pfh: It's psyco, not psycho. A PowerPC port doesn't seem imminent, sadly.
I think I have discovered my two criteria for being happy to use a machine as my mail server:
dalke: Oh! I'd kind of assumed you were already married :-) Congrats, anyway.
As trailed, this is my EuroPython diary. Also find it here. I was too busy to be as verbose as last year.
Were spent at the PyPy sprint. Lots of interesting things happened but they will not be described here.
Like the dominant calendrical system not having a year 0, EuroPython 2003 does not seem to have a day 0.
It's harder to write a conference diary when you're running the conference.
Maybe I'll write some more tomorrow.
First lesson of conference chairing: when you've said that a speaker can use your laptop to show his slides, it helps if you put the slides on your machine before the talk is due to begin.
Still, some running around sorted that out and if that's the worst thing that happens during the conference, I'll be very happy.
Radio mike fun.
Franis Glassborow is giving a keynote on "Building Successful Distributed Communities", theme "People Matter", which, um, I'd kind of spotted by now. Still, it's probably a point that people could do with having hammered down their throats.
There are a lot of small comapnies here. This makes my NTK "£80 million in venture capital" T-shirt much appreciated :-)
Beer!
My day off: no Python language talks today.
Something about libxml2 by Daniel Veillard. I think he might be French :-) I personally don't care much about XML related technologies, but it does occur to me that if you are having problems parsing a multi-gigabyte XML you have two options:
- writing a stream parsing API.
- finding something else to do with your time.
I can't believe XML is really a sane way of carting that much data around.
WebTK seems pretty interesting, although the speaker is spending a lot of time being introductory. I'm going to ask for a demo at the end of the talk.
He did one anyway (although the final, most complicated one refused to work).
Skipped a couple of talks during which time I tried and mostly failed to upload my speakers' slides to the EuroPython database. Grumble.
Went to Martijn's talk on his XML database, and found it surprisingly interesting given that I'm not interested in XML or databases :-) Well, the theoretical aspects of database querying are quite cool, and that's what he was mostly talking about.
Lunch. Can it only be a year since thel last EuroPython? One of the biggest changes is that I had met noone who attended EP2002 in the flesh, and this year I spent the four days before the conference at a sprint with 12 people who are here (and there was a Zope3 sprint next door with another 10 or so people). Also, standing at the front of the auditoria and introducing speakers does mean that people know who you are :-)
Random musing: good talks are probably more interesting than reading the equivalent material on some webpage somewhere. When Martijn & I did Moshe's "Do's & Don'ts In Python" talk yesterday (he got the dates wrong...) we tried hard to make it very interactive, asking people in the audience to raise their hands if they'd made the mistake under discussion, disagreeing with each other. We had a lot of fun giving the talk, and I think the audience enjoyed being there (at least, no-one's said "you really sucked!" yet). Basically, I should remember to use the audience if I ever have one. Being in the large auditorium probably doesn't help.
Another tip: have an editor window open in the background prepared with a really big font.
Umm, this only takes me to middle of the afternoon and I don't remember in detail what I did in the afternoon. At some point, though, something finally went click in my brain and I now understand (more or less) PyObjC, and I spent about two and a half talks (including Guido's keynote) hacking around in Interface Builder. Jack Jansen would like me to help him write a tutorial to help other people get to this point, and I'm not sure I can :-( Before I didn't really get it, and now I do. Oh well.
Guido's keynote was interesting. He mentioned PyPy a lot, and talked about some dreams for the far future of Python. Oh, and the ternary operator is dead (yay!).
More beer!
The last three Python language talks are today. The first is a 90 minute Stackless Python presentation, which was probably my biggest gamble: would anyone turn up? Would they care? Also, Christian wanted to demo a 3D MMORPG largely implemented in Python, which meant we needed a beefy desktop machine and network on the stage. This nearly did lead to disaster: when the game tried to get the projector to do 1280 x whatever, the projector barfed and switched to 320x240, which the game wasn't too fond of. So, get Christian to ad lib, plug machine back into monitor (which crashes the machine, so reboot), fire up the game again, tweak game to do 1024x768 (which we hope the projector can do), unplug monitor, reconnect projector, observe game appearing nicely at 1024x768, but also that the machine had now hung (again). But one more reboot does the trick, and Christian can finally demo his apps. It was lucky I was around to do this fiddling, so the speaker could go on doing Q&As.
I think the talk went really well, in the end. It was in the bigger auditorium so the 40 or so people who turned up hardly filled it, but this was a pretty reasonable fraction of the attendees of the con, given the other talks that were on at the same time. It was also interesting that Christian (and I) had guessed the experience of the audience completely inaccurately. When Christian asked "how many of you have used stackless?" only about three people raised their hands, which meant that most of Christian's slides weren't really suitable. He did a godd job of giving an instant crash course in programming language implementation and activation records and the issues Stackless attempts to resolve. I think I could have presented the same material, but would have needed considerably more that five seconds to prepare. This and the fact that I spent about twenty minutes repeatedly crashing Christian's computer meant that this was an almost slide-free talk, but this didn't seem to matter -- it may even have been a good thing.
Next up was Anna Ravenscroft talking about how to teach Python to absolute newbies to programming -- a slight contrast to the previous! But still good stuff.
My final talk is Alex Martelli -- Anna's SO, though I don't think I knew this when I scheduled their talks next to each other -- talking about "What's New in Python 2.3". This is a bizarre talk to me, as I implemented many of the things he is talking about...
By this point my brain was utterly fried, a state no doubt exacerbated by the incredibly loud building site outside our hotel. And Jacob is trying to get me to evaluate (very simple) patches to the compiler package of Python 2.3. The mind is willing, but the brain is not cooperating.
The afternoon, for me at least, is entirely lightning talks. My talk (on reStructuredText) is scheduled for 1410, but is swapped to 1610 (and eventually gets given at 1550) which is a very good thing because at two o'clock I was feeling really grotty. Lunch, and plenty of coffee woke me up enough to keep me upright during my talk. Most people were polite about my talk, which is reassuring, although I suspect I was talking too fast (I usually do when talking to groups).
Some of the other lightning talks were interesting and/or fun, but others sufferred from the standard problem of basically being 45 minute talks crammed into 5 or 10 minutes. The absolutely standard problem with talks is overlong introductions. The WebTK talk sufferred a bit from this -- giving examples of why you might want a dynamic website, for example.
Anyway, that was EuroPython 2003. It's hard to be objective when you are so involved, but I've really enjoyed this year, and I would hope that the organiser's having a good, fun conference means the attendees should too. Next year? Only time will tell...
I did wonder if this was a deliberate piece of irony by the author though (I didn't finish it).
And yes, *everyone* should read 'Politics and the English Language'
Last day of PyPy sprint, EuroPython tomorrow.
I leave for another PyPy sprint, to be followed by EuroPython. Registrations have topped now 200 now, but there's still time to register if you're dithering!
Hopefully life will get a bit less insane after that. Apart from two weddings, Ashton Court Festival and trying to find someone to take the spare room in our flat...
So, lets see. Tomorrow I'm flying out to Gothenburg to take part in the second PyPy-sprint; I'm staying with a friend in London tonight.Secret Packing WeaponThen a week later Gothenburg -> Edinburgh via Stansted for a week-long Workshop on Derived Categories, then back home.
Two weeks after that, will most likely be at another PyPy-sprint before EuroPython, then EuroPython itself.
July should be more restful, but I still have two weddings (one family, one friend).
I'm hoping to get away with the flights involved in the next two weeks without checking anything into the hold. My secret aid in this is my many-pocketed rain coat -- at the moment its pockets contain a usb mouse, an ethernet cable, some mouthwash, my meds for the gum infection, my travel documents and a pen. And maybe some other stuff.I hope this doesn't freak out the security scanner at the airport...
Wow, having an infection around one of your wisdom teeth give you more-or-less constant pain really doesn't help your mood. I would be hard pushed not to hurt someone who tried to keep me away from my ibuprofen tablets right now.
Am on anti-biotics & it's clearing up, but this sure makes me appreciate the generally very good health I've had these last few years.
OS stuff: not a lot (see above).
I've more or less finished sorting out my track for EuroPython 2003. I'm certainly looking forward to hearing the talks I've scheduled!
Relatedly, online registration is open, with early bird rates until the 31st of May (so a bit more than a fortnight from now). Register early! Register often! No, wait...
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!
