Tuesday, June 26, 2007

China has a Norwegian speaking techie population

And from the looks of it, they like OpenBSD or at least PF. I can see it in my logs.

As regular readers will know (hi, all three of you), while I was attending EuroBSDCon 2006, I moved the PF tutorial's home to NUUG's server. As a consequence, I now have read access to the web server logs related to the material I have put there.

That's how I, thanks to a tiny little statistics script, know that at this moment 13022 unique IP addresses or host names have hit one or more files in the tutorial directory tree since I moved the files to this location.

It won't surprise you that I sometimes when I really should be doing other things, glance at the logs and occasionally see something interesting.

At times I see stuff like somebody at fetching http://home.nuug.no/~peter/pf/no/langbrannmur.html - the Norwegian version of my PF tutorial (which is sadly behind times, unfortunately), as one long html file.

If you do a 'whois' you will see what I mean. Somebody in China is reading about how to set up PF, in Norwegian. The funny thing is that something like this happens fairly regularly.

If we can trust the whois data to be correct, this could mean several things:
  1. Educated Chinese prefer to read networking literature in Norwegian over English, even when the English version is more up to date
  2. There could be a number of Norwegian network people in China who feel better after reading about PF in their native language
    -- or --
  3. China is so big and has so many people who are potentially interested in PF that at least a few times a week one will fetch the Norwegian version of my tutorial by accident.
I just thought I'd share it with you guys. And now it's 13026.

[and yes, there is option 4: Chinese robots, slurping away anything they can find. But the first theory is a lot more fun.]

Which Windows XP version is it on that laptop?

Q: Which Windows XP version is it on that laptop?

A: None. I run OpenBSD.

That exchange happened in my office a few days back. The reason? At my day job, we do a number of different things, and at times we do tests on new hardware for customers who require that. So a customer asked us to do an assigment which involved going to an airport in a different city and test some equipment there. And they asked if we would have a laptop with English-language Windows XP on it to bring along. That's when I answered that actually my laptop does not run Microsoft Windows at all. I added that if I could dig out the restore disks for my old one, it would probably be Norwegian.

Frankly I'm not sure what my response my colleague sent back to the customer, but it probably said something along the lines that we may be able to dig out one with Norwegian XP on it. It took me only a few minutes to dig out the restore CDs (for some reason my old Fujitsu-Siemens laptop (an Amilo 1840W if it matters) came with two apparently identical restore disks), and put them on my desk. Then of course I forgot to bring them with me (that machine was in my attic at home), but finally I brought the machine into the office this morning.

Why bring out a 2004-vintage machine in the first place? Well, the ThinkPad R60 I'm typing this on did come with some sort of Microsoft system on it, halfway installed with no backup media, and since this is the machine I rely on for stuff I need to do every day, it has been running recent OpenBSD-current snapshots since I got it last October.

So for a detour into the world of mobile Microsoft computing, I needed to get the older unit up and running again. It had worked reasonably well up until I bought the newer machine - with a 3.2GHz Pentium IV and three fans in it it sounded like a hangarful of F16s and it never did have more than about 72 minutes of battery life (since reduced to zero), and in the end small bits of plastic started coming off it here and there, but it still worked and it did come with restore media.

So arriving in the office this morning with a sligtly heavier backpack than usual, I plugged the machine in along with the others, turned it on and after a bit of fiddling with firmware menus got it to boot from the restore CD. (That is, to actually get the thing to boot the CD, you need to watch for the right moment to press Enter, I almost forgot.) The installer takes a while to load, and if we forget about the 'you have to agree to this EULA' screen it's on par with FreeBSD's sysinstall for intuitiveness or lack of it.

The first thing that really amazed me was how long it took to create a file system on the disk. The last system on the machine pre-windows restore was OpenBSD, so of course creating a new partition and file system was necessary. Past lunch the thing was still only in the early stages of copying system files.

Then for a while it just stood there with a dark screen making CD crunching noises, so I power cycled the machine. It came up again with a the traditional teletubby background and a dialog which demanded to see the restore disk for a while more. I gave it that, and it went on for a while, finally rebooting. Over the next few reboots the system consistently tried (but failed) to find the correct resolution for the internal display (1200x800), trying 800x600 alternately and 1024x768, never finding the physical resolution at all.

Any X I've encountered just magically found the optimal configuration for this kit; for the first few months I had the machine it ran fine without an xorg.conf. After I started taking it to lecture halls to speak to projectors it turned out I needed one to fiddle with, but finding something that worked on the interal screen was never ever a problem with any freenix and X. But then of course I guess Windows had to be different. By the time I had XP's Service Pack 2 installed, it was 3.30 pm.

What would you do - go on mucking about with the Windows machine and try to make it behave, or tell the customer "if you need me to go there with something that speaks wifi and I can use to take notes, I can bring my OpenBSD Thinkpad"? It runs X with KDE, so I won't look too scary I guess.

Bah. It's late in the day, SWMBO just called in for supplies, and I have other writing to do. In the meantime, I could see how far installing that FreeBSD-current snapshot on it gets.

The book, it's still progressing.

Sunday, June 3, 2007

Summer's here and the time is right ...

... for working on the book. At long last, summer has found Bergen, with temperatures in the twenties C and only occasional rain. Great for heavy duty interaction with my laptop and various network gear I need to write about.

BSDCan was excellent (with total time spent on Canada Customs in excess of four minutes this time), lots of good material presented such as Michael W. Lucas' Netflow on the cheap tutorial.
The conference has turned into THE place to meet BSD people - I have a longish list of people to contact over various things (book related and others). Unfortunately with the Calgary hackathon due to start only about a week later, there were fewer OpenBSDers attending this year than last.

My two sessions both had something like thirty people in them each. The tutorial was well received and I think well executed; for the malware paper presentation my slides were still not good enough and I got a little too distracted by the microphone's oddities. Mostly people who read the paper think it worthwhile, what puzzles me is that some of the feedback I get is extremely hostile to any mention of spamd. Maybe it's the mention of my all time high (almost 12 hours) for having a spammer hang around to never finish delivering the message after all.

LinuxTag is past too. I ended up not going for various practical reasons, and I certainly hope Stephan's 1-hour version of the PF tutorial went well. He told me he'd cut out the spamd part entirely, but I still think he would need to talk fast.

Order of the day: Beat that chapter into shape, then prepare for a backyard barbecue.