It's New Year's Day yet again. These things keep happening. I don't know why, but I find
the day depressing, and I have the feeling that I have a hangover even if I haven't been
drinking. Somehow didn't get very much done today.
The insulation in the ceiling has been there for 3 weeks, and it shows:
bits of it are falling out of place. Not a serious issue, I suppose: when then put in the
plaster, they can reposition it.
I've been taking a number of free courses on biology and related concepts lately. One,
Introduction to Genetics
and Evolution, starts today (really tomorrow, since it's in the USA). But they've
released the first lectures already—nearly 1½ hours of lectures proving that evolution
really exists!
What sane person can doubt that? I know that a surprising number of people in the USA
really believe the Torah's explanation of
the origins of the world, or this
nonsensical Intelligent Design
concept. But in the Real World?
It seems that in all countries there's a sizeable minority of people who reject the theory
of evolution, in particular the idea that humans evolved from animals:
In Turkey, it seems that it's not a minority who reject this assertion: it's a majority.
And even in Australia, only 42% (or maybe 71%) of people believe in evolution.
But OK, anybody participating in this course won't believe this nonsense, right? Wrong. A
surprising number of participants disagree, or find no proof of the theory. I'm amazed.
Hopefully they'll change their minds by the end of the course.
We've been talking with Garry Marriott about putting some horses on his south paddock to eat
up the grass, but somehow hadn't got round to doing anything about it. Today I got a very
garbled (mobile) phone call from him, in which he told me that his granddaughter Emily was
planning to have the paddock slashed on Monday, because she had received a notification from
the council. She even wanted to do the area round the dam, where we have horses.
Can we (or, preferably, our horses) eat it down by then? Over with Chris Bahlo to take a
look, and to put the horses on the dam area. The latter was easy: they were already there,
so we just needed to close them in. Over to talk with Emily and (mainly) her partner Sean,
and took a look at the letter. It had arrived nearly a month ago, but since it was
addressed to Garry, they didn't open it. But the content was very interesting:
Lot 2? That's our property! Sean and Emily didn't know that, of course, but the
council certainly does, judging by the rates notices we have for it. And as a result of the
misdirection, we only have until Sunday to do the work.
That's not possible. Today and tomorrow are days of total fire ban, which includes
slashing. So what can we do? Called up the council (that's the number at the bottom) at
14:49 and was told to call back during business hours. Forget exactly what they are, but
they included the time I called. Clearly they have shut down the council for Christmas and
not bothered to change their message.
OK, it'll have to be Monday. How can the council be so disorganized? Will I get an
extension? On the face of it, it looks as if we should get Sean's mate to do at least the
front of our property too.
So it's fairly clear where the blame lies with my DHCP problems, though I still need to
trace the other connection. I'll do that when the temperature has cooled down a bit. But
there's another issue, and potentially it's related: from time to time ssh setup
fails. More tracing, with very little to show for
it. This is on the server side:
For no apparent reason, the remote sshd has just closed the connection. Why? It
doesn't say. There's nothing in auth.log, which is arguably not the correct log file
anyway, since it doesn't get as far as trying to authenticate, but there's nothing elsewhere
either. And a retry usually succeeds, though sometimes it takes up to 6 attempts. More
head-scratching.
Looking more closely, they were a roughly mixed group
of Sulphur-crested
cockatoos and Corellas. Yes, they're also cockatoos, but it's surprising that they flock together.
I've been looking for good kitchen scales for some time. It's certainly not an issue of
price: the last one I bought cost $10.86 including shipping, and it works well. So why do I
need a new one? It's tiny! The bowl that comes with it is tiny too, and even
so it obscures the display as often as not.
This week ALDI has a set of scales on special.
What size? What range? Who cares? They're nutritional! And they come in three colours,
just what we want:
But then, the best thing about ALDI is the 60 day no questions asked return guarantee, so
when Yvonne went into town to pick up some horsey stuff, she
stopped in and bought one. So did Chris Bahlo, because she liked the bright red one. So
much for my irony.
And how are they? The size is OK, and they go to 5 kg in 1 g increments. And it comes with
codes for 958 different foodstuffs, so that you can work out exactly how many calories their
specimens of the same foodstuffs would have delivered. I thought that stuff went out of
fashion decades ago.
More to the point, it didn't work. I powered the thing on and watched the no-load display
start at 0 and climb to about 10 g before I turned it off. Heat problems? It was very hot
today, and maybe it doesn't handle changes of temperature well. On the other hand, Chris
had no trouble with hers. So an hour later I tried again and watched a different behaviour:
it jumped at random several hundred grams in each direction.
Played around with the batteries, and the problem went away. So somehow the battery
connections are not good, and it's too stupid to notice. Now we'll have to see how useful
it is. Yvonne likes it because it's easy to clean (she says). The dozens of irrelevant
numerical displays irritate me somewhat.
It's high time I got round to upgrading eureka, my main machine. It's still running
FreeBSD 9.2, and I've been meaning to upgrade
it for nearly a year. Today I got as far as connecting my spare BenQ monitor to two test machines. It wasn't easy. BenQ monitors say “No cable connected”
when in fact they mean no signal. But there was a signal, as other monitors confirmed. Was
the thing set up to only accept input from one source? The menu would tell me that. But
the menu didn't display.
After much cursing and monitor swapping, discovered that it was set
for HDMI input, and that for some reason the
menu wouldn't display under these circumstances. But I could still switch between the
inputs with the Enter button, and in the other modes the menu works. Is this some
misfeature of HDMI, or just firmware breakage?
One of the most consistent weather predictions we've had recently was that yesterday and
today would be very hot, and that it would end with a storm. And so it was. As the rain
approached, I predicted on IRC:
<grOOgle> <15:57:41> Prepare for the next flower pail.
At that time, the outside temperature was 36.4°. An hour later it was 22°, a drop of 14.4°,
or nearly ¼° per minute:
And then the power failure came, killing 200 days of uptime. Grrrr, especially since I went and got the emergency
generator, and we were able to continue with cooking more or less normally. I'm surprised
how resilient this generator is: it was able to run the water pump with no trouble, and we
even ran a microwave oven, which, however, wasn't very happy about the quality of the power.
The Kenwood Chef refused, however:
the control circuitry appears to be very primitive. The first time round, it was unable to
maintain constant speed. On the second attempt it ran full speed, independently of the
setting. Damaged? No, fortunately. After the power came back, it ran normally.
The power came back after about 2¼ hours, and that's about the time I needed to
get eureka back up and running again. I really must save my config files and check
that the machine can start up without so much manual intervention. But then, that's part of
the plan for the new machine.
The storm last night sounded violent, but there wasn't really that much to show for it—lots
of scattered eucalyptus bark and a couple of small twigs broken off:
Over to look at Chris Bahlo's building site today. Construction was to have commenced on
5 December
2014 and be complete by 13 March 2015. Now over a third of that
time has passed. What does the site look like?
It's been months since Chris Bahlo complained to David Yeardley about his sheep running wild
over her property, and a couple of
weeks ago I saw that David had noticed it, and thought that they would finally stop
things. As the photo above shows, they haven't. Called up the Council and was put through
(ultimately) to Neville, the ranger, who told me that they would pick up the sheep after
getting a copy of a letter to the Yeardley's demanding the removal of the sheep (in 24
hours!) along with proof of delivery. Does Chris want to do that? It's just pain.
Wouldn't it be nice if the Yeardleys would behave like responsible members of society?
Called up the Golden Plains Shire
Council to speak to Simon Howland, who issued the incorrect fire prevention notice. Not
back until “Monday”, by which she meant a week's time. So how had he intended to verify
that the work was done by yesterday?
That's all that was in the box. No packaging, no instructions. I can live without
the latter, but clearly the components didn't all live without the former. The broken part
is a simple 6.3 mm to 3.5 mm adapter, but what if the other components are also defective?
Without the adapter I can't know.
I've signed up for a Coursera course,
Dinosaur Paleobiology. When
did it start? 3 January, it claims. But there's no way to access the course. Where other
courses have a button “Go to class”, this one has a button “Started 2 days ago”, and it
doesn't link anywhere:
Fought my way through the Coursera help pages and found a way to submit a problem report.
To my surprise, I got a response in only two minutes. Less to my surprise, it was “clear
all cookies, install latest browser, log out, try again”. And of course that didn't work.
Another message, another almost immediate reply:
Although the course is due to start 2 days ago, it's still inaccessible now. This is a
known issue and we're working on it.
The first sentence is a repeat of my error message. And the second begs the question why he
required me to mess around with browser settings.
Synchronizing files across the net is not necessarily a sequential matter. If I'm
synchronizing a whole directory over a slow net, how do I know whether a specific file is
full or not? It's not that easy: you can find out how many data blocks are allocated, and
you can compare that with the EOF. But even if the blocks are allocated, it doesn't mean
that they contain data all the way until the EOF. And because they're blocks and not
individual data, they normally extend beyond the EOF. Spent some time playing around, and
came up with this program, which is still very flaky and
utterly dependent on UFS. But
it helps.
It'll be months before the new house is complete, but it'll also take me months to throw out
all the old stuff that I really don't want to throw out. Today made another attempt, and
filled a washing basket with about a dozen old (but mainly functional) disk drives and a
whole lot of ancient documentation. It's like pulling teeth.
To add insult to injury, I dropped the washing basket in the skip and couldn't get it out
again: it was too far down, and the skip is so high that I can't conveniently climb in and
back out again. Finally managed it with the aid of a bamboo pole.
Fifty years ago I had the crazy idea of fasting
for Ramadan, which started on
4 January 1965. I'm not a Muslim, so why should I bother? I suppose
it was for the experience. I only kept it up for a couple of days.
Now I have very different reasons for not eating: I'm getting old, and I don't need as much
food as I used to. Today we tried getting by with a substantial late breakfast and a
substantial evening meal, missing out on lunch. Still no go, and I don't think we'll repeat
it in the near future.
When I first started using computers at home, it was normal that it was difficult. In the
course of time, though, things have generally got better. 16
years ago I gave a whole day tutorial on installing FreeBSD at the AUUG 1998 conference. I
didn't finish.
In the meantime, things have got easier. I can do the same installation in about 30
minutes. But what do I have? A base system. How do I configure it? How do I install
useful programs (ports)? How do I keep it up to date?
Commercial systems like Microsoft and Apple have solved that problem long ago. So, to a
greater or lesser extent, have most Linux distros. But somehow FreeBSD is lagging behind,
and there's a good reason why I still haven't updated eureka nearly a year after
buying new hardware. But finally it's time, and I started by taking an image of the root
file system of lagoon, Yvonne's computer, which in my
configuration includes /usr, and putting it onto
the SSD that will become the
new root for eureka. The build, in a test machine, worked as well as ever. What
will the ports bring?
Craig appeared on site today to dig the holes for the footings for the shed. That's all he
can do this week; we need the building inspector to inspect the footings, and he won't be
back until next week. In the meantime did my own inspection and discovered that the holes
were in the wrong position, so Craig will have to pick one up and move it 7 m south.
Also talking to Sean and Craig about the slashing that needs to be done. Craig tells me
that they can't issue a request to slash grass if it's being used as fodder, so we may get
by without having to do anything.
Borzois
are sight hounds, right? Right. But
sometimes I doubt it. When walking the dogs, we frequently see rabbits. So does Leonid, but when he runs after them, they frequently go
off in a completely different direction, he doesn't notice, and they escape.
Today was slightly different: he found a second one to follow, and it went into the bracken,
where I suspect the dogs have an advantage, but he still didn't manage to catch it. I let
Nikolai off the leash, and he shot in, and in
no time they had it. After all, Borzois are also supposed to hunt in packs, even if their
intended prey is considerably bigger than a
rabbit.
But who gets the rabbit? The one with the stronger teeth?
On the face of it, Nikolai (left) is winning. But equally clearly, the rabbit is
losing. So I disengaged Leonid, leaving him with only some fluff in his mouth:
And then, of course, it was gone. But not the way we expected: he seems to have hidden it.
Ten hours later, in the evening, we found him still playing with it:
OK, he had his chance. Took him out of the dog run and put Leonid there instead. Came back
30 minutes later to find him still playing with it. Took him out and put Zhivago in instead. Gone, this time for good. And when
we put the others out later, they spent lots of time looking for it. When will they learn?
The photos of the dogs this morning were taken with my “standard” settings:
aperture-priority f/8, ISO 24°/200. So the exposures were round 1/60s, and there's a fair
amount of motion blur as a result.
I've already taken to using the 14-150mm lens
for walks. It's not the best lens I have, but it offers great flexibility, and it's far
more than adequate for this kind of work. So why don't I allow auto-ISO sensitivity?
Probably I should. Then the next question is: how high can I go? The recommended maximum
is round 33°/1600 ISO, but if I only use auto to be flexible, maybe it makes more sense to
put the upper limit at 36°/3200 ISO, like I've done for Yvonne's camera.
Jim Lannen and apprentice along this morning to look at some odd jobs to be done on the
Kleins Road house. I nearly didn't hear them: the battery in the doorbell was flat. They
did their jobs without event, but it's clear that we need a better doorbell solution in
Stones Road (and not just the inevitable problem that most Australians ignore doorbells and
just knock). So they put in a cable for a real cabled doorbell (or chime, as they seem to
be called). The button will be on the right, where hopefully it will be very visible:
Duncan was there, and curious about the power points for the studio flash units that I'm
putting under the ceiling of the lounge room. He was particularly interested in the one in
the kitchen area, because it didn't have any cable going to it. Oops. Hopefully that will
be the last one,but it's done now, along with the cables to the shed:
I wonder if there's something else that we've forgotten.
The house is now approaching lock-up stage, and there are still discussions about the front
door. As Duncan showed me, the contract contradicts itself. Item 11-3 states “Provide
BFR13 Front Entry Door in lieu of XV10”, a sum of $182, and the colour selection states XS3.
Should I have noticed that? Just finding out what these codes mean is serious work. In any
case, the current situation is unacceptable: the door they delivered is designed to be
painted, not what we wanted, the side panel (not mentioned elsewhere in the contract) has
been supplied with non BAL 19-conformant glass. More negotiations needed, but Duncan
confirmed that Tom Tyler is on holiday until next week, so I have a little time.
One issue with the lock-up stage is that yes, it's really lock-up, and I don't get a key.
From now until we move in, we only get in for inspections with Duncan. That will be fun.
Later in the afternoon, Warwick Pitcher came along to spread the sand base for the water
tanks. A good thing, too. It seems that the top of the tanks needs to be at least 50 cm
lower than the gutters to avoid flooding with heavy rain, so they had to dig them 50 cm into
the ground:
It seems that they have now decided that building began on 19 December—no idea why—so
presumably they're committed to getting it finished by 27 March. That's roughly when we're
hoping to be able to move in to Stones Road. It'll be interesting to watch.
For some reason, Olympus lenses come with a plethora of different lens caps. Including the 15 mm Body Cap
Lens, I have 6 different kinds. The one on the 14-150mm lens, though identical in construction to the one that fits both the 14-42mm and
the 17mm lenses, is a real problem. I have had it fall off and get lost at least 5 times, usually in
a forest somewhere. There must be a better solution. I've thought of tying it on to the
lens, but that's a nuisance most of the time. Why don't they make cheap, throwaway plastic
push-on caps any more?
Yvonne has different problems. She's still having trouble
accepting her new camera. One of the issues is definitely lens caps, and you can understand
that when she's on horseback and doesn't have time to fiddle around with them. That's why
she likes the body cap lens. Just about all compact cameras have integrated lens
protection. Why isn't anything like that available for higher quality cameras, maybe
fitting into the filter thread?
As I feared, updating my packages wasn't all plain sailing. In fact, it was almost
impossible. Started on my stable machine (that's the name) with the disk cloned
from lagoon. Away it ran, then:
The process will require 210 MB more space.
2 GB to be downloaded.
Proceed with this action? [y/N]: y Fetching zziplib-0.13.62_2.txz... done
Fetching zsh-5.0.7_1.txz... done
pkg: cached package zsh-5.0.7_1: size mismatch, fetching from remote
Fetching zsh-5.0.7_1.txz... done
pkg: cached package zsh-5.0.7_1: size mismatch, cannot continue
What went wrong there? The “cached package” information is not very relevant, since the
package had just been downloaded from the net, twice. And because of this single error, the
whole update of 474 packages was aborted.
That proved not to be as big a problem as it might seem: after deleting the zsh
package, the next one failed in exactly the same manner. So did pkg install
of zsh. Peter Jeremy suggested pkg clean, but of course that didn't make any
difference.
After some head-scratching, wondered if it had something to do with cloning the
disk. lagoon clearly needed the same updates, so it was worth trying it there. Sure
enough, it started off quite well, until I got:
Checking integrity... done (20 conflicting)
pkg: Cannot solve problem using SAT solver:
conflict rule: The following packages conflict with each other: py27-imaging-1.1.7_5(r), py27-pillow-2.6.0_1(r)
conflict rule: The following packages conflict with each other: py27-imaging-1.1.7_5(r), py27-imaging-1.1.7_5(r)
upgrade rule: upgrade local py27-imaging-1.1.7_3 to remote py27-imaging-1.1.7_5
dependency rule: package py27-imaging(r) depends on: py27-imaging(l)xbmc(l)
cannot install package py27-imaging, remove it from request? [Y/n]: pkg: cannot find py27-imaging in the request
pkg: cannot solve job using SAT solver
Checking integrity... done (0 conflicting)
Conflicts with the existing packages have been found.
One more solver iteration is needed to resolve them.
How did that happen? These are clearly dependent packages of
some Python port.
I didn't install them: pkg did. So how did these conflicts arise? Still, I don't
need them now, and I can always reinstall if I do, so I could remove them. Or could I?
=== root@lagoon (/dev/pts/2) /eureka/home/grog 19 -> pkg delete py27-imaging-1.1.7_5, py27-pillow-2.6.0_1 Checking integrity... done (0 conflicting)
Package(s) not found!
=== root@lagoon (/dev/pts/2) /eureka/home/grog 20 -> pkg delete py27-imaging-1.1.7_5 Checking integrity... done (0 conflicting)
Package(s) not found!
=== root@lagoon (/dev/pts/2) /eureka/home/grog 21 -> pkg delete py27-imaging Checking integrity... done (0 conflicting)
Deinstallation has been requested for the following 2 packages (of 0 packages in the universe):
Installed packages to be REMOVED:
py27-imaging-1.1.7_3
xbmc-13.2_4
The operation will free 73 MB.
Proceed with deinstalling packages? [y/N]: y
[1/2] Deinstalling xbmc-13.2_4...
[1/2] Deleting files for xbmc-13.2_4: 100%
[2/2] Deinstalling py27-imaging-1.1.7_3...
[2/2] Deleting files for py27-imaging-1.1.7_3: 100%
In other words, pkg doesn't understand the package names that it outputs. This isn't
new, but it's frustrating, and in cases of duplicates it's inaccurate.
Duplicates? What duplicates? These duplicates:
=== root@lagoon (/dev/pts/2) ~ 23 -> mailme pkg upgrade ...
Processing candidates (769 candidates)... done
Checking integrity... done (19 conflicting)
pkg: Cannot solve problem using SAT solver:
dependency rule: package linux_base-f10(r) depends on: linux_base-f10(l)linux-f10-sqlite3(l)
(many more) conflict rule: The following packages conflict with each other: linux_base-f10-10_9(r), linux_base-c6-6.6_1(r)
conflict rule: The following packages conflict with each other: linux_base-f10-10_9(r), linux_base-f10-10_9(r)
upgrade rule: upgrade local linux_base-f10-10_7 to remote linux_base-f10-10_9
dependency rule: package linux_base-f10(r) depends on: linux_base-f10(l)nspluginwrapper(l)
dependency rule: package linux_base-f10(r) depends on: linux_base-f10(l)nvidia-driver-304(l)
cannot install package linux_base-f10, remove it from request? [Y/n]:
What do I do now? lagoon is Yvonne's machine, and I
don't want to break Linux compatibility on it. So for the moment I stopped, after only 2
hours and nothing achieved but a little experience.
What a pain! So far we appear to have:
Can't clone a disk with pkg data on it.
Inadequate error checking after loading a package from the web. I suspect that it
didn't make it there at all, but at the very least it could report the expected and
actual sizes, like the old Ports Collection stuff did.
pkg somehow manages to install conflicting packages.
pkg doesn't (apparently) offer a way to specify exactly which package you're
talking about.
Some of these points may be incorrect, but that's the way things look at the moment. So
once again the update is on hold.
Over to the Stones Road site this morning to take a photo that I forgot yesterday. There
was less going on, only the bricklayers. And Craig's ute, the one with the water, was
gone. Brendan, the brickie, confirmed that he hadn't used the water. Why was it gone?
Craig is in Ballarat in the daytime, so
back home to fill up another tank. While there discovered that Craig was complaining
because Michael, one of Warwick's men, had moved the ute (it was in the way of an entrance
to the paddock, and locked) with
his Bobcat, and that Craig was
quite upset about the fact. Later we heard that the gearbox was damaged—I can't see how
that could have happened—and that the chassis was twisted. Looks like there will be some
fixing to do.
Yvonne uses a smart phone. That in itself isn't an issue,
but she has real problems using it. I have a tablet with phone function, along
with Bluetooth headphones, but I use it
so seldom that I still have issues. In particular, the user interface is really bad.
I've been using touch screens for years now, and I still find them one of the most obnoxious
developments I've had the displeasure to experience. Answer a phone? Not so long ago you
used to pull it out of your pocket, feel for the “Talk” button, and press it. Now you pull
it out of your pocket, look at it to locate the green flashing area on the screen, and swipe
it in a specific direction, which you have learnt by trial and error (why?). Everything
seems to be designed to be “cool”, rather than functional. Not for the first time, I wonder
how these interfaces will develop over time. Desktop computer interfaces such as Microsoft
and Apple have improved marginally, so there is hope for portable devices.
In the meantime, ALDI has a “back to school”
special for the kids who thought that the school holidays would last another 3 weeks: a dumb
phone (well, not that dumb; it does have some network functionality):
My guess is that no schoolchild would be seen dead with something so old-fashioned. But it
has keys, though admittedly it acknowledges modernity by not having a labelled on-off or
talk button. And possibly the user interface isn't as bad as the smart phones I have seen.
In particular, it costs $18 and can be returned if I don't like it, so I got Yvonne to buy
me one, and we'll try it out.
First problem: it was dead. Put SIM card and battery into the device, connect charger.
Nothing. Something wrong with the battery? Looks OK to me:
That's in a particularly dark part of the house; even with the lights on, it was difficult
to recognize the distinction between the cat and the background. Clearly a case for flash,
but I was concerned that he would walk away. So I cranked up the sensitivity to 45°/25000
ISO and took it hand-held, still 1/10s. At the standard 24°/200 ISO it would have needed
12 seconds.
And the quality? Acceptable or unacceptable, depending on your standards. This was done
with DxO Optics
“Pro”'s “DeepPRIME” noise reduction.
This was reprocessed 10 years later, mainly to address white balance; at the time I
used “PRIME”, which was the best they had. There's not much difference to be seen.
Without it looks like the first image. Run the cursor over either
image to compare with the partner:
On the whole, I don't see a big improvement, though of course the highlighting makes for
part of the problem. And the colours are really not good. Although white balance has been
set to the minimum of 2000°K, it's all too yellow. Shadow doesn't really look like that.
In passing, it's interesting to note that the lack of a built-in flash is a real problem for
photos such as this.
It's been over 3 years since we
last ate cannelloni, and we thought it would be
time to do it again. Things went relatively smoothly until we discovered we had forgotten
to add the chicken livers to the stuffing. And I forgot to read to the end of the recipe,
where I wrote:
In particular I had a bad experience with
San Remo. It's too early to say where
the problem lies: their other products are of good quality, but the tubes didn't cook
properly, and the results tasted like glue.
They didn't taste like glue this time. It's hard to describe, but they were coarse and
grainy. Next time, if we make it again, we'll have to make fresh pasta.
Brendan has now finished with the brickwork, but he'll come back next week, or maybe the
week after, to wash down the brickwork. And for that he'll need another 1 m³ of water.
Not a problem, we have water tanks now, and the weather forecast suggests we'll get round 50
mm of rain over the weekend. That's enough to fill the tanks to about 40% (about 20 m³).
And Mari Hendriks was supposed to connect them up today.
Problem: it's raining. He can't glue his pipes while it's raining. So all that water goes
to waste, and the water is on the wrong side of the tanks:
What does that imply? Does Google know? Of course. This
page discusses the matter, and “SirDice” comes up with a suggestion:
pkg clean # cleans /var/cache/pkg/
rm -rf /var/cache/pkg/* # just remove it all
pkg update -f # forces update of repository catalog
rm /var/db/pkg/repo-*.sqlite # removes all remote repository catalogs
pkg bootstrap -f # forces reinstall of pkg
He also adds the important warning:
You can delete quite a lot but whatever you do, do NOT
delete /var/db/pkg/local.sqlite. It contains the database with your installed
packages. If you remove it the system will think you have nothing installed.
I've already run pkg clean, and it was fairly clear that the issue wasn't with the
contents of /var/cache/pkg, since new contents get loaded. And pkg update -f?
Yes! After that, I only had the pain that I had already experienced on lagoon. But
of course I still got the issues that I had already seen with py-imaging. In the
meantime, though, Callum Gibson and Peter Jeremy had pointed out that my problems deleting
with the exact package name were of my making. The name wasn't exact:
conflict rule: The following packages conflict with each other: py27-imaging-1.1.7_5(r), py27-pillow-2.6.0_1(r)
conflict rule: The following packages conflict with each other: py27-imaging-1.1.7_5(r), py27-imaging-1.1.7_5(r)
upgrade rule: upgrade local py27-imaging-1.1.7_3 to remote py27-imaging-1.1.7_5
...
=== root@lagoon (/dev/pts/2) /eureka/home/grog 20 -> pkg delete py27-imaging-1.1.7_5 Checking integrity... done (0 conflicting)
Package(s) not found!
Of course the package wasn't found. That was the one pkg was trying to install, not
the one I already had installed (py27-imaging-1.1.7_3). Using that name, it worked.
So that's one objection less.
After that, it ran until the same place where I stopped on lagoon, with Linux
compatibility. I didn't want to break lagoon, because it's Yvonne's machine, but I can risk it on stable:
=== root@lagoon (/dev/pts/2) /eureka/home/grog 21 -> mailme pkg delete linux_base-f10 Installed packages to be REMOVED:
Lots of Linux stuff, of course, and ... acroread9-9.5.5
nspluginwrapper-1.4.4_4
nvidia-driver-304-304.88_2
acroreadwrapper-0.0.20130208
Why nvidia-driver? That's FreeBSD-specific. Still, noted the packages so that I could reinstall them later.
After that, yet another error:
pkg: Cannot solve problem using SAT solver:
conflict rule: The following packages conflict with each other: gcc48-4.8.4.s20141218(r), gcc-4.8.3_2(r)
conflict rule: The following packages conflict with each other: gcc48-4.8.4.s20141218(r), gcc48-4.8.4.s20141218(r)
upgrade rule: upgrade local gcc48-4.8.3.s20140424 to remote gcc48-4.8.4.s20141218
cannot install package gcc48, remove it from request? [Y/n]: Enter pkg: cannot find gcc48 in the request
pkg: cannot solve job using SAT solver
Another package to delete. Looking back, I wonder if the “SAT solver” ever solves a problem
or a job.
And after that pkg upgrade ran! How many packages have I lost? gcc isn't an
issue, because it's always a dependency. But what about the collateral damage removing the
Linux emulation? nvidia-driver was fine, but I couldn't
find acroread, nsplugin-wrapper or flashplugin:
The good news today: it didn't rain nearly as much as forecast. But yesterday's rain was
enough. Not only was it on the wrong side of the water tanks, they were floating in it:
At some time in the past I must have joined the FreeBSDforum site, though it's not
really my thing. But what's the password? No worries, I can get it reset. But first I had
to go through a CAPTCHA like I've never
seen before:
It looks as if Craig is so upset about what Michael did to his ute that he won't even
communicate with us, and Leah didn't come to clean the house today either. That's sad. And
it means that I have to do more work myself, which is even sadder.
Out today to the shed to throw away some stuff. I'm getting better: threw away a whole lot
of old floppy disks, mainly with old Microsoft-based software. But then there are other
things I can't throw away, like my
home-made Z-80-based
CP/M machine, old listings, and a number of
ancient copies of AUUGN. And, of course, various old vague and strange hardware. Must I really throw it all
out? Doesn't anybody want this stuff? 25 years' worth of c't, 20 years' worth
of Byte, lots of other stuff?
Ah, the pain!
The dumb phone I bought last week was so interesting that Yvonne wanted one too. And it's complicated enough to require an instruction manual. But paper
is expensive, so they printed the A4 original in a format of 10×7 cm, roughly 3 pt, and
almost illegible without a magnifying glass. Called up Tempo, who claim to supply service, on 1300 886 649. 20 minute wait, during which I
was continually asked to send email to them. I've been there before: they never answer.
Finally connected, spoke with Ben, who promised to send me the document. And so he did, but
despite clear enunciation of the email address, he got it wrong, and it bounced. All that
work for nothing. Still, maybe this is a case where the info@tempo.org
(why .org, incidentally?) might work. Sent them mail. So far no response.
But where are the cables for the ceiling lights? Gone!
Called up Jim Lannen and discovered that yes, this is OK. Apparently they bring them
through from above, presumably to be able to place them more accurately.
Finally got round to connecting up the electric fence actuator that I repaired last month. As I feared, it's still
not working effectively: only 2.1 kV unconnected, or 1.3 kV connected. It seems that it's
just not discharging the capacitor properly. That's enough effort wasted on this ancient
device. Time for a new one.
A call on the phone today, ostensibly from a Fiona of “Collection House”, maybe this company, regarding a debt to Telstra. I asked her to authenticate herself. “How
can I do that?”. Indeed. That's her problem. She asked me to call back, but why should I
do that? As I've said before, if Telstra wants money from me, they should send me an invoice. End of discussion.
Only later did it occur to me that once again I didn't get enough information. Last time I
didn't note the name of the company, nor the amount they are asking for. This time I got
the name, but not the amount. If they're scammers, there's a good chance that one or the
other, maybe both, would be different. In any case, why can't they just send an invoice?
This somewhat dated report suggests that Telstra sells its debt to the agencies. Can they do
that without ever having issued an invoice?
Once again to Stones Road, mainly to walk the dogs and put Chewey back in the horse
paddock. But the plasterers have left signs of having been there, including not shutting
the doors. And just under half of the internal walls are installed:
The first image shows something of interest: no cables for the power points or network
connections. My guess is that this is the way they do it. In each case the other side of
the wall hadn't been installed yet, and that looks like this:
Heard on the radio this morning:
something along the lines of “I have received a text about the background
of Beethoven's piano sonatas”. Only in
the later discussion did it eventuate, by implication, that it wasn't a text at all: it was
a short message. More
bad language? It was all the more
ambiguous because a few minutes later he referred to the text
that Robert Schumann had written
at the top of some work or another—clearly this case a real text.
Is this becoming mainstream? My aunt Frieda is turning 90 next weekend, and today we
received an invitation to a
party. RSVP was “Text to
mobile Gillian Kline ...”.
More language evolution. OED has it as a draft
addition dating as far back as March 2004, with quotations going back to 1998. I must have
missed something.
Mari Hendriks over to do the downpipes this morning. It wasn't a good day for it: we had 5
mm of rain, and Mari was correspondingly happy.
While there, took a look at the plastering. Oh horror! They had plastered over all
the power points. Spoke to Luke, the foreman, who said “That's the way we always do it.
The electricians will know where the cables are”.
By that time it was too late. Called Jim, who produced some colourful expressions, and
Duncan, who told me “some do, some don't”. I had met up with a plasterer some time last
month, and he was clearly one who did, so I didn't think any more about it. Jim thinks he
can recover, but it would really have been trivial to drill a hole through the plasterboard
in the appropriate position.
Over to the site yet again later on: Dave Tudor, the shed man, finally showed up, and it
looks as if he will be able to start next week. That would be a load off my mind.
I've bought a new GPS
navigator, not because the old one is more defective than it was when I bought it, but
because the battery is as good as dead, and the maps are out of date.
I had looked at some expensive models three months ago, and come to the
conclusion that they weren't worth the additional price. So in the end I bought an
el-cheapo one, 7" for only $69.96 (isn't that an inaccurate way of saying $70?). I
couldn't replace the battery and the maps for that.
It arrived today. Yes, it uses iGO software, with which I'm familiar. But they have a solution for that: complete overhaul
not only of the GUI, but of the terminology as well. Still, the manual should help.
Yes, it should. But the 56 page manual I received with it contains almost no information
about the navigation software, though in passing it's interesting to note a screen shot of
the main screen of the old navigator, which looks nothing like what this one presents. Most
of the manual describes the tablet-like functions which the thing also has, and which are
completely irrelevant to a GPS navigator. The navigation software gets a little under a
page of useless commentary, the most important of which is:
Please refer to the Navigation Software Instruction.
And where's that? Went searching, and of course iGO's web site was down, but somewhere I found a
manual online.
Have things got even worse? I have a recollection that this is fairly typical for navigator
instruction manuals. Why do they do it?
During walks, Leonid has been surprisingly
good at returning from whatever he was doing whenever Yvonne calls him. But now, as he's coming up to his first birthday, that obedience seems to be
waning. Today he wandered off into
the Dereel Lagoon and paid no attention at
all when Yvonne called him. Clearly time to keep him on the leash more.
But before we could do much about it, we saw him chasing a group of kangaroos. Bad idea.
Catch a kangaroo and it will slit open your stomach. And then a yelp. Clearly time to
help. But how? I held Nikolai and
Zhivago on the line, while Yvonne went off to
recover him.
I couldn't follow because the other two dogs would have gone crazy. But there's a signpost
right where I was:
Just the thing to tie the dogs to while I went to help Yvonne. But while I was doing it,
Nikolai saw something and dashed off, pulling my hand across the bottom of the sign and
cutting a deep gash:
No hope of helping Yvonne. Enough to help myself. Back home, where Chris Bahlo was, and
she helped me put a provisional bandage on it. Then off to
the Ballarat Base Hospital
Emergency Department, labeled “Triage”, a word I didn't know in English, but which seems to
have a certain currency in medical circles. I had to ask to confirm that this was, indeed,
the Emergency Department. But then, that's relatively typical of things
in Victoria. Even the
Base Hospital web site is titled
“Ballarat Health Services”, and though it clearly refers to the Base Hospital, the only
current explicit reference is to the car park, and even that reference is truncated.
Long wait, nearly an hour—I hope that in cases of life or death they're faster—and then a
Dr. Nigel (apparently no surname), who took a look and went into amazing detail about what
was happening, what to do, and what would happen later. He first injected 6 ml of
anaesthetic (a sodium channel
blocker, he says), which made my finger swell up alarmingly, and then tidied the whole
thing up and stitched one of the wounds up (3 stitches):
It seems that they throw the instruments away after a single use, so he offered them to me,
with the exception of the needle with which he made the stitches. They could come in handy:
The second image of the leash clearly shows where the plastic has been scraped off.
And Leonid? Yes, he caught his kangaroo, a juvenile only slightly larger than he. He
didn't quite know what to do with it, and just jumped at it every time he moved. When
Yvonne arrived, the kangaroo escaped. And Leonid wasn't injured; no idea what the yelp was
about.
Those power cables in the first image should be 2 m long. And to the right of the door (the
one that goes back) in the second image is where the main switchboard should be. I hope Jim
can find that.
That was taken with a Nikon
“Coolpix” 880, with 3 only megapixels, not really enough for him. But I
still have the scope (and it's up for grabs!). Now I have a decent camera, so why not take
new photos?
Today should have been the day where I take my monthly flower photos, but I didn't feel up
to it. Still, there area couple of flowers of interest.
The Strelitzia reginae has
only a single flower this year, and even that doesn't look very happy:
I didn't think to bring my camera to the Emergency Room yesterday, but of course I had my
Android tablet, which I use as a glorified eBook reader. And it has a camera, so I used it to take
some photos of my hand. Today I moved them to eureka for processing. This is a pure
Linux base, right, no Microsoft in sight. So why do I get this?
-r----x--x 1 grog lemis 1,279,950 15 Jan 11:33 IMG_20150115_113355.jpg
-r----x--x 1 grog lemis 1,194,553 15 Jan 11:34 IMG_20150115_113405.jpg
It's no longer clear what I meant with this comment, but I suspect that it's the use of
the .jpg extension to the file name, where .jpeg would be more correct. Sadly,
that train has left.
Apart from that, of course, the photos confirm my prejudices against smart phone and tablet
cameras. But at least I got some images, so for an emergency they're adequate.
For some time now I've had a sign on my letter box “Please do not fold envelopes”. It's
mainly intended for postmen who like to fold A4 sized envelopes or magazines. But yesterday
I received perfectly normal envelopes which had been scrunched up:
I took the photos of the envelopes with my mecablitz 58-AF-2. Like with all on-camera flashes, I've had trouble with the
exposure. This time I put my Sekonic L-308S next to it
and measured:
f/4.0.5 is a silly way of saying f/4.75, in other words intermediate between f/4 and
f/5.6. But the camera was set at f/8, and I had +2 EV flash compensation. Why are these
values so out of line?
That looks like a scratch next to his eye. It's slightly inflamed, but it doesn't look as
if any serious harm has been done. But he's very lucky; just a couple of millimetres
difference and he could have lost an eye.
It's the middle of summer again, time for the monthly flower photos. Things are gradually winding
down here—I don't see more than 2 or 3 more series before we move house. And we're not
paying as much attention to the garden as we did in previous years. Still, there are some
surprises:
That's the Mandevilla laxa that I
planted 3½ years ago, and
which is flowering for the first time. Why so long? I also planted some Jasminum polyanthum in the same
place, and though that has flowered profusely in other parts of the garden, it's still not
flowering here. Either the soil is bad, or it's not getting enough water. In any case, the
Mandevilla is pretty and looks and smells like a
small Frangipani, so we'll find a place
for it in the new house.
And then there's this plant, which seems to be a volunteer:
And our Gladioli are flowering, but not
standing upright. It looks as if the bulbs grow towards the ground, and then offer
insufficient grip to hold the stems up.
Seven years ago I took some photos of a large group of
Sulphur-crested
cockatoos. In those days I was just getting back into photography, and the lens I used
was an old, manual focus Tamron 80-210 mm lens. Partially for this kind of photos I later
bought my Zuiko Digital ED 70-300mm F4.0-5.6 long telephoto lens, and of course since then I've
changed camera twice.
Today we had another similar swarm, conveniently while I was taking the flower photos.
Changed to the 70-300 mm and tried again. Are the photos better? One of the issues was
flare from the sun, and by the time I had gone back and got a lens hood, most of the
activity had ceased. And focusing on individual birds is really difficult, even at
relatively short focal lengths. In the end I didn't get any good shots of the birds
swarming, though these ones differ significantly from 7 years ago:
My Internet link has an uplink speed of 5 Mb/s, so theoretically I should be able to upload
data at about 600 kB/s. In practice, window size and latency limit it to about 125 kB/s.
Yes, of course I could increase the window size, but that requires rebooting, and I'm an
uptime fanatic. So I put up with “only” 125 kB/s.
But today I got much worse transfer rates, over an extended period of time:
House photo day today. It was really yesterday, but it has been quite windy both days, and
that's murder for my HDR images. On the other hand, it was also overcast, so I didn't have the extreme
contrast range typical of bright sunshine. And I always used to make these photos without
HDR. So how about doing it again, partially making up for the brightness differences with
automatic exposure for each image?
The results were acceptable. Here last week's and this week's view of the verandah (run the cursor over an image to compare it with its neighbour):
It's not until you look at the extreme left corner that you see the difference. But that's
after a couple of rounds of tidying up the image. Without that, things look considerably
worse (second image, again run the cursor over an image to compare it
with its neighbour):
Our water tanks have had the water on the wrong side almost since they were put in place.
Tomorrow Mari Hendriks is coming to connect them up, so over with Chris and her fire pump to
pump the water out:
It's a good thing we took Chris' pump. We have an electric one as well, but it doesn't pump
more than a fraction of that amount of water. We also have a petrol-driven one, in fact
quite similar to Chris' model, but we got it with the house nearly 8 years ago, and we've
never used it. Time to have it looked at.
Talk with Mari Hendriks about the water tanks today. Yes, we need the trench completely
drained, so back over there with Yvonne to pump it out.
After some confusion (Chris had put the hoses on the wrong way round), got the thing going
and got a surprising amount of water out:
Into town to see the doctor for a number of reasons:
a Cobalamin injection, inspect my
wounded hand (all coming up well) and get the results of last month's blood test.
On the way home, stopped off at a roadside plant sale just south
of Napoleons. It's a little
early for plants for the new house, but I took a fancy to
this Fuchsia:
Mari Hendriks really wanted to do some work on our site today, but it's not clear that it's
a good idea to move the tanks manually. At least the remaining water in the trench has
drained away:
He came along anyway to peg out the septic tank, which he will hopefully install next
Friday. It's a very different layout from what we wanted, but it seems that it's the only
way, given the lack of gradient on the property. On the other hand, he didn't want to use
what gradient there was. I'm confused.
Tanya loved carrying twigs and small branches
around, so we dubbed her “branch manager”. When Leonid came we thought that he was going to follow in her footsteps. But for some
time he hasn't shown much interest—until today:
Last year I got the impression that Powercor's reliability was improving. You certainly wouldn't get that impression this year. This
morning we had yet another power failure, at 2:01. That's the 6th power
failure in 21 days, an average of 3½ days per outage, or roughly 104 per year. Of course,
it could be related to the same still-unfixed fault.
Into town early this morning for another fasting blood test. The regular pathologist (Pam,
I think her name was) wasn't there, and it seems that her replacement is completely
confused. After waiting 1½ hours and getting increasingly hungry and annoyed, I finally
asked what was going on. She said “number 6? You're on now”. But by that time I was so
hungry and angry that I asked her to get confirmation from the doctor that it would still
make sense to take a blood test. No, not under those circumstances. While there, noted
that she had already processed numbers 7 and 9—no idea if 8 also suffered my fate.
All were apologetic, of course, and have promised to allow me immediate access next time.
But what a waste of a trip to Ballarat!
Still no sign of Dave Tudor. I wonder if I'll hear from him again. And it seems that Norm
Baker forgot part of the slab, a rebate for the sliding door. Not the only thing that he
didn't do.
I've been bitching and moaning about GPS navigators since I first got one, and somehow the
problem isn't going away. When I went
into Ballarat today, I took my old one
and my new one with me. I still can't work out how to do some simple things with the new
one, like muting the sound.
But then the display on the new navigator went crazy, like an old analogue TV with bad
horizontal sync. And I couldn't turn it off. In the end, I had to reset it. OK, I still
have the old navigator. But it wouldn't charge. Are the batteries that bad?
In any case, the new navigator went from bad to worse. By the time I got home, it wouldn't
power on at all. Time for a return? Return postage costs about a quarter of the price I
paid for it. Still, I can recover the other ¾. What a pain!
Looking once again for navigator apps for my Android tablet. And once
again there was little to be had. Why is this all so difficult?
The Olympus OM-D E-M1 is a lot smaller than the old DSLRs that
it replaces, such as the E-30. For Yvonne it's still far too big, but a number
of more serious photographers think it's too small.
And that's a full
frame SLR design,
nearly 60 years old. Older cameras were even smaller, such as
the LeicaIIIf. Here dimensions in mm:
Height
Width
Depth
Leica IIIf
70
135
30
E-M1
93.5
130.4
63.1 (34)
E-30
107.5
141.5
75
The E-M1 has a typical modern bulge on the right; that's the 63.1 mm. The rest of the body
is “only” 34 mm thick, 4 mm more than the Leica. The real question is why modern DSLRs have
become so big.
Our dogs can all catch rabbits, though Leonid is still learning.
But after my accident last week we've been keeping him on the leash, and let Zhivago run free.
There's a patch just south of the “eagle's nest” where there are a lot of rabbits, and the
dogs are always very interested. Today Zhivago walked through the bracken, dropped his head
and killed a rabbit—all so fast and with so little effort that we didn't notice at first.
And he also dispatched it with amazing speed:
Why? Normally he gets a fresh “chicken frame” (the rib cage and other bones) every evening,
so he's used to eating and swallowing bones, but he chews them properly. He must have been
in a hurry when he caught it.
That's about the stage we were at a little over 2 months ago. I
wonder if they will catch up. Certainly nothing much has happened on our site this week.
Somewhat garbled call (Thanks, Wendy) from Jason Bruty this morning, so over to discover
that we were short 10 sheets of wall sheeting.
How do we explain that to Widespansix months after delivery? Jason had
that answer too: they're missing from the Bill of Materials, so the delivery was correct.
He even arranged with Widespan to have the sheets delivered. Clearly made the right choice
to get him to do the work; I wonder what Dave Tudor (who still hasn't shown any further sign
of life) would have done. The only problem is that this weekend is the
long Australia Day weekend, so
there's a good chance that it'll take a little longer to complete the shed.
So why did both of my GPS navigators die at the same time on Wednesday? Conventional wisdom has
it that the problem is probably elsewhere. Did some investigation and discovered that both
still worked—in my office. It seems that the USB cable (with the standard mini B connector)
was not making a good connection. That's not the first time I've had this problem:
four years ago I
destroyed a GPS navigator with a flaky connector, and I've had continual problems charging
my Android tablet. So I found a connector that made better contact and was able to resuscitate both of
them.
Tomorrow my aunt Frieda is celebrating her 90th
birthday—in Briagolong, only about 400
km away. How do we get there? How long will it take? That's what Google Maps and GPS navigators are for, of course. But
which do I choose? I have two potential ways to get there,
via Geelong,
via Ballarat. Then the navigators offer
four different profiles: “fast”, “short”, “easy” and “economical” (or in modern English,
“green”). Google prefers the route via Geelong, and the navigators prefer the route via
Ballarat. And here's what they come up with:
Software
via
Distance (fast)
Time (fast)
Distance (short)
Time (short)
Google maps
Ballarat
395
4:36
Geelong
391
4:36
New navi
Ballarat
395
5:04
362
8:31
Geelong
392
5:09
367
7:13
Old navi
Ballarat
395
5:10
378
5:46
Geelong
390
5:09
382
5:20
Which is right? In particular, the “short” times are ridiculous, and why are they so
different between the old and the new navigator? But I've learnt never to trust the values
anyway. From a more practical viewpoint, it's clear that we can't average more than 100
km/h, so the question is really whether it's 4 hours or 4½ hours.
Some months ago I made a
filet de bœuf sous vide, and it tasted very good. “Rump” (for some definition
of rump) didn't. So today we tried a
filet again. And, sadly, it wasn't as good. Definitely not bad, but not in the same class
as the last one. I cooked them both exactly the same way, so it looks as if the success
last time was at least partially due to the quality of the meat.
So what time should we have left
for Briagolong this morning? We had
to be there at noon, so based on yesterday's calculations, some time between 7:30 and 8:00. In the end, we left at
7:30, which turned out to be a good choice.
Coming across the West Gate
Bridge, I noticed that the new navigator had stopped updating. Another hang?
But that was about the worst time to do anything about it. A few kilometres later, though,
the other GPS unit stopped too. That was serious: I was relying on at least one of them to
get me there. Stopped in the emergency strip and discovered that once again I had
charge problems, this time not with the cable, but with the socket adapter. Possibly a
blown fuse, but of course not the time or place to investigate. By this time the new
navigator had come out of its hang to crash with a Microsoft-level error message:
The error message was more serious than it looked: the navigator had hung itself up, and
there was no way to reset it. I must remember to carry a paper clip in every car. Finally
Yvonne managed to power it off with sustained pressure on the
on/off button, and when it came back, it displayed the message again. It wasn't until the
third attempt that it ran normally. Fortunately at one of the many ensuing stops we found
another cigarette lighter USB adapter, so we were able to continue using it.
But only a few hundred metres further on from the initial stop, I got a warning on the
dashboard:
What does that mean? It didn't stop the car from moving, though it seemed a little jerky.
That could be due to the road surface, of course. Yvonne checked in the instruction manual—no mention of that particular message. What's a
powertrain? I'd expect it to run on rails. It's clearly an American word, and I'm a little
hazy about its exact
meaning. Wikipedia says it's just
about everything from the engine to the wheels. Is that what Holden means too? OED spells it the way I would expect, and confirms that it can have different
meanings:
power train n. Mech. the mechanism that transmits the drive from the engine of a
vehicle to its axle; the transmission; (also) this together with the engine and driven
axle(s).
I turned off the ignition, waited a while, and restarted the engine. Passed the power-on
check. But a bit further on, it happened again. And after the third time it failed the
initial check when turning on the ignition. Clearly it was getting worse.
Yvonne checked again and found a number to call, but it seems that that was a paid service
that had lapsed 5 years ago. Wouldn't this have to happen on a long weekend far from home?
In passing, it's interesting to note the price of fuel. In recent years it has hovered
between $1.20 and $1.50 per litre, but despite the long weekend, the prices were remarkably
low. We filled up a bit later and paid $0.999 per litre; later on the way back we saw a
price of $0.987. It's been years since I've seen anything that cheap.
Continued to Briagolong, and got there without mishap, coincidentally arriving in exactly
the 4 hours, 36 minutes that Google Maps had
predicted. On the way home, though, things weren't so good. After crossing a dirt road and
accelerating out, the car gave up, and I couldn't move it. And that in the middle of
nowhere.
Time for emergency roadside assistance? It seemed unlikely that that would help; the best
we could do would be to have the car towed
to Maffra and have it serviced there,
which would mean returning home by train. What a pain! Yvonne tried stopping some cars,
and one kind soul did some searching for us in the area, but to no avail. Yana, who was returning with us, went a nearby house and did some enquiry.
I tried calling back at the Klines, but all I had was mobile phone numbers, and I didn't get
an answer on either. Decided to bite the bullet and call RACV. They wanted me to sign up for 2 years, at a price
of about $240. OK, we can live with that. Next, “where are you?”. All I had was the
coordinates from the GPS navigator. And they can't handle that! Why not? What use was it
to pay my $240?
In the meantime, Pauline, the resident at the house where Yana had gone, had arranged for
Tom, a mechanic, to come along and take a look. He immediately diagnosed transmission
issues, which seemed reasonable, though some of the details didn't. In any case, he claimed
that the car would start and run, and there he was correct. So he led us to Maffra, and the
car ran normally. Should we really leave it there? Discussed with Tom, and he thought we
would probably get back home with it, and if we didn't, we'd be closer than Maffra. So we
set off, by the slightly longer way
of Sale, where there's a railway
station, but it gave us no further trouble, and we actually made it back to Yana's house,
and then on home via Ballarat. Interestingly, we saw a surprising number of breakdowns on
the freeway. Is that related
to Australia Day? Arrived without
further incident, this time after 5¾ hours and a total journey of nearly 800 km. What a
day!
But we weren't done yet. What does Google say a about Check Power Train?
Quite a bit, as it turns out, notably this article, which suggests that it's relatively specific to
the Commodore VZ. And it can mean
just about anything. Time to read the diagnostic codes out of the car.
At least for the first 1½ hours I had both GPS navigators running to compare them. One
thing's clear: the new one is so much easier to read. We were in bright sunshine,
and the display of the old one was almost unrecognizable. So if I can get used to the
interface of the new one, it might be worth keeping.
Twelve years ago we celebrated my father's 80th birthday
in Briagolong. Today it was his
sister Frieda's 90th birthday. And what a difference! There must have been twice as many
people there, many of whom I didn't know, and others whom I didn't recognize.
Frieda herself looks remarkably fit, in fact better than her younger brother and sister:
About the only thing wrong was that there were so many people there that we hardly had a
chance to talk to anybody for any length of time. It's nice that Frieda has so many
friends, and we're not the only ones who came a long way, but there's something to be said
for small gatherings.
After yesterday's excitement, spent most of the day picking up the loose ends, processing
photos, investigating potential causes of the car problems (though not to the extent of
reading out the error codes from the car). Yvonne made up
for not taking any photos yesterday by taking 191 of horses today, about half as much again
as I took yesterday.
Tony Abbott has lived up to my
expectations, unfortunately.
In fact, he's remarkably creative, and having long ago annoyed me and most people I know, he
now seems to be turning on his friends. Today
was Australia Day, a day on which
many Australians are recognized for their services to society. And this year our glorious
Abbott has revived the highest honour,
the Knight
of [the Order of] Australia. Two people were so
honoured: Angus Houston and
Philip of
Greece, the consort of Elizabeth
II of the
United Kingdom.
I've never understood these things, and it seemed somehow typical of Abbott that he would
recognize a soldier. But Philip? Firstly he's already a prince; why make him a knight?
Looking at the list of knights, though, it seems that his son is also a knight,
so there's some precedent. I still would have thought an Australian to be more appropriate;
Philip and Charles are the only non-Australians to receive the honour.
But other people haven't seen it as calmly as I. It seems that Abbott was so concerned not
to be refused that he didn't tell anybody in his party about it. Many were upset, some even
openly, and many have ridiculed him:
But now we have a real live error condition. What do they display now? The same! I have
no idea what they're there for, but they don't help with out “Check Powertrain” condition.
So there's nothing for it but to call up Ballarat Central Auto
Electrics and get them to read the information out, for which they charge the princely
(or is that knightly?) sum of $66.
Talking about the matter on IRC, Jamie Fraser said that his father had bought a Bluetooth coupled diagnostic interface for “about 6 bucks”, and that it was also
capable of reading out the data. That's less than 10% of the price that VACC ask for a single readout.
Did a bit of investigation and discovered that the one Jamie had found cost nearly $9 US,
but that's still cheap. And of course on eBay they're even cheaper. In the end decided on one costing $5.99 Australian, including
postage—currently the equivalent of $5.24 US. It looks as if it will work for both cars,
and at least it will serve us next time.
In the afternoon into town to have the data read out, and on the way the error indication
disappeared. No complaints any more. What causes that? In any case, since the problem
doesn't seem to be getting any worse, it makes sense to wait for the diagnostic dongle to
arrive and read out the information then. Now that we're near home, if we really run into
trouble, we can tow the car to where it needs to go.
Three weeks ago I bought a
very cheap “dumb phone”, one that runs neither Android nor iOS, but which has a real keyboard. It
was attractive enough that Yvonne bought one too to get away
from the pain of scraping glass to do anything useful.
But it's still a relatively modern device, with an illuminated colour display. And there's
the problem: in bright sunshine it's illegible, while old-fashioned unilluminated displays
have no difficulty. So we're thinking of returning them. While in town today, went to
Dick “We can't be bothered to serve you” Smiths and JB HiFi to see what they had. Nothing
as cheap as the $18 that we paid, of course, but also nothing that could be
demonstrated. How do we know if the display is any better? In the end we gave up; we'll
have to find a way to make phone calls outdoors.
While in town, also went looking for light fittings. Somehow people don't seem as
interested in them as they used to be. The first place we went to had closed down some time
ago, though it's still very much in evidence on the web. The second one wasn't quite where Google Maps had placed it,
but it had quite a number of light fittings, all but a few of which were out of the question
for each of us. Unfortunately, the ones that did come into question for me had little
overlap with the ones that came into question for Yvonne. It
looks as if we'll have some discussion to do.
Progress on the house is waiting for the plasterers to finally get back from holiday; Duncan
is hoping for the second half of this week. But it seems that Jason has done some work with
the shed:
Leonid has a particularly curly coat. Or he
did. I'll have to go back and look again, but at the moment the middle of his back has
relatively short, straight fur. Not very easy to recognize in these photos:
Received two 37 mm UV filters in the mail today, for Yvonne's
Olympus E-PM2 (well, more specifically for the 14-42mm
f3.5-5.6 and 17mm F2.8 lenses,
both of which have the same filter diameter).
Why UV filters? Digital cameras have UV filters in the camera. They're really workarounds
for the lack of an easy to operate lens cap. The real solution would be something like just
about every compact camera has, some form of robust diaphragm in front of the front element.
That could be quite difficult for wide-aperture lenses, but the very fact that the filter
diameter is only 37mm shows that it could work here. In fact, something
similar, the LC-37C auto open lens cap, is available, but only for the 14-42mm
F3.5-5.6 EZ. And even Olympus sells “protection
filters”, which aren't filters at all. And they cost between $37 and $59. Why?
CJ along today with Sue with more problems relating to living in the age of the Internet.
Seems he was a little too trusting of some company or another, and now we need to get some
money back from them. Hopefully it'll work. It's sometimes good to be reminded of the
transition that people have to make when living in the Internet.
Yvonne has been training the dogs to sit on command, mainly,
I think, because Sandra the dog trainer told her that you can't train sight hounds to sit or
drop on command. Not so here:
Call from Jason Bruty today about the missing components of the shed, and also about the
(“Personal Access”) door. The former are still not there, though he has been trying to get
some sense out of them for a week now. And the latter is a clear case of damage at
delivery: it seems that they put the PA door at the very bottom of the package they
delivered, which looked like this:
There was also a phone call on the answering machine from David Fitzgerald of Widespan, wanting to know when somebody would be on
site. Jason had already told me that he had tried to get him to pick up the components
(probably in Geelong), and that he had
pointed out that they had the obligation to deliver the components. David was very vague
and repetitive about the delivery, asking whether somebody was on site all the time. I got
the impression he wasn't listening, but on the third time I think he understood when I said
that they should contact me before delivery and that I would be on site.
And the PA door? No mention. When I mentioned it, he pointed at some conditions that I had
allegedly received, saying that the doors must be inspected on the day of delivery. Apart
from the fact that that's downright impossible, I didn't have any such document in my files.
But no, that's the way it is, and I definitely received the document. Asked to speak to his
superior. “My superior is the General Manager”. “OK, connect me to the
general manager, then”. “He'll only tell you the same thing I did”. After a bit more
nonsense, I hung up on him, arguably not the best choice, but these people irritate me.
He called back and started the whole rigmarole from the beginning—even the question about
whether I wanted the missing components delivered. I don't know if it's just him or company
policy, but I got the impression that he was doing everything possible to avoid having to do
anything at all. It's clear that I won't get anything useful out of him. He even refused
to send me a copy of the document that I allegedly received—“It's on the web site”. In the
end, I hung up on him again, not once, but twice more.
Looking in my mail, I finally found the text he was referring to, in a purchase agreement
sent to me as application/octet-stream, which I was able to identify as
a barely legible PDF document. The first
page says:
Rather to my surprise, it's possible to read the second enlargement of the conditions:
9.1 At the point of taking possession of the building, it is the Purchaser's
responsibility to ensure that all materials nominated on the delivery dockets are supplied
and that they are in a satisfactory condition. Any damage or short supply must be marked
on the delivery dockets and signed by the driver.
9.2 ... Any concerns in respect of defects or damage to any P.A. Doors or roller doors
must be dealt with and advised on the day of delivery. Failure to so advise may negate the
Purchaser's ability to claim for any damage.
OK, that looks like a CYA clause
to me. What's wrong with it in practice?
The whole page of conditions contradicts the first page of the agreement.
They were supplied nearly two months before delivery.
They're encoded incorrectly and written in script that is barely legible on a computer
screen, and they weren't supplied to me in writing along with the other documentation
that they supplied.
At no point were these alleged conditions drawn to my notice.
It's completely impractical to inspect the package on delivery. These sheds are sold to
end users, not companies who will erect them immediately. To inspect it would have
required complete unpacking, something that the delivery driver would have had to do.
The clause states that “failure” to do so may negate the ability to claim for any
damage. That's reasonable, I suppose. If a door is damaged, it's important to be able
to establish when the damage happened. But we know that: Jason has photos showing that
the door was damaged due to being unloaded on a stone. So the intention of this clause
is not relevant.
Apart from these issues, neither Jason nor I are very happy with the finish of the shed.
Sheds aren't supposed to hermetically sealed, but the gaps under the roof here are much
larger than either would have expected:
Conventional wisdom has it that it takes the best part of two days to install a septic tank
system. But Warrick Pitcher and Mari Hendriks are not conventional. They installed it
today, starting admittedly at 7:30. By the time I came on site in mid-morning they were
well on their way:
My weekly house photos require a lot of
processing. To make a panorama I take multiple views of a scene and stitch them together.
To get the requisite dynamic range I take multiple images of each view, with exposure
bracketed in intervals of 3 EV.
First I use align_image_stack and enfuse. to create blended photos, and then I use Hugin to create a panorama from the views.
That's a lot of processing, and I'm still
producing JPEG intermediate results. Why?
I have the space, and I have the processing power. So today I tried
using TIFF intermediate files. That's not
for the first time, but last
time I didn't have very good results, at least because of the performance of the
machines I had at the time. This time was different: my machines are faster, and the
network link between dischord (Microsoft) and eureka (FreeBSD) is now 1 Gb/s. But things still didn't work as
I expected:
I've never used that much memory before, but why am I not using any CPU? Did some
investigation and discovered that the disk was the bottleneck: I was running
12 align_image_stack and enfuse processes in parallel, each of them working on
70 MB files, and the disk just didn't manage it. High time to finally finish my new
machine with its SSD system disk.
So I reduced the concurrency to 2 processes, which sped things up a lot. And this time I
didn't have any trouble with stitching; I even got finished relatively quickly. The only
question that remains is what to do with the intermediate images:
=== grog@eureka (/dev/pts/7) ~/Photos/20150131 142 -> du -sc * 27769 C
19 Pano
4764 orig
...
33444 total
C/ contains the intermediate images, and orig/ contains the original images
and their TIFF conversions. The values are in megabytes; 33 GB per weekend is rather a lot,
so I think I'll remove them again. They can always be recreated relatively quickly.
Chris Bahlo's gelding Chewey has been on the Stones Road property for months. Problem:
despite a new actuator, the electric fence is not very effective, and Chewey goes through
the fence at will. Yesterday Warrick and Mari saw him walking over the freshly laid sewer
soakage and warned me what could happen. So today Yvonne picked him up and walked him home:
Those three views are as close to perpendicular to the house wall, for future reference.
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.
