: the photo below shows a similar yellow tinge.
I recently read an explanation for this phenomenon by Andrzej Wrotniak: the mirror of DSLRs is much smaller than the photo area, so light that comes in to the side won't be measured. This doesn't happen with modern digital lenses, but there's a big difference in the size of the rear element of the lenses:
|
|||||||||||||||||||||||
Andrzej only experienced this phenomenon (with the Olympus Zuiko 50 mm f/1.8, which by coincidence I used to take the photo above) when it was fully open; the Pentax Super Takumar shows it at all apertures. Presumably an artifact of the lens design.
| Monday, 7 April 2008 | Dereel | Images for 7 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Spent most of today with the pleasant task of designing an irrigation system for the garden. That needs to be done quickly so I can fill in the trenches before it starts raining again, which the Bureau of Meteorology has now decided will be Wednesday evening.
Philmac obligingly provide documentation, hidden somewhere in their state-of-the-art web site (“Philmac is proud to present our new website, which has been upgraded to help you find the information you need as easily as possible”). The truth is something like this, which conforms fully with the modern “don't show them anything more than absolutely necessary” principle, and to all excess uses up 100% CPU time on my 2 GHz machine:
|
||||||||||||||||||||||
In any case, they have a planning guide for setting up irrigation, apparently scanned in in low resolution, but their web site has no hits on “planning”—nomen est omen. There are several different kinds of sprinklers, and it looks to me as if the “micro sprays” are the most suited. The spec sheets give important details like flow rate per hour (including US gallons per hour, but not Australian gallons) and sprinkler radius (for the 360° units) or radius (for the others).
What they don't say is the amount of water that falls on the ground in a given time (that's lph / (M_PI * diameter * diameter / 4) * 360 / degrees). In the end wrote a little program for it, and discovered that, if I believe their data—which in at least one case appears to have suffered a transcription error—the 360° sprays deliver between 1.7 mm and 4.1 mm water per hour, the 180° sprays between 6.2 mm and 8.6 mm, and the 90° sprays between 9.7 mm and 14.2 mm per hour, all at a pressure of 100 kPa. How can you build a system with that? My guess is that the flow rates are for the 360° sprays only, and that they reduce accordingly for 180° and 90°. But why isn't this information readily available?
The other issue is the price. What do these things cost? If I install micro sprays, I'll need about 100 of them. Nobody, not even the retailers, can give me a price list, though at Midland Irrigation they wrote down a couple of prices for me on the back of a business card, including these particular sprays—they're well under $1 each, so that's acceptable. But they need pipes and mounting equipment. And if I use larger sprays, I may get by with far fewer. How much will that cost? Who knows? What unnecessary frustration!
| Tuesday, 8 April 2008 | Dereel | Images for 8 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Darah still has Greasy Heel. Yvonne took her to the vet again. Kurt is on honeymoon, and I'm not convinced that his colleagues have the same level of expertise. Looks like it's just treat, treat, treat.
Ian Donaldson sent me a message about the yellow tinge in my 50mm f/1.4 Super Takumar:
Yellowing of Takumar lenses is a common problem and is caused by radioactive elements. See here: http://photo.net/bboard/q-and-a-fetch-msg?msg_id=008UdG Note that you can clear the lens quite easily with exposure to UV, sunlight will do.
I suppose that's something worth trying. The alternative is simpler, of course: on the few occasions where I use that lens, postprocess the result to remove the cast.
Using old lenses has one very significant disadvantage: modern digital cameras aren't designed to handle them, the viewfinders are much smaller than the SLRs of 40 years ago, and it's very difficult to focus them correctly. Even with pretty much ideal conditions, this is the best I could do with a 600 mm telephoto:
|
|||||||||||||||||||||||
The full-sized image shows distinct lack of focus, even though I stopped down to f/32. So I need an auto-focus lens.
I'm still trying to decide where to buy a new lens. I've already had a disastrous experience with Cameta, who refused to communicate, went back on their promises, returned my payment and then gave me negative feedback on eBay because I described the transaction as it was. That ended up with a mutual retraction and a very bad taste in my mouth.
Now I've looked at 47th Street Photo, who now say they can accept overseas payments. In the auction I read:
International Bidders - Sorry, but we currently CANNOT accept credit cards from international buyers. We accept Paypal (Unlimited), Money Orders in US funds, and Wire Transfers.
Being careful, though, I went on further, and on their International Buyers page I read:
International Customers Payment Options
- 3. Bank Wire Transfer (Contact us at 718-998-6666 for further information or e-mail us at sales@47stphoto.com)
- 4. Paypal.com: Up to $200.00. Your shipping address must be Confirmed with Paypal.
The $200 isn't an issue (but why?), but the “Confirmed with Paypal” (sic) is. That's the problem: that option doesn't exist in Australia.
So last week I sent a message to sales@47stphoto.com, as they suggest on the pages. I've received no reply. Maybe that's a good thing, though it makes you wonder how completely disorganized these people are.
Thinking about that, it's clear that some big companies, at least Cameta, maintain their positive score on eBay by bullying anybody who gives them negative feedback. Thus the following relationship is probably more important than the ratio of positive to negative feedback:
| Company | Feedback received | Ratings mutually withdrawn |
| Cameta | 283,218 | 863 |
| 47th St | 168,023 | 231 |
| Wednesday, 9 April 2008 | Dereel | Images for 9 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Finally finished my planning for the irrigation system, not helped much by the discovery that the spec sheets I had for the microsprays didn't match the description in the planning guide (the one I couldn't find in the Philmac technical library, but which showed up under the obvious heading of Single Piece and 2-Piece Jets and Sprays, a page that first displays the list properly, then shrinks it to the upper half of the page, puts in a scroll bar which uses incredible CPU resources, so much that scrolling takes several seconds). It doesn't link to any spec sheet for the single piece jets I was planning to use.
Then decided to print out a list of distributors for Philmac products. I had already noticed that Midland Irrigation wasn't on the list for Ballarat, so looked again. This time I put in my post code, though it's clear that there are no distributors out in the sticks. Who cropped up? All the people I know on La Trobe St, Ballarat, including Midland Irrigation. But if I entered “Ballarat”, they didn't show up. This appears to be a problem in the search engine which relates to specific post codes, and Ballarat has a different one. What a crock!
When I got to Midland Irrigation I discovered that this didn't really matter: they had different, correct spec sheets (some of which even included mm rainfall/hour columns) and the components to match them. I'm more and more amazed that a new web site can so completely miss the point.
Finally got most things I was looking for, with the exception of low density poly(propylene) pipe and pipe saddles (not in stock), some solenoids (too expensive at Midland). Off to the Ballarat Pump Shop, where they didn't really have too much, and on to Celsius (or is that Indoor and Outdoor Trends? Or just Outdoor Trends? They use all three names, but it's the last that shows up on the invoice). There they were in the process of changing the sales software, and it took a little longer. They had the 19mm LD poly pipe in stock, at about $137 for 200 metres. By contrast, the price at Midland Irrigation would have been about $65, and even the rural grade 25mm poly pipe at Ballarat Pumps cost only $109. I can only imagine some database problem. They finally sold it to me for $70. Also picked up the remaining components.
Then off to Middendorp Electrical, where I spent nearly $500 on very few items, nearly all on 60 metres of power cable and 52 metres of conduit.
Finally back home, at about 15:45, pondering how much trouble badly implemented computer sales solutions cause. No more work today, especially since Essey Deayton (or is that now Jensen?) came along to stay a couple of days. Since we last saw her she has moved to Queensland, USA, back to Queensland, and now she's on her way back to South Australia.
| Thursday, 10 April 2008 | Dereel | Images for 10 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Spent most of the day today in the trenches in the garden. First we had to lay power cables, which proved to be very difficult for one person and very easy for two. Somehow it seemed to take up most of the day.
In the process, found one of the biggest ants I've ever seen. It measured 3 cm end to end:
|
|||||||||||||||||||||||
Put it in the fridge to slow it down, but it was still fast enough to be difficult to photograph. Put it in the freezer for what I thought was only a couple of minutes, but unfortunately it was too much, and it literally curled up and died. I was rather unhappy about that.
In the process, rediscovered a strange problem taking very close-up photos: the middle of the photo showed extreme flare:
|
|||||||||||||||||||||||
I saw this before when taking photos of my Savary bassoons. I can't work out what causes it. It seems to happen more at small apertures (this one was f/16). I've only ever tried this sort of photo with the 50 mm f/1.8 lens from the Olympus OM10; maybe there's a connection there. On the other hand, it may have something to do with the length of the exposure.
Essey did the cooking this evening, enchiladas, and so we invited the Yeardleys over. Strangely, we didn't take any photos.
| Friday, 11 April 2008 | Dereel | Images for 11 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Unpleasant news in the mail today: last month a so-called “safety” camera caught me driving dangerously on the Prince's freeway, doing all of 108 km/h where only 100 are allowed. I strongly doubt that; close to where this happened there's a speed display camera, and it shows different speeds every time I go through it. And I know there are traps there, so I'm more careful than usual. But what stupidity! I'm fined $138 for doing a speed only 8 km/h over the stupidly low limit (other freeways of lesser quality are rated at 110 km/h), and in other countries this kind of road would have limits no lower than 130 km/h. In Germany, where the road toll is lower than in Australia, it would be unlimited. “Safety”, indeed!
Spent most of the day working in the trenches, ahead of the rain to be expected in the evening. We managed to fill in most of the trenches before it arrived—what proved to be all of 0.5 mm. Now there are just the surface connections to do.
Over to the Yeardleys for dinner. David has bought a pool table, which lives on the verandah:
|
|||||||||||||||||||||||
Tonight was also the convention of the giant moths:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The first one was at the Yeardleys, the other outside our kitchen. They're about 5 cm long—those are bricks underneath the second one.
| Saturday, 12 April 2008 | Dereel | Images for 12 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work in the garden today. Finally got all the pipes connected up, so now I can at least get water at various points round the garden. Next step is to connect up the solenoids and decide how to lay out the sprinklers.
There are two issues here:
Controlling the solenoids. Three years ago I started a project for a sprinkler controller based on an old laptop and a relay board. That had ultimately died because I had burnt out the solenoids (put DC instead of AC through them), but the equipment itself was still functional—and nowhere to be found. Before I go crazy trying to find it, I think I'll buy a controller on eBay.
How many emitters should I put on each (pipe) circuit? That's actually quite a complicated consideration. Ideally the pump should run all the time, which means that the number of emitters should be matched to the flow rate of the pump at the desired pressure (which seems to be in the 150 kPa range). And how do I find that out? And what if I change a pump? I suppose the best way is to decide on a flow rate and just keep adding emitters until the pressure is maintained. And if I change a pump, I may have to add or remove emitters. What a pain!
One of the really annoying things about MacOS X is that it's built upside down: first they made the GUI, then they put an operating system underneath it. The result is that many things can only be done from the GUI. This is a particular nuisance to me, since I do most of my access to the system via ssh.
Today I did a bit of uninformed research into how to undermine the GUI do two things that I frequently want to do: unmount cameras (yes, there's the umount command, and it works fine, but the icon on the GUI is too stupid to notice when things have been umounted, and it complains anyway), and hibernate the machine (sleep(1) is, of course, something completely different). Found that ktrace exists and produces intelligible output. With its help I discovered that WindowServer starts programs from the task bar.
What about the umount command? I did a bit of searching and found a couple of likely looking processes:
Unfortunately, they don't seem to be involved in umounting; they didn't do anything useful during the process. I fear that I may have to trace the GUI itself, and I can't begin to imagine how big the trace files will be.
| Sunday, 13 April 2008 | Dereel | Images for 13 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Essey left this morning. It's nice to have visitors, but it's also nice to be by ourselves.
Yvonne wants to start using youtube, so I did a bit of looking around. The first thing that hits me is: thank God for youtube-dl! It's so much easier than the canonical interface.
In my case, one problem was that I couldn't even access it. It claimed that I didn't have Flash installed. On further examination, that proved to be correct; after my troubles some time ago, I reverted from the Linux firefox to native FreeBSD firefox, and it doesn't directly support Flash. Tried again with Linux firefox, and it claimed that I still had Flash 7 installed, though I had removed all Flash versions and installed Flash 9. Compared with the iceweasel on ceeveear, and it reported the version that I had installed on dereel, “Shockwave Flash 9.0 r115”. Further investigation showed that the plugins were, indeed, identical: firefox on dereel was lying.
Finally renamed the .mozilla directory and started again. This time it worked. Clearly there was some invalid information in the .mozilla directory that was causing it to report the incorrect version of Flash. Further investigation showed that it was in ~/.mozilla/firefox/pluginreg.dat:
1195601097000:1:13:$ Shockwave Flash 7.0 r25:$ Shockwave Flash:$
Replaced the old ~/.mozilla and confirmed that it now thought it had Flash 9. Went to youtube and it crashed the browser.
On a suggestion from Juha Kupiainen, tried nspluginwrapper, which enables the use of Linux plugins under FreeBSD. It doesn't have any docco, of course, but it's relatively straightforward to use:
=== grog@dereel (/dev/ttyp6) ~ 87 -> nspluginwrapper -v -i /usr/local/lib/npapi/linux-flashplugin/libflashplayer.so
Still, that's a pretty terrible state of affairs. Nowadays you'd expect to be able to install a web browser without this kind of trouble. Given that plugins are so difficult to install, the port should really include them by default.
Apart from that, didn't do much. It's been a tiring few days, so watched TV. Even that wasn't as peaceful as I would have liked: at 19:40 we had another power failure. Only 2 seconds, but when the power came back, teevee didn't. Further examination showed that the motherboard had been fried. Spent 90 minutes cursing and swearing and putting together an old system to replace it.
| Monday, 14 April 2008 | Dereel | Images for 14 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Not much to do today, apart from a few phone calls, one to sort out yesterday's destruction. Did some work on the sprinkler system, but there's not much more I can do until the sprinkler controller comes, and that will probably be Thursday.
Daniel O'Connor is working on MythTV, and I've also received a message from the MythTV bug tracking system about fixes to build under FreeBSD, so we set to to build the newest released version. It hasn't improved; they've added stuff with Linux-specific things such as off64_t and including Linux header files. Clearly portability is still not an issue for the developers.
| Tuesday, 15 April 2008 | Dereel –> Geelong –> Dereel | Images for 15 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
A fair amount of feedback on my comments on Apple software today. Michael (the one without a surname) wrote:
Have a look at 'man diskutil' Specifically diskutil list, diskutil unmount and diskutil mount
Indeed (and notice the misspelling of “umount”). That does exactly what I want, at least some of the time. I also get messages like:
=== root@boskoop (/dev/ttyp2) /Users/grog 9 -> diskutil unmount /dev/disk1s1Volume failed to unmount=== root@boskoop (/dev/ttyp2) /Users/grog 11 -> /sbin/umount /dev/disk1s1=== root@boskoop (/dev/ttyp2) /Users/grog 12 ->
No idea why it didn't want to umount, but it's clearly not the old “Device busy” issue. After that, the GUI went through the motions of umounting, but didn't perform it. When I then turned off the camera, I got the old “The device you removed was not properly put away” message.
The presence of diskutil unmount actually makes things worse, not better. Apple does have a command to unmount and inform the GUI. But it's not called umount! That's even more broken than I thought: if the code has been done, it should be present in umount. But it gives me the ability to write a single-line script umount to do what I want:
diskutil unmount $*Michael also writes:
If you run you queries through the google debugger first, you might save a lot of time in trace... :) This answer found via google search for 'mac osx unmount from terminal'
That's true, but then people like Michael are even more efficient at finding solutions.
Callum Gibson came up with another one on IRC:
<callum> groOgle: oh yes. I was going to say I have program for mac to make it
go to sleep. I think it's called SleepNow.
<callum> Very useful for when you tell the kids to get off the computer and
they keep saying "just wait, just wait..."
<groOgle> callum: :-)
<callum> http://www.snoize.com/SleepNow/
This is based on MacPorts, a clone of the FreeBSD Ports Collection, and I don't have it installed, so I'll try it some other time.
In other mail, Joacim Melin writes:
Maybe this could be an idea for your garden? Silly name but it looks cool: http://www.engadget.com/2008/04/15/cyber-rain-xci-gives-you-wireless-control-of-your-sprinkler-syst/I know I will look into it for my garden (2600 square meters).
This is a computer controlled sprinkler, something like the project I was working on a couple of years ago. To summarize, it:
It connects wirelessly rather than by cable.
My equipment is in a metal shed, and though to my surprise I do have wireless connectivity (small gaps in the metal, soon to be closed), it's marginal; that's why I'm putting Cat 6 there, at least to connect to my brewing hardware.
It checks the weather forecast and bases the sprinkler schedules on that.
My system doesn't do that, and I think that's a point for my system: it's an open-loop approach. I frequently rant in this diary on how inaccurate the weather forecasts are here; they certainly don't say how much rain will fall, and that's a good thing, because they're seldom correct.
Also, the device doesn't seem to factor in wind speeds. Evaporation is one of the main issues here; on an annual basis it exceeds precipitation almost everywhere in Australia. To get round that, we mulch heavily. And how much water remains in the ground depends on the porosity of the soil. None of this seems to be taken in in this approach, which IMO makes it worthless.
In Australia, better systems (but not the one I've bought) have inputs from rain sensors, which is a small start, but still better than the approach that this system appears to take.
The correct answer, of course, and one that I've been considering for some time, is the closed loop approach: measure the soil humidity and water accordingly to maintain a particular level. But nobody seems to offer the appropriate sensors.
On a slightly amusing side note: my introduction to microprocessors, in 1975, was an Intel 8008 based system that was originally designed for exactly this purpose. It did have humidity sensors.
While I was outside, a phone call came from Andrew Mackay, one of the people who sued the American Saddlebred Horse Association of Australia. He claimed to have been successful in the case, something I hadn't heard, and which was subsequently denied by a representative of the ASHAA. I had voiced my opinion not only of their actions, but also of their invasion of my privacy, last December. He was calling to ask me to take down that entry. They should have thought of that before passing on my mail address to the spammers. I certainly see no reason to cooperate with people whom I consider unworthy of any contact.
One of the things I didn't mention last December was their abuse of the legal system to bully people into doing what they want: they had already bullied Chris into taking down a diary entry last year. In view of the court action and the potential legal costs, she found it better to do so. Possibly they intend to try the same with me, but I don't have any property, so they can't expect to threaten me with the cost of legal action—that would all be their own expense—but they still annoy me greatly. I suppose I should at least have offered to him to correct any factual inaccuracies, though.
In the afternoon to Geelong to the dentists, which took up most of the afternoon. Found a Vietnamese grocery there, and spent far too much money. I don't like Geelong: it has a somewhat rough feeling about it, the shopping centres look half-finished, and even the approach to the waterfront is less than pleasant:
|
|||||||||||||||||||||||
| Wednesday, 16 April 2008 | Dereel | Images for 16 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Didn't do much today. Baked some bread, did some work on my web pages, and also a bit more work on the irrigation system. Somehow I can't be overly bothered. I suppose things will change when the controller is here (currently it's waiting for me in Sebastopol).
| Thursday, 17 April 2008 | Dereel | Images for 17 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Mail from Achim Patzner today, talking about the disparity between Mac OS X umount and the GUI application, apparently a visible part of diskarbitrationd. He explains that umount has to comply with the UNIX specifications, and that as a result it can't inform the GUI (on Apple in this case) about what it has done. I find that hard to believe, though standards can be funny things. But there's no reason why the GUI application should not check things properly. It sees that the device has gone away; why doesn't it check whether the file system has gone away too before producing incorrect error messages?
Into town for a variety of things, including picking up my new sprinkler controller. It didn't match the description on eBay: I bid on a 6 station controller without any mention of a rain sensor input, but got an 8 station controller with rain sensor input. I can't complain about that.
Also had a haircut and took in the damaged computer stuff for a repair quote; that alone will cost $45.
Back home, pottered around a bit looking for connectors for the solenoid cables. Didn't find any, but I did find the old relay board for the sprinkler project. I already had the laptop, so all that's missing is the power supply. Still, with my motivation it'll be easier to install the sprinkler controller, and I can think about the rest later.
| Friday, 18 April 2008 | Dereel | Images for 18 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Another bloody power failure! That's the third in 11 days. When I moved here from Wantadilla, I thought that the power supply situation was better than with ETSA, but since the beginning of the year we've had 9 power failures, and they seem to be increasing. In addition, the damage to my computer on Sunday indicates that they're becoming more dangerous.
Spent most of the day installing the sprinkler controller and the first sprinkler section, in the garden bed to the north of the house. The installation itself went fine, but I spent most of the day looking for tools and then mislaying them. I really must get more organized.
| Saturday, 19 April 2008 | Dereel | Images for 19 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work in the garden today, setting up the second section of the sprinkler system. As I had expected, that took longer than others might have expected: getting the positioning of the sprinkler emitters is quite tricky, not helped by lack of adequate documentation. The documentation on the Philmac web site doesn't seem to relate to what they sell, and the only information on the wide sprinklers was that they had a sprinkler radius of 3.5 m at 150 kPa.
What I found was different: the pump was cycling between cut-in pressure of 220 kPa and cut-out pressure of 350 kPa, and the radius was closer to 2.5 m than 3.5 m. Do I have that much head loss in the system? Maybe; a pressure gauge would be interesting. That would also mean a maximum run for the 19 mm low pressure poly that I'm using.
Learnt a positive thing too: it's trivial to move sprinklers around, so did a fair amount of that. Now, of course, I need more sprinklers. That'll have to wait until Monday.
| Sunday, 20 April 2008 | Dereel | Images for 20 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
In the following, you'll probably want to enlarge the images by clicking on them.
Now that Yvonne is taking video clips, it makes sense to be able to display them on the web. But of course, firefox can't display them; it needs a helper application. Which one? That depends on the file format, and the correct thing to do is to look at it and determine the content. Unfortunately, file(1) is too stupid for that:
=== grog@dereel (/dev/ttyp4) ~ 119 -> file ~yvonne/Photos/20080419/DSCN2406.MOV
/home/yvonne/Photos/20080419/DSCN2406.MOV: data
I have a little script called mpid that is a little more intelligent. It uses mplayer and it's primarily intended to find information about MPEG-2 streams:
=== grog@ceeveear (/dev/pts/0) /myth/Images 23 -> mpid 1020_20080420125500.mpg
1020_20080420125500.mpg VIDEO: MPEG2 1280x720 (aspect 3) 50.000 fps 9600.0 kbps (1200.0 kbyte/s)
So this is a 720p stream.
It didn't give much useful information with Yvonne's image until I used the -v option, when I got all this and more:
=== grog@ceeveear (/dev/pts/0) /myth/Images 21 -> mpid -v ~yvonne/Photos/20080419/DSCN2406.MOV
Quicktime/MOV file format detected.
VIDEO: [jpeg] 640x480 24bpp 15.000 fps 0.0 kbps ( 0.0 kbyte/s)
No idea what Quicktime is, but it's clear that mplayer can play it. That's not firefox's way, of course. Web browsers give you the opportunity to trick them by believing that the name you give the file has something to do with the contents. So I had to tell firefox how to handle file names that end in .MOV.
I really, really hate firefox! I've complained about it on numerous occasions, and really there's nothing new today except my recognition that I am no closer to accepting its behaviour than before. It seems to be getting worse, not better, in its understanding of computer environments. At one point it at least added the possibility of putting real file names into its stupid dialogue window for choosing the auxiilary programs. Now that's gone again; what hasn't gone is that it starts searching for the programs in my home directory:
|
||||||||||||||||||||||
Even then, it shows all files, even if they're not executable. To find, say, /usr/local/bin/mplayer, I need to double click back to the root file system, then double click through each individual directory until I get to the file. What kind of brain death is that?
Instead, chose a random file (in fact, ~/Bug/nfsmount, which isn't even executable; it didn't complain about that, not even when I tried to display the file with it. Then I went looking for the file where the name nfsmount showed up; it proved to be ~/.mozilla/firefox/7v0n6ir5.Default User/mimeTypes.rdf. That's right, with spaces in the file name, which meant that the first time through my search failed with things like
grep: .mozilla/firefox/7v0n6ir5.Default: No such file or directory
Yes, with a few special incantations I can modify the behaviour of these tools to work around the breakage. But it shouldn't be necessary. Why do people do these things? Are they completely illiterate?
It seems you can change the name of the directory, as long as you update ~/.mozilla/firefox/profiles.ini. Now the directory is called ~/.mozilla/firefox/7v0n6ir5.horrible_broken_firefox_with_no_understanding_of_UNIX. It doesn't make me feel much better, unfortunately. When will somebody who doesn't think in terms of mice come up with a usable web browser? To calm myself down a bit, moved firefox up a rung in Dreckstool.
Later in the day I tried to pay a bill with ANZ's horrible web interface. It gives me a full screen display with just the top middle occupied:
|
||||||||||||||||||||||
Why it wants to fill the screen is beyond me. It probably says more about the web programmer's assumptions or prejudices than anything else.
Went to the “Pay anyone” screen and filled out the details. Here are the relevant parts (obviously dummy data for publication):
|
||||||||||||||||||||||
The validation scripts didn't like what I had to say:
|
||||||||||||||||||||||
Why? Are they still using the same 48 character set that was in use in the 1960s?
|
||||||||||||||||||||||
I suppose that's marginally acceptable, but only marginally. It's not very clear: the example format contradicts the “greater than zero” requirement. What it's trying to say is “Please enter values with a decimal point followed by two digits”.
Fixed that and got the next message:
|
||||||||||||||||||||||
That's completely stupid. It's complaining about the comma in the number. This is exactly the format that the ANZ page uses for representing values of at least $1,000 (for example my balances, not shown here, are in that format). Not only is it clear, it also enables competent programmers to add a level of protection. If I had written 9,9999.00 it would be reasonable to assume that I had suffered from keyboard bounce. Without the comma, there's no way for the software to recognize this fact. With it, it's clearly invalid.
Corrected that one, and things moved on from Java to Melbourne. It still didn't work: I was asked for an account name and description. For me, the account name is the description, so I left the description field empty. It complained (only the first time, unfortunately, so I can't show the message here). So I entered the description “Required field”, which it used instead of the account name.
So I went back to the previous screen with the Back menu selection. Don't do that! The ANZ programmers haven't worked out how to handle duplicate PUTDATA. The message I got was multipurpose and vague. My comments in italics.
Why Your Session Has EndedTiming out of sessions
The most likely reason is that you have exceeded your set time limit for submitting information.
OK, what's the timeout? By how long?
Other possible causes
You may have:
- Resubmitted a request.
OK, why should that log me out? Even if the ANZ programmers can't handle normal web navigation, they should be able to return to a known state.
- Asked for a page from a banking session that had ended.
How? And again, why should that log me out?- Used a web browser that is not fully compatible with this service.
“We can't program, so this is your fault.”
So I returned to the login window (in another page) and tried to log in again. I was taken straight to the “session ended” page. Presumably the page had some data associated with it.
So I went from my local bookmark to the page again and logged in. This time it worked.
I went the Edit Payee page and changed the description. Then I returned to the Pay Anyone page, filled in the details from the saved payees, and tried again. I got a malformatted message (gratuitous capitals, no mention of what the BSB is supposed to be) that the payee account exists with a different name.
Back to the Edit page. Nope, everything there was correct.
Back to the Pay Anyone page. This time it worked (so I also can't show the messages). This may have been a database synchronization problem, which in itself is horrifying enough.
All in all, it took me about 45 minutes to pay. This isn't the first time I've had trouble with ANZ; every time I try to use this terminally broken Verified by Visa application, I have similar problems.
Spent much of the day working in the garden. It's clear that the incorrect radius of the sprinklers is causing significant problems.
Found my first bug in the firmware of the Olympus E-510: exposure compensation is not registered in the EXIF data if flash is in use. Here are three photos taken with exposure bracketing (0.7 EV):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
As the EXIF data for the first and third photos show, the Exposure Bias is reported as 0.0. It works correctly if flash isn't enabled.
| Monday, 21 April 2008 | Dereel | Images for 21 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
I had intended to go into town today, but I found myself with lots of documentation to do (not the least yesterday's rants, of course), and by the time I was done, it was lunch time.
Yvonne into town after lunch with Darah to have her hoofs looked at again, and I joined her a little later. Kurt Enzinger is back from his honeymoon, and he voiced the opinion that maybe what she has now isn't Greasy Heel after all: indeed, it's not greasy, and there's a little pus. He took a biopsy, scraped off all the scabs under sedation, and now we have to leave the legs bandaged until the results of the biopsy come through, probably Friday or Saturday.
Then I on to buy more fittings for the sprinkler system. Somehow the smallest details take up the most time: one run of the pipe will go along the base boards around the outside of the house, and I need to screw clamps to them. Simple enough, and I had no difficulty finding the clamps ($0.25 each in metal, $0.35 each in plastic), but what screws do I use? The salesperson at the hardware shop recommended screws 40 mm long, which would go right through the board. That seems a little excessive to hold a 2 metre length of plastic pipe that will weigh about 300 g per metre when full of water. And then there was the question of corrosion protection: zinc plated, galvanized (yes, there appears to be a difference, probably the way the zinc is applied) or “golden”? In the end, disregarded the advice of the salesperson and bought some zinc-plated screws 12 mm long.
On to Midland Irrigation to discuss the throw length of the rotor sprinklers (supposed to be 7 m, is 5 m). It seems that the rotors require a pressure of 300 kPa, while the jets require 150 kPa. Why do they make different fittings with such drastically different pressure requirements? They work at the estimated 200 kPa that I have in the system, but my distances are all wrong.
Back home, but didn't get much done before it got dark. Gradually I'm filling in the gaps.
| Tuesday, 22 April 2008 | Dereel | Images for 22 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
James Andrewartha commented on one of my recent rants:
Hit Ctrl-L (or the icon in the top left) to bring up a text-entry widget in file open dialogs.
There's no icon at top left, but Ctrl-L does seem to work:
|
|
||||||||
That's helpful, but it highlights all the weaknesses:
It doesn't understand the PATH environment variable.
It doesn't understand execute privileges, and lists files that can't possibly be executed.
It's not documented, not even a “press Ctrl-L for a sane way of navigating this tree“.
Even then, it's clunky; it does file name completion, but it highlights the completion, which slows it down by a factor of about 20.
While checking this, I performed the following sequence:
Stopped firefox on tvremote, which I was using for this test.
Started iceweasel on ceeveear, a Debian Linux box: James uses Debian, so I thought there might be something different there. iceweasel is a version of firefox that doesn't have copyright issues. And yes, there is an icon where James describes, and it produces an input field in the same window. So this is just a way of showing how the interfaces can change from one platform to the next; that was part of my initial complaint.
Started firefox on teevee again. It didn't start! Instead, it got the iceweasel to open a new window. Is this really helpful? What are the security implications?
Spent much of the day adding sprinklers to the system, and now I have everything set up that I had planned for the first stage. There will be more when we know what we want to do with the space to the south of the house, which currently looks like this:
|
|||||||||||||||||||||||
After that, spent the rest of the day planting things that will hopefully look OK come spring.
| Wednesday, 23 April 2008 | Dereel | Images for 23 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
| Topic: technology | Link here |
Another message from James Andrewartha today, partially pointing out some of the misassumptions I made yesterday. He also says:
A new iceweasel window instead of firefox is the designed behaviour, due to the mozilla “remote” system that I think works over X. Use firefox -no-remote to work around it, but this might go badly if you've got the same home directory nfs mounted on both machines.
That may be; I hadn't expected that it was a bug, just a bad misassumption on the part of the designers. I really can't see any justification for this “there may only be one” attitude. It's good to have ways to solve the problems, but why introduce them in the first place?
|
Preparing the east bed
|
Topic: gardening | Link here |
More work in the garden, which kept us busy most of the day. At least we now have some things planted, nearly all of them cuttings from existing plants in the garden (the only exception was the bulbs that I had exchanged last month.
It's my guess that this is the day we planted the Gazanias and Dianthus to the west of the garden path.
Despite multiple attacks with glyphosate, weeds still continue to come up, so put in a layer of old packing cartons with only holes for the plants:
|
|||||||||||||||||||||||
They'll get covered over later, of course.
Some of the plants show remarkable ability to grow from cuttings. Last Friday we had to prune a number of bushes in the north bed, including a succulent that blooms bright red in spring (and maybe longer if it gets enough water):
|
|||||||||||||||||||||||
We had put the cuttings in a bucket of water; today, 5 days later, many of them had developed root shoots up to a centimetre long:
|
|||||||||||||||||||||||
While digging in the garden, I find a number of weird and wonderful life forms. This one is obviously a spider, about 1 cm long:
|
|
|||||||||
And autumn doesn't mean just yellow leaves (though the birches are developing them):
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
| Thursday, 24 April 2008 | Dereel –> Geelong –> Dereel | Images for 24 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More mail on the topic of firefox today, this time from Marco Perez:
The problems with the so-called filepicker in Firefox you're describing may be eased a bit. The default filepicker is in fact the one from GTK2, which many users love to hate. But the Mozilla browsers include a native XUL filepicker as well. You can activate it in the advanced browser settings. To do so enter the following in the address bar:
about:config
Then enter the following into the filter bar near the top:
ui.allow_platform_file_picker
Indeed, that was it. It seems that the old version I was referring to was the native “file picker”.
The other objections remain; while playing with it, I found a new one:
|
|
||||||||
I don't use /tmp for anything except small files, but the various browsers seem to love it, and I can't find a way to change it, let alone “try saving in a different location” (is this yet another word for “directory”?). Searching through the about:config shows that I can indeed change it on dereel (preference mozex.general.tempdir), but apparently not with the version of iceweasel that I have on ceeveear.
A bit more work in the garden, finding interesting things in the process:
|
|||||||||||||||||||||||
|
|
|||||||||
I have no idea where the pipe running to the top of the photo comes from; it's the same kind as the stuff I have been laying, but it was there already, and I wasn't aware of either end. This will document it.
The grub is one of many I've found in the ground; they're about 3 cm long. I wonder if they're beneficial.
Into Geelong for another dentist's appointment for Yvonne in the afternoon, running into the inevitable unnecessary traffic controls:
|
|||||||||||||||||||||||
This came after 3 km of empty highway with a 60 km/h limit. Why do we need traffic control on a road like that? Because there are machines on the side of the road. Sometimes I wonder if regulations aren't completely replacing road sense.
In Geelong, had intended to do lots of things while Yvonne had her teeth looked at, but in the end only managed to do some more shopping in the Vietnamese grocers, and even then didn't get everything I wanted, and was late picking up Yvonne.
On the way home, did a little shopping, and was faced with yet another example of obfuscatory terminology:
|
|||||||||||||||||||||||
What's the difference between buk choy and pok choy? Looking at the vegetables, they're pretty much the same, but one is darker than the other. According to the information I have on my page , they're the same thing. Wikipedia only knows the name bok choy, but at the moment that page is broken anyway. I suspect that these are two different suppliers, one using the real name and one the stupid name, and Safeway hasn't been able to tell the difference.
In January 2009, James Andrewartha pointed me to a web site that suggests that the NSW bureaucrats have mandated this distinction too.
We've been buying a fair amount of new kitchen equipment recently: last week I bought a table-top cooker at the same Vietnamese grocery, ostensibly for bulgogi, but it came without the corresponding grill fixture, one of the things I looked for in vain today. This week ALDI had a new raclette grill and a bread making machine on special, so picked up both of them: the plate on our old grill is wearing out.
I've been baking bread in the oven for a while now, but the idea of a machine to take over the timing work for me sounds like a good idea. Also ALDI's two month no-argument return policy gave me the security that I needed to decide to buy it.
I've never even examined a bread maker before. What I found didn't encourage me:
The bread maker doesn't handle humidity; it's up to the user to add the correct amount of water. How much? On the bread mix packet I had it stated 420 ml. The measure I got was clearly inconsistent with itself: the 2 oz (fl oz?) mark corresponded to about 45 ml (should be 57), and the 4 oz mark corresponded with about 105 ml (should be 113 ml):
|
|||||||||||||||||||||||
Further investigation showed that “50 ml” was really 60 ml. I've never seen such an inaccurate measure before.
On the positive side, the measure only had space for about 200 ml, less than half of what's needed to make bread, so I would never have used it anyway.
In passing, it's interesting that the bread mix (from Laucke in Strathalbyn, just across the road from where Yana went to school) comes in 600 g sachets, to which you add the mentioned 420 ml of water, for a total weight of 1020 g. The packaging suggests that it's a 1 kg loaf, which would mean losing only 20 ml to evaporation. That's a nuisance, since no bread machine bakes such round figures; they're still firmly rooted in the middle ages. This one is typical: loaf sizes are 900 g (32 oz/2 lb), 1150 g (40 oz/ 2½ lb) and 1300 g or 1350 g, depending on where you look, or 48 oz/ 3 lb. I chose the 2½ lb setting more by accident than intention, but of course when it was finished it weighed only 880 g:
|
|||||||||||||||||||||||
Are they giving the weight before baking? That makes no sense at all, and makes it almost impossible to compare with commercial bread, which gives the real weight of what you buy.
The dough is first kneaded, then baked, in the same container. This leaves the kneading paddles sticking into the bread, and you have to pull them out manually:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
On this occasion at least, the dough rose high enough to coat the lid of the machine, requiring considerable trouble (soaking) to clean off again:
|
|||||||||||||||||||||||
There are very few options for controlling the way the bread is made, and clearly it's not baked. The results look pretty ugly, particularly considering the holes in the slices:
|
|||||||||||||||||||||||
How much of this is the fault of this specific el-cheapo bread machine? Certainly the unbelievably bad instructions are. But I suspect that the paddle issue is general, and probably the colour (though all, including this one, offer the option of making the bread darker). The loaf looks quite like the loaf on Laucke's home page, so possibly that's the way it's intended to look. In general, the results are an order of magnitude worse than doing it manually, and the problems cleaning the machine mean that it also doesn't save any effort. It'll go back.
One thing that is probably not the fault of the machine is the dough on the cover. When making the bread, I took one package of bread mix and one package of yeast. The package originally came with four packets of bread mix; only later did I discover that there was only one package of yeast, intended for all of the mixes. So presumably with the correct amount of yeast it wouldn't have risen as much.
In this connection, it's worth considering a test I read in the September 2007 issue of Choice magazine, which had been sent to me as part of a trial subscription. I had found the reports extremely superficial, and so didn't continue with the subscription. My experiences today confirmed that impression: no mention of the paddles stuck in the bread. Do they all do it? Does only the ALDI machine do it? I doubt the latter. The test also only tested performance for making this ugly white bread, something that I probably wouldn't do again.
| Friday, 25 April 2008 | Dereel | Images for 25 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Darah's hoof: not Greasy Heel
|
Topic: animals | Link here |
Call early this morning from Kurt Enzinger, the vet, despite the public holiday: the biopsy of Darah's hoofs showed that it was not Greasy Heel, but a “photo-activated vasculitis”, probably caused by excessive ultraviolet light. So change (and simplification) of treatment: the legs will need to be bandaged for a while, and we'll apply some ointment, but we won't need to scrub as much as before. One thing is good: the coming winter will improve the conditions for the current condition (less light), while they would make it worse for Greasy Heel (moister).
Into town in the afternoon to find the injection that Kurt had left for us outside the office at Miner's Rest. It wasn't there. Tried the office door, and to my surprise it was open, though there was nobody inside, only a friendly smoke cat. Finally somebody showed up and started looking for the stuff, but while they were doing it I found it outside: the bag was so light that it had blown away.
|
New trees
|
Topic: gardening | Link here |
On the way home dropped in at Avalon Nursery in Haddon to look for a couple of Callistemons (or is that Callistema?). Left after spending far too much money with two Grevilleas, a Kaffir Lime tree and a strange-looking plant calling itself Sapphire Dragon, which proved to be a Paulownia kawakami (or is that Paulownia kawakamii? Googlefight seems to think so), which on the photo looks something like an enormous jacaranda.
|
Raclette: no progress
|
Topic: food and drink | Link here |
Tried out our new raclette grill in the evening. Getting real dry-cured ham in provincial Australia is almost impossible, so we ended up with salami instead, which surprisingly didn't taste too bad.
The grill itself was another matter: on the face of it it's very attractive, with an oval form allowing 8 trays instead of the more common 6, and with a thermostat. In practice, though, it just didn't get hot enough, and it was also uneven in heating. When I finally got one part of a tray grilled (over-grilled, in fact), the other end was still only barely cooked enough:
|
|||||||||||||||||||||||
Comparing the old and new only gave a partial answer:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The element on the older grill is further out from the middle of the tray area than on the new one, explaining the uneven grilling. But why is it so slow by comparison? I had suspected a lower power rating, but according to the specs it does 1200 W, while the old one only does 900 W. I wonder if it's honest.
| Saturday, 26 April 2008 | Dereel | Images for 26 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Woken early this morning by the arrival of Gordon, our farrier, one of the few people we deal with who doesn't have a Germanic name. Nevertheless he seems to have German habits: he always shows up in the middle of the night.
Planted the trees we bought yesterday; the garden is filling up. In the process discovered a couple of identical-looking spiders, about 4 cm across, who promptly attacked each other:
|
|||||||||||||||||||||||
By the time I got my camera one of them had burrowed away.
I'm still trying to come to terms with the necessity of having to dig up the entire 300 m² of garden to suppress the weeds. In the meantime, decided to clean up the ground cover under our mystery yellow flowering tree. The immediate ground cover the “succulent daisy” that we've planted in many places. One of the disadvantages is that old growth dies and new growth comes up over it, getting weaker as time goes on. There were also a couple of bearded irises in the area, so I decided to remove everything and start again:
|
|||||||||||||||||||||||
The couple of irises were literally the tip of the iceberg:
|
|||||||||||||||||||||||
I wonder where I'll find space for all of them.
Also came across some bulbs that must have been there for ever, but couldn't work their way through the undergrowth. Planted them in one of the planting boxes:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
I wonder what they are.
| Sunday, 27 April 2008 | Dereel | Images for 27 April 2008 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More baking today, with mixed results. After the disappointing results with the ALDI bread maker, decided to see what it was like for just kneading dough. Once again, the results were disappointing. Tried a French-style baguette, but greatly misjudged the amount by which it would rise during baking:
|
|||||||||||||||||||||||
Then tried a rye bread, which showed a number of the weaknesses of the machine:
Inserting the baking pan is non-trivial: the drive for the paddles needs to be aligned correctly, and just finding the correct position is difficult. On this occasion (but, surprisingly, not before) it took me four attempts to get both paddles to engage, not helped by the lack of firm confirmation that they've engaged.
On this occasion, the removable paddles somehow got dislodged; only one of them gripped, I was able to turn the device off (it holds state for 15 minutes) and reengage the things, but that was only possible because I was watching it. Normally this would have resulted in a wasted loaf.
It only allows the dough to rise once. To do a typical two-rise loaf you have to knead it again. This may have been the explanation of the extreme size of the previous loaf.
The problem with the dislodged paddles and the difficulty in inserting the pan is enough reason to return this particular bread maker. Peter Jeremy tells me that his has fixed paddles, but one way or the other they all leave holes in the bottom of the bread.
More garden work, and finished the bed with the ground cover. I'm gradually coming to terms with the idea of digging up the entire garden.
| Monday, 28 April 2008 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Last month it was really warm, but things have changed. The following graph shows the ambient temperature in the brewing shed over the last 2 months. The extremes are probably 5° either side of this value, so the maximum on 16 March was round 40°, while the minimum today was round 0°:
Did some work in the garden anyway, mainly digging. I'm wondering how many of the things we had intended to transplant would actually survive at this time of year; we might be better off preparing the soil and transplanting in early spring, when there are no more frosts.
From time to time I've had problems with my X display on eureka, which is connected to dereel's mouse and keyboard via x2x. It used to be sporadic, but in the last couple of days it's been happening all the time, and it effectively makes the window manager useless.
The problem seems to be in the reported modifier keys. My displays have the standard 5 modifier keys, which are mapped to bits in the range 0 to 0x1f. What I'm seeing are spurious modifier keys in the range 0x100 to 0x700. Here's some debugging output that I put in there, followed by a standard error message:
Button state: 0
Button state: 100
Button state: 300
Button state: 700
Button state: 700
Button state: 700
Button state: 700
X Error of failed request: BadValue (integer parameter out of range for operation)
Major opcode of failed request: 147 (XTEST)
Minor opcode of failed request: 2 (X_XTestFakeInput)
Value in failed request: 0x78
Serial number of failed request: 57987
Current serial number in output stream: 57992
Part of my trial modifications was to limit the range to 0x1f, but I think that's the wrong end of the stick: for some reason, the X server is reporting these values to x2x and believing them itself. Why that should be so is beyond me. Unfortunately, I can't upgrade it to the latest version of X, because I have these two BenQ P992 monitors connected to it that report their highest frequency incorrectly, and recent versions of X believe it and offer no way of overriding it. sigh
The cold weather is exacerbating the problems with the temperature control of our air conditioners. I must finally get something done about that.
| Tuesday, 29 April 2008 | Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
My satellite connection is getting no more reliable; we're continually being disconnected for up to a quarter of an hour at a time. I wish people would finally install ADSL here.
More work in the garden. Finally things are beginning to look more like a garden, though there's still much digging to be done.
To the Yeardleys for dinner in the evening.
| Wednesday, 30 April 2008 | Dereel | Images for 30 April 2008 |
| Top of page | ||
| previous day |
Still more work in the garden, and transplanted a lot more flowers, some of which turned out to have long root systems much longer than we had expected:
|
|||||||||||||||||||||||
I don't know if it'll survive, but Yvonne didn't like it anyway, so it had to go. The Yeardleys got it, so hopefully it'll come good.
Our Sapphire Dragon is deciduous, and today it lost a leaf, which I put to good use:
|
|||||||||||||||||||||||
More firmware breakage: my new sprinkler controller has a function to multiply the watering times by a factor between 0.5 and 1.5 to cater for differing seasonal requirements. But when it's set to anything except 1, the manual “run program” doesn't work any more. I think I'll have to start a list of all these strange little things attributable to firmware.
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 |
