| Greg's diary |
| October 2004 |
| Today's diary entry |
| About this diary |
| Next month |
| Greg's photos |
| Greg's home page |
| |
|
|
This is a shortened version of my personal diary for October 2004, shortened to those entries of interest to computer people. See http://www.lemis.com/grog/diary.php for my personal diary.
In the meantime to Mount Barker to look at our kitchen oven, which has been having temperature control problems. Fun to be able to do some simple debugging; it's looking either like a thermostat with fused contacts (the mechanical part works fine, and so does the electrical part most of the time), or possibly a short to the chassis; it seems that the elements are live, and that the thermostat shorts to neutral. A strange arrangement, meaning that a short to earth will also cause it to heat. The only explanation I have is that this oven is German. German power plugs don't distinguish between live and neutral (you can insert them either way round), so probably the constructors didn't consider this lethal arrangement to be a problem.
Sat Oct 2 23:59:13 CST 2004 root 666 80.9 0.2 1264 868 dco R+ 12:04PM 710:27.57 find /src -type fThat's 710 minutes out of 715, so clearly the 81% indicated CPU is incorrect. I wonder how long a single conditional breakpoint takes when the breakpoint condition isn't met.
Apart from zaphod panicking, we still have problems with the oven that I looked at with Barry on Friday. Had it connected to a different power socket, one on a residual current detector; it blew, thus confirming our hypothesis of a short to earth.
Apart from that, documentation is the flavour of the week. There's the possibility that I'll be presenting my work at a conference next year, so it's high time to start thinking about documenting what I'm doing.
During the night I had had a flash of inspiration about the oven, and in the evening confirmed it. In fact, it had been staring us in the face: when we first got it back from repair (replacement of a heating element), one of the connectors to the element had shorted to the rear panel, tripping the circuit breaker. I had fixed that some weeks ago by turning the spade connector around, increasing the distance to the rear panel by a couple of millimetres. The other one didn't seem to need it. On Friday I had actually shown Barry the spark marks of the short circuit—and there were two of them. But it didn't twig until this morning. Obviously the first one was live (throwing the circuit breaker), so the other one had to be neutral. Turning the second spade terminal around was all I needed to do. But I should have realized that on Friday.
Apart from that, more documentation. It's difficult to find a good way to explain the concepts, and I spent most of the time trying to structure the document.
Documenting my program is not easy. I know what I want to say, but how do I structure it so that other people will understand?
Documentation is gradually fitting into place. It's a big job, though.
In preparation for our Europe trip, installed Apache on my laptop, eucla. To my surprise, it worked out of the box: I had already copied the configuration files from echunga (also known as http://wwww.lemis.com/, my local firewalled server).
Filling out a senate voting form is also a pain. We had only 47 candidates; in Victoria it was 61. We have the choice of choosing one of about 10 “tickets”, predetermined (and undisclosed) sequences, or entering our own sequence for each candidate. The ballot form is so big that it needed to be folded to fit in the ballot box; that doesn't exactly suggest careful planning.
Spent the rest of the day tidying up AUUGN. It's about ready to go now. I've started including the Lions Book in installments; I wonder how popular that will be after 27 years.
Finally finished and sent off both AUUGN and my tutorial notes for BSDCon Europe. Thank God for that!
Off to Europe today, leaving at 9 am. It's the first time that Yvonne and I have been on holidays together for nearly 10 years, and it's a funny feeling. By midnight we were in Singapore, this time terminal 2, where there was no free Internet access at all, despite all claims. There are computers everywhere, all offering Microsoft and “Internet Explorer”, but no other protocols, nothing secure at all. I mentioned it to a couple of people using the facility, but their response was moronic.
Arrived at Frankfurt at 6 am and had to wait until the very end for our “priority” baggage; somebody had removed the priority tags. Still got to the railway station relatively quickly, and we left on a train within 1½ hours of arrival. Got to Kleinenbroich with no trouble, only a little after 3 hours after arrival. We keep forgetting how small this country is.
I hate Braun electric toothbrushes! They work well, but they must be some of the most unreliable electric devices I know. Today my new toothbrush failed, the one I bought in March this year to replace another one that never worked properly. It always seems to be an issue of electrical contacts, notably the switch; shaking it would turn it on or off. Today it decided to stay on until it discharged the battery. Sure, it's still under guarantee (in Singapore), but that doesn't clean my teeth.
Second day of the horse training today, and I spent more time catching up on computer stuff, including updating my web page generation scripts. Got that finished and then discovered “It's just what I asked for, but not what I want”. More work, but not today.
In the afternoon off to visit Gabi and Gerd Reichert, old friends of Yvonne (she's known them longer than me). They live just on the other side of Passau, so we thought of dropping in to Passau and buying a present for Gabi, whose birthday is coming up.
Quiet day. Spent some time trying to access the network with Gerd's ISDN router, which is connected to t-online.de, and kept getting messages like:
*** Connection closed from efnet.xs4all.nl: Connection reset by peer
After far too much time came to the conclusion that something was disconnecting the IP connection before dropping the ISDN line; a simple ping -i 30 to a remote machine solved the problem, but that took hours of continual reconnects to understand. They still have big problems with name resolution, though, and they seem to block local name resolution traffic. t-online may not be Deutsche Bundespest, but they still wouldn't be my choice if I lived here.
Also did some work on the slides for Karlsruhe. Finally got a workaround for the hard-coded page size in the PostScript output, but it could still do with some improvement.
Up through the Black Forest through less than excellent weather, and finally made it to Karlsruhe in the early afternoon.
In the evening other people started to trickle in, and off to the Wolf Bräu brewery pub, ending up with about 18 of us. The food was acceptable, but not more; the beer was quite good.
My debug workshop again today, with a surprising number of participants. I didn't count them, but I was told there were 40. Tried to emphasize the kernel aspects more this time, but shortly after lunch it was clear that I was losing them. Tried instead to debug a user's problem, but he was running a recent FreeBSD CURRENT system, and it seems that remote gdb no longer works. Spent the last ¼ of the workshop trying to get things to work, without success. I suppose that's an illustration of real-life debugging, but hardly satisfying.
Out with the locals for dinner in the evening, missing the majority of people who were just leaving when we returned. That may have been just as well.
Today the conference proper started, and I saw a few more familiar faces at breakfast, notably Ollivier Robert and Jordan Hubbard. Nevertheless, despite the good turnout—over 200 people—a lot of the old familiar faces were missing.
Jordan started with a keynote on MacOS and BSD, pointing out, rather amusingly, that far from dying, BSD has (with Apple's help) overtaken Linux as the most used free operating system. Then Antti Kantee talked about checkpointing long-running applications, rather the way Tandem did it decades ago; reminds me of HIBUG, a debugger I wrote at the time which had the ability to save (re-)executable dumps. After that, Alistair Crooks and Veleriy Ushakov demonstrated some miniature machines, halfway between a small laptop and a pocket calculator, running NetBSD. That might be just what I need for a more comprehensive video remote control.
In the afternoon more NetBSD stuff about the packages collection and cross-building, and Yannick Cadin did a useful tutorial on DHCP, unfortunately running out of time. I must check the slides that he's going to put up; they could be very useful.
In the evening to the Nancy-Halle to listen to Luigi Colani talk about his understanding of technology. I've never been impressed by him, and he didn't change that with his talk. Seeing a model of an aeroplane with the comment “Aerodynamics by Colani” almost sounded like a joke. Had some food there as well, not too bad, but I was more than a little put off by the fact that we had to pay for all drinks, even water (and pretty steep prices at that; water was € 8 per litre). Left pretty soon and off to the Wolfbräu for some more substantial and much less expensive drinks. Again early to bed.
This morning for the first time, we were asked for our room numbers at breakfast. I was wondering about that: it seems that a number of people not staying here had breakfast. At least at our table, all people were really staying.
I only found one talk worth listening to today, Massimiliano Stucchi and Matteo Riondato talking about FreeSBIE, GUFI's (Italian FreeBSD User Group's) answer to Knoppix. Interesting talk, going into much more detail than I had expected. Others seemed to think so too: the talk came third place in the “best talk” competition.
Apart from that, talked to a number of people, though not as many as I had hoped. Towards the end, Erwin Lansing brought some samples of his own beer; he got the idea when visiting us about this time last year.
In the evening, off to the Alten Brauhof where we ate on Friday. Somehow ended up with just about everybody who was left there; took a number of photos. Nonetheless, the conference finished without me meeting a couple of people I had wanted to meet.
| Greg's home page | Today's diary entry | Next month | Greg's photos |