Tuesday, December 27, 2005

Day 238

[reading: A. Zee, "Quantum Field Theory in a Nutshell"]

Got around to doing a quick and dirty digital version of a collage I did a year or two ago (even at this cut-down resolution, it's still pretty large).

Sunday, December 25, 2005

Day 236

And it's now official. As previously rumoured, the new Globe season includes Titus Andronicus (plus Antony & Cleopatra, Coriolanus and Comedy of Errors, but obviously they're rather less interesting).

Saturday, December 24, 2005

Day 235

[reading: Joel Spolsky, "Joel On Software"]

OK, vegetarians and vegans look away now.

No, really, I mean it.

This evening's dinner. Before cooking:

After cooking:

After eating:


Thursday, December 22, 2005

Day 233

Now that work seems to have officially finished for the year, it seemed like an ideal opportunity for nipping in a Slacker's Lunch. The strenous activity this time round was table football (foosball for US readers). There was a period, back when I worked in an office that had a table, when I was vaguely competent; sadly I seem to have lost the old magic. (I suppose it's just possible that there might be another reason why I wasn't playing particularly skillfully).

Anyway, I also managed to persuade my fellow slacker to come round for food on Saturday; I think the key deciding factor was revealing exactly what we've got planned for the food...

Wednesday, December 21, 2005

Day 232

Things seem to be moving along on the Project X front. My co-conspirator has officially started on working out her notice at her job, so she should be able to work full-time on it by April. Project Y will probably finish around then (at least according to the estimates I put together earlier this week), so that should dovetail nicely. In the meanwhile, she's going to spend a day a week on the project, and I should be able to spend one or two days a week, so things shouldn't completely seize up before then.

However, hopefully I can finish off a couple of other things on my to-do list in the next week and a half to get them out of the way.

Tuesday, December 20, 2005

Day 231

[reading: Frank Miller, "That Yellow Bastard"]

Last day of work on Project Y for this year today, which was neatly followed by the office party. I'll do another day or two of work on Project X; then I have some Copious Free Time until January, which will be good.

Sunday, December 18, 2005

Day 229

[reading: Frank Miller, "The Big Fat Kill"]

Bugger. My chances of getting through my to-read pile have suffered a serious set-back: I returned from a trip into town with seven more books (and rather fewer presents for other people than I'd intended).

Saturday, December 17, 2005

Day 228

Came across another sport where men and women compete equally, and this one isn't even particularly for rich people: sheep-dog trials. Yep, it looks like "One Man and his Dog" is back.

Somehow, it's marvellously soothing to watch; it's like a visual version of that old traditional classic, the shipping forecast on Radio 4. Mike and I were actually discussing whether there might be a market for a CD of shipping forecasts: you could get a couple of hundred of them onto a CD, then set it to shuffle play.

Which does slightly remind of something I heard about some years ago. Apparently, it used to be possible to get hold of a mag tape full of random data. Random data is very useful, for generating tests and performing statistical analyses, but most of the time people just use pseudo-random data generated by the computer itself. This has gone spectacularly wrong on occasion, so having true random data available is useful.

So how do you get real random data? Well, in this case by monitoring output of radioactive decay; in a particular time period, if an odd number of particles are detected, then a 1 is added to the data, if it's even then a 0 is added. There was also apparently a slight statistical tweak, to allow for the fact that zero is an even number (because the distribution of the numbers of particles produced is obviously cut off at zero: there's no way, even in quantum mechanics, to emit -1 particles).

Friday, December 16, 2005

Day 227

[reading: Tom Phillips, "A Humument"]

Just completed my first week with a full five days of work for a long time (er, 227 days I guess) and I'm very tired as a result. Obviously, I'm not expecting to get any sympathy :-), but it's interesting how quickly you get used to a life of leisure. It's not even like I've done the same number of hours as I did at my last job—that was minimum 45 hours a week, and I've probably only done around 35 this week.

Anyway, we went out to see some show-jumping this evening, which culminated in a puissance. I'd not encountered one of these before, but the basic idea is high-jumping for horses. Each round, they raise the main wall a little bit and anyone who fails to clear it is out of the competition. For this particular puissance, two riders shared first place by successfully jumping a 7'3" wall. That's a long way up. Even the riders could barely see over the wall, and they were sitting on horses.

Show-jumping is also interersting as it's almost unique as a sport that has an even playing field: men compete against women; over-fifties compete against under-twenties; rich folk compete against, er, other rich folk. I can't think of many other sports where that happens. Sailing and motor-racing, I guess—sports where the motive force is provided by something other than human muscles.

Wednesday, December 14, 2005

Day 225

[reading: G.K.Chesterton, "Father Brown: Selected Stories"]

Following on from my monster tax refund, I just got a cheque from the government refunding some National Insurance. I guess those virgin sacrifices on stacks of inverted tax forms really did the trick.

Also, my latest Amazon aid parcel has sadly undone my good work, and my to-read book stack is back into double figures.

Monday, December 12, 2005

Day 223

More on Project X today, but I digressed to have an interesting discussion with my co-conspirator. We were talking about the differences in approach between ad-hoc standards bodies like the IETF and the more official committee-driven international bodies (like the ITU, ISO or the OIF). The more lightweight approach always seems to win out; I guess the main reason is because the committee-driven approaches have a tendency to settle an argument between approach A and approach B by including two distinct modes in the resulting standard, thus implementing both. By the end of the standards process, the net result is almost unimplementably complicated.

The more open approach also seems to work well in competition against standards that aren't decided by committee, but which are fairly tied-in to particular manufacturers. Not so sure why that should be the case, but the comparative expense presumably plays a part.

As an aside, both of these are actually illustrated by the very first piece of software I worked on. This was back in 1987, and I was part of a team implementing an email gateway—back when almost no-one in the world had heard of the concept of email. To this day, almost no-one in the world has heard of the two different email protocols that we were converting between: X.400 (an ISO standardized protocol) and SNADS (an IBM-specific technology). It was another four years before I actually got an email account of my own, and that was of course an SMTP account.

The list of open, simpler standards trouncing the expensive, complex standards is pretty long (and this is just the ones I happen to know something about).

This doesn't seem to apply so much to the lower layers; the standards for things like Ethernet, Token Ring, X.25, SDLC, SONET, SDH and even ATM have all been pretty successful. I guess it's harder to end up with an all-inclusive compromise for hardware: it's not possible to have a pin that handles either nine volts or six volts. (Although they seemed to try hard with ATM: 53-byte cells just seems like such an odd number, you suspect it was arrived at as a weighted average between those who wanted 48-bytes and those who wanted 64-bytes).

Sunday, December 11, 2005

Day 222

Went out for a walk this afternoon, and at first it looked like it was about to rain heavily, with dark clouds on the horizon . . . but then I realized that it was actually just burnt petrol, and unlikely to turn to rain.

The Space Cadets show has become rather dull; so far, it's degenerated into the usual reality show format, populated by the usual media whore contestants. Perhaps it will pick up when they actually go into the faked up spacecraft.

Friday, December 09, 2005

Day 220

Had my co-conspirator round to work on Project X today, and it turned out to be surprisingly difficult to share data between two computers attached to the same LAN. I'm guessing that I need to frob some ZoneAlarm settings, but nothing obvious made a difference in the first five minutes so we just resorted to sneakernet. I also need to sort out a longer piece of CAT5 to lower the probability of me either

  • a) tripping over a network cable, or
  • b) garroting myself with a network cable
or just possibly
  • c) both of the above simultaneously.

Wednesday, December 07, 2005

Day 218

And in other shocking news: a TV show that I actually want to watch. And it's a reality TV show, too.

Actually, the more interesting parts of the first episode were the behind-the-scenes bits about how the production crew went about converting a disused military base in Suffolk into a mock-up of a Russian cosmonaut training facility (buying supplies from markets and shops in Moscow, raiding an aircraft scrapyard). Hopefully they'll keep including some of that kind of thing rather than just degenerating into Big Brother with claustrophobia.

Tuesday, December 06, 2005

Day 217

... in which I learn that when they say there isn't much stack space available in the Linux kernel, they really mean it.

Actually, the real reason it took me so long to debug this particular problem wasn't just the stack space; the compiler optimizer wanted to play as well. I'd foolishly put something large onto the stack, in preparation for some later code, but not actually used it—and so the compiler had quietly stripped it out. When I did eventually start using it, some time later, things began to go awry.

At this point, the usual rule of "cause of bug == most recent change to code" no longer applied, and as I don't have a debugging system set up yet, I pretty much had to resort to shotgun debugging.

In fact, a kernel debugger wouldn't actually have helped much in tracking down the problem. The key piece of evidence that I eventually twigged to was that the crash would appear and disappear as I included or removed chunks of the code, even though the code was never run. That is: if I added in code that referenced the big thing on the stack, the optimizer could no longer quietly elide that buffer, and the code would crash before it got anywhere near the new code. Doh, problem solved.

On the plus side, I got paid today. When I first looked at the payslip, I had a worrying moment when I saw the size of the Tax number in the deductions column. Then I noticed it had a minus sign in front of it—which was just as well, given that the tax number was bigger than the pay number. So I ended up being paid about treble what I expected, presumably because of the mysterious machinations of the PAYE system.

Sunday, December 04, 2005

Day 215

Nipped in a bit of time this morning on Photoshop; following the monster scanning session, there are some of the better individual pictures that I'm slowly returning to and tweaking. Nothing too serious—scratch and dust removal, a spot of Levels and maybe a bit of sharpening. The eventual intent is to have a digital portfolio of my better pictures, the ones I tend to frame up and put on the wall.

Then it was out to the Barbican for a bit of culture: Yo-Yo Ma playing three of the Bach cello suites (plus three encores). 3 and 6 were great; 5 was very intense but perhaps a little ponderous for me. It's hard to believe he's fifty—he looks like he's in his thirties.

The coughing of the audience during the performance was kind of odd; there was the usual few coughs during the actual playing, but lots of people seemed to hold it in until the end of a movement—the gap between some of the movements was a veritable cacophony of coughs, and Yo-Yo Ma would just pause and wait for it to die down. It's a long time since I last went to a classical music concert, so I don't know if this is normal or not.

I was also reminded of how awful the Barbican itself is, both inside and out. Horrible concrete, which I'm told is now Grade 2 Listed and so is now stuck there forever. It's also amazingly difficult to find your way around, again both inside and out. The inside would work well as a map for a multiplayer FPS game—lots of nooks and crannies to camp in, and lots of stairwells and mezzanine levels. The outside is just as confusing, with different levels of walkways and stairs and corridors. There are occasional maps displayed on the walls, but these are more likely to confuse than help because they've taken the bizarre approach of orienting the maps differently in different places. (My only theory for this was that it might be a misguided attempt to align the maps with the walls that they're on: a map on an east wall has west at the top and so on). They obviously know it's a problem—the walkways that lead through the residential part to the theatre part all have yellow lines painted on the ground to help lost tourists ("Follow the yellow brick road! Follow the yellow brick road!").

One of the other folk (whom I'd not met before) in our little group has a flat in the Barbican, so we stopped off for a coffee there after the concert. Quite a nice flat, in a 1970s kind of way, but what was more impressive were the brochures that he had for a couple of other flats he was keen on buying. They were also at the Barbican, but up at the penthouse level, over three floors and with a great view. Price: a cool 1.8 million pounds. At this point I felt much less guilty that a complete stranger had bought an expensive round of drinks for us all.

Friday, December 02, 2005

Day 213

Some Linux confusions this morning. I'd installed Fedora Core 4 over RedHat 9 on one of my home machines the other day, and all had seemed to go well. However, when I booted the relevant machine this morning, I could get in past the graphical login ("Can't start session due to some internal error"). Following a tip or two here, I tried booting from a rescue CD and running yum remove gdm.

This did seem to fix the graphical login problem, so I could log in as a normal user, but then I discovered that I couldn't run su ("Cannot execute /bin/bash: Permission denied"). Given that I'm playing with kernel development, I really need to be superuser.

There seemed to be a variety of advice out there as to what might fix it, so I went back to booting from the rescue CD and blew away /var/log/wtmp and /var/run/utmp and altered /etc/selinux/config to run in permissive mode. One of those fixed things; not sure which, but I strongly suspect the latter.

So I guess I'm still not sure that Linux is ready for the great unwashed masses yet . . .