Sadly arriving two days late for wearing to FOSDEM, my trial-run of Sound Juicer t-shirts arrived on Monday. They rock! The back reads freshly squeezed audio if you can't make it out. Go and get yours now! NP: Ben Folds Five, Ben Folds Five

I haven’t been blogging lately. I do have a running list of ideas and topics, but my blog is going through a bit of an identity crisis. When I was a law student, it was easy for this to be a ‘law student blog,’ but now that I’m practicing I haven’t figured out how to find a balance between the personal and professional (and, perhaps on a third axis, the technical). Also, since I had a daughter, my time has been pretty scarce and I’m not sure that blogging is the best use of it. I’ll probably pick up in frequency before too long, but in case anyone is wondering, I’m still here. In the meantime, my Webloyalty is a Scam entry continues to get more traffic than anything else on this blog—close to 31,000 hits at the time of this writing, and nearly 1,000 writebacks (plus nearly as many personal emails to me on the topic). See also this follow-up post. One guy has even started a blog dedicated solely to criticizing Reservation Rewards—but now the link seems to broken. When I write these consumer protection entries, I inevitably get comments from low-level company insiders defending their practices (see, e.g., U-Haul Responds). These writebacks are often a much better condemnation than I could ever come with. For example, the following was apparently written by a “WLI Call Center Rep.” The POST request came from the same area of Connecticut as where Webloyalty is located, and it seems pretty genuine:

I’m a rep in the call center at Webloyalty. I have no pity for any of you people. You’re all suckers, plain and simple. Didn’t anybody ever teach you that nothing is free? When all of you idiots made your purchases on whatever website you did business on at the end of your transaction there was an offer asking if you wanted to save $10, or get award miles, or whatever. When you clicked on that link you were not automatically signed up. What happens is that you’re brought to the Reservations Rewards website. On that website it tells you that you are on the website for Reservations Rewards. You see, you can tell that because the banner at the top of the site says “Reservations Rewards”. Unfortunately you were too stupid to notice or remember. It then gives you instructions on how to redeem your “reward”. At that point you are instructed to enter you email address in twice and click accept. Now you have to manually type in the email twice in those boxes. No cut and paste is allowed. Then it tells you to click accept. Now, if you had any fucking brains in your head you would have noticed that right above the box where you enter your email mail address its says, in regular sized type, in plain sight, right out in the open, that entering your email twice will act as an electronic signature and that by clicking accept you are accepting that the website you just made a purchase on can share the billing information with Reservation Rewards. Also in the big box next all of this it gives the exact details of what you are signing up for, again in regular sized print, in plain sight, right out in the open. If you are too stupid to take the time to notice all of that then you deserve what you got. which was a membership in a overpriced bullshit ptogram. Webloyalty depends on idiots like you to not notice this stuff. To be blinded by the idea that you are getting something for “free”. To not look at you credit card statement so charges go through every month. Its unbelievable the amount of dummies out there that fall for this stuff. Even if you do catch the charges eventually, and get a refund they still made money off of you by collecting interest on your money when they had possesion of it. Stay a member or cancel its win-win for webloyalty. So, anyway I hope all you dummies learned a valuable lesson and wont fall for this again. I’m sure many of you will, though. You’d be shocked at how many people are repeat members where they canceled the service a while back but fell for the scam again a few months later.

This past couple of weeks saw a lot of Debian and Ubuntu work from me in (hopefully) equal measure. For Debian, I whipped up new versions of robotour, xshisen, libmemcache, and gtklp; robotour finally installs in /usr/games as it should be; xshisen acks NMUs and finally has a sane source package (well, a little bit tipsy still, but at least it’s now fixes the NMU source b0rkage;) libmemcache’s -dev now installs its headers to the right location; and gtklp bumps to a new upstream, with touch-fu replacing AM_MAINTAINER_MODE for seamless build. Thanks a lot to my wonderful sponsors Margarita Manterola, Sylvain Le Gall, and Andreas Metzler for the uploads! On the other hand, opendchub has migrated to testing, and its initscript failed (not to mention missing a manpage) I’ve fixed the latter, but I still have to figure out how to fix the former; I’ll probably have to consult the reference and add an unprivileged opendchub system user and configure the daemon to use that, or fix the source and allow flexible redefinition of opendchub’s working/configuration directories (currently hardcoded to $HOME/.opendchub). Either way, I have to touch the source and make patches; and I would appreciate it very much if my dearest readers can help out On the lighter (or heavier, depending on your POV) side, I noticed that Steve McIntyre posted an RFH for cvs, and I’ve decided to help out a bit. Seeing the source, I was pleasantly surprised as the age (and design) of this package, and it would be a blast for me to study this (and fix bugs, prepare patches, and release new versions while at it. It would also prepare me better for NM as well. As for my Ubuntu work, I’ve upgraded my Breezy Badger desktop to Dapper Drake last Monday (just in time for Flight 4,) and I’m quite amazed at the new features of this release. Like many folks I just loved xchat-gnome’s notification feature (especially when you’re moving in between workspaces very often) and was very much pleased with the speed of Gnome 2.13.91 (so much so that I’ve departed from my normal Ion3 desktop without regretting it. But of course, since Dapper Drake is the latest development branch, there were some snags that I got to encounter. One particular bug was with masqmail, because the 0.2.21-1ubuntu1 version created a /var/run/masqmail directory, but since /var/run is mounted as a tmpfs in Dapper, masqmail will complain the next time it restarts. Another (probably cosmetic) bug(?) was with tuxpaint, since it placed a couple of launchers in the Gnome Applications menu, one on Education and the other on Graphics. I’ve yet to see if tuxpaint’s is a genuine bug or just a design decision, but I’ve fixed masqmail’s, and 0.2.21-1ubuntu2 should be on Dapper now, thanks to Daniel Holbach. So far I’m loving the new Ubuntu release, and with FeatureFreeze finally here that means I’ll have to drag my desktop box again to the local internet cafe to do yet another aptitude dist-upgrade. Gaah, I need a laptop!

Each morning, CodeCon started later and later. This is not just because us volunteers couldn't get there in time, but also because the attendees would show up more hungover and sleep-deprived as well. Thankfully, it's a hacker conference so we started at about noon. Or 12:30. Which is "morning" for some. We kicked off the day with a talk on Dido which is meant to help you write voice-menu systems. Quinn did a neat demonstration involving a real telephone, which was quite novel. However, since I despise voice-menu systems, I have to hate his project on principle. In the Q&A; period, I nipped over to Bill Pollock's booth. He runs No Starch Press which is absolutely fabulous. He gave me a free mug, some tea, a book, and a T-shirt. In exchange, I had to buy a few books as well. I'm afraid that my book buying habit is starting to get uncontrollable. Every time I return from a trip, I bring five or ten books back. Now I have a copy of madduck's book which I will get him to sign at DebConf 6. I completely zoned out on the Deme presentation. I think I was talking to someone who wanted to convert me to Evil, after hearing that I was a hacker. I'm afraid that I'm too noble for that. But what I did glean from the presentation is that deme is an actual English word. My presentation on UniConf came up next, so I put on my red "Speaker" badge and hopped on to stage. You see, I was at CodeCon as a backup speaker, in case one of the others missed a flight or fell down some stairs. I had to resist doing any of them bodily harm. For my altruism, I gave a talk about how UniConf glues configuration systems together, and why people should use it instead of writing their own. I got a lot of questions afterwards, which was pretty cool. At lunch, wealhtheow and Bill came by to ask me what I'd like on my pizza. Wow! I said that it didn't matter, and thought little of it. Then they showed up with boxes and boxes from the local Costco. Bill had just bought everyone lunch! We organised a queue that would have made the British proud and everyone got a slice to eat. I remember first hearing about Monotone from Graydon, back at GCC Summit. So I was a little surprised to see Graydon sitting on stage looking pretty while Nathaniel was presenting. Monotone has inspired a couple of version control systems but it's never really gotten popular. But one thing it hasn't passed on is its use of Merkle trees to express differences. All throughout the conference, I had been selling drinks from behind the bar. Standing for all that time really hurts the feet, so I was sadden to discover that my shoes were not good for this. Still, I had a good time there. I've definitely found a calling in the service industry, if the computer bubble bursts again, I'd make a great bartender. I never knew! You make eye contact, smile and say something friendly. Then you chat as you pour a drink, take an interest in someone, and just be nice. If you give a little extra, smile and wink, like it's secret just between you. I was busy chatting to my customers for the Rhizome presentation. Reading its description, I'm not really sure that I could tell you what it is. And I work with computers! As I was selling drinks, I was also selling CodeCon T-shirts. Very snazzy T-shirts. Every time I sold a drink, I tried to sell a T-shirt along with it. Show off the front, show off the back, offer an appropriately sized shirt. Flattery seems to work very well, as does smiling when the person looks at it. You know they're sold when they're holding a shirt, so you offer to fold it for them and name the price. It was so easy! Daniel came up again to do his Cqual++ presentation. Not only was this piece of software more difficult to write than delta, it's also less popular. Alas! He does some very impressive static dataflow analysis on C++ to find nasty, horrible bugs. But parsing out C++ is scarily difficult! After the presentations were over, we had a very nice set of closing remarks which made all of us blush. And then we tore down all the stuff we put up and packed everything away. This is where I packed my own T-shirt in a box by accident. Jonathan's going to send it to me by post, which is very nice of him. Then I bid sad farewells to gorgeous Wealh eow and charming akashayi who went off to dinner without me. Six of us climbed into a taxi and we went to a sushi restaurant called Osaka, where we met up with breezip. Bree was the friend that the lovely and indefatigable spider88 kept trying to meet, but failed to. We were all quite exhausted from socialising with people since, after all, we are a group of introverts. So we enjoyed an excellent dinner with subdued, but fascinating, conversation. I was treated to the best uni I have tasted in my entire life. I think Bree was quite amused at my sea-urchin ecstacy. After dinner, Len paid with a thick wad of cash, which in conjunction with his suit, made him look like a mob boss. I had a very good night.

Last night I uploaded the last major pieces of Xorg 7.0 to experimental, including the server itself. Today I came home and found that Joerg Jaspert had already approved them to go in. Big thanks for him for processing these things in record time. The packages currently only have a few of our Debian-specific fixes applied, and the next major step before uploading to unstable will be to get those ported from 6.9 to 7.0, and also to pull in patches like benh's radeon patch that fixes a lot of crashes in 6.9. As things stand now though, you can pull the packages from experimental for i386 and be running Xorg 7.0 in all its modularized glory. Hopefully a few of you version junkies will try them out and report bugs so we can hammer them in to shape for release!

I wanted to give another note of thanks to Eugene Konev. Eugene helped out quite a bit in the past, getting 6.8 and 6.9 in good shape for Debian. For 7.0, I had done a pretty poor job on the xfonts and he put in a ton of effort to polish them up. In addition, he packaged xdm and even ported all the patches for monolith over to that. He really helped me get this last bit over the hill, and he deserves a lot of credit for it.

I'm feeling pretty good about this because it represents the end of about a year's worth of work. I won't feel like it's the end until a set of packages I'm happy with ends up in testing, but until recently I wasn't sure I'd be able to get this far in time for etch. Yet somehow, here I am, looking like we'll definitely be shipping modular Xorg with etch. There's still this last 10% to go through though (and you know what they say about that) but I'm confident that the end of this task is in sight and I can start to move Debian's X work from playing catch up to helping define the cutting edge. Hopefully with everything going on people will be getting excited about X again, get involved, and we can keep making cool things happen.

Xulrunner finally reached unstable on sunday, and it already needed some adjustments. I forgot a conflict, misnamed the libsmjs-dev package, the hppa assembler code got somehow duplicated, so xulrunner failed to build on hppa, and some preference files needed to be moved around the packages. It also fails to build on alpha, but that’s not my fault. I also changed the way we identify a Debian built xulrunner in the user agent. I replaced the "Gecko/yyyymmdd" string with "Gecko/Debian/x.y.z.t-r". That has several advantages over the previous "Gecko/yyyymmdd Debian/x.y.z.t-r" string:

• It shortens the already long user agent string so that additions such as product name are not painful,
• removes pointless information (the date in the original string indicates the date of the build, not that of the API),
• keeps the "Gecko" string (which some site might want, seing how Apple and Konqueror did put a "like Gecko" string),
• and finally avoid confusion with other Debian release informations that may be present in the product specific part (I think Galeon puts one, for instance).

That needed 2 updates, because I realized in between that while you can set the general.useragent.product and general.usergagent.productSub variables, they are not actually taken into account until you change them a first time… The HTTP protocol initialization in xulrunner forces the respective values Gecko and yyyymmdd… I just completely removed that part of the code in xulrunner. I took advantage of this needed fix to work even more on the package, enabling the typeaheadfind module, needed by galeon and epiphany, removing some old perl code that does nothing except producing useless errors during the build, and, last but not the least, enabling the “flat style” chrome, meaning that instead of being stuck in .jar files, everything is just here in a standard tree. It appears upstream provides a way to build such trees (option --enable-chrome-format=flat to configure), but it’s useless because it doesn’t install the trees when running make install. In addition to that, the configure script still requires zip, even if it doesn’t use it, except for some obscure .jar files in libnss, that are not installed either. So after disabling the build of these useless files and removing the zip strict requirement if we build flat chrome, and fixing the script responsible for installation of the chrome files, everything got much better. These changes have been applied to firefox as well, so stay tuned, it’s for next upload. Having highjacked spidermonkey, I did some bug triage there as well. It appeared they could all be closed by the fact libmozjs0d is there. I still need to make some triage in my patches to xulrunner and send upstream those that I still have not sent, if they are of any interest there. Apart from this xulrunner stuff, I also did some bug clean-up on libxml2 and libxslt, which I’ve been somehow neglecting for some time.

As I told before, I’ve been working for a few days on a way to display japanese characters on a web page when you don’t have japanese fonts locally. That involved taking japanese characters from a font file, and turning it into a png image. So I just wrote a small and simple program using libpng and freetype. Which works quite well. But that’s not the point. During this “development”, I lost 2 or 3 days trying to understand what was wrong with my code that made it work properly from a shell command and fail halfway when wrapped in a php script. I still don’t know what’s wrong with php, but it works well with a perl cgi wrapper, and it still fails if I raise the php memory limit. Interestingly, other libpng tools (such as pngtopnm) fail in the same way. It wouldn’t surprise me if it was libpng’s fault. The first thing i did to understand why it’d fail, was to run the program under valgrind. And what I saw was not encouraging. A lot of “Conditional jump or move depends on uninitialised value(s)”. I tried different combinations of code, using png_write_image or png_write_row, with large or small buffers… It turns out libpng can’t write a png file with a width smaller than 148 pixels without those “Conditional jump or move depends on uninitialised value(s)”, which happen, as the stack trace shows, in the zlib, called by png_write_row, but i doubt it to be a zlib issue. You would think it would be a buffer size issue, but if you use a larger buffer for each line you call png_write_row for, you still get the errors. If anyone has a clue, please leave a comment. Update: Thanks Mark for your comment, though I don’t get how it can be safe to have such accesses to uninitialized memory. Deuce, the php issue can’t be a timeout issue, the program runs in fractions of a second.

I've bought a USB keyboard today. The one with 88 keys.It's called M-audio keystation 88es, full-sized lightly-weighted keys.Auto-detected with ALSA usb-audio driver, and *it just worked*.Nice.

(This is a copy of my response to the LWN article Toward a free metaverse, which won't be freely available itself for another week or so.) There's an unconcious bias in this article toward 3d graphical worlds, and I think it's important to realise that that's not a necessary component of a virtual world, although it's certianly a component that will appeal to a lot of people and that fits in nicely with some current ways of working with computers -- ie, sitting in front of expensive screens weilding a mouse. Long ago we had MOOs, MUDs, etc, and they were text based. Just words, you know. And nearly all of them were free software. It might well be that text, or speech, or the like ends up being a better interface to virtual worlds of the future, especially if they are tightly integrated and layered on top of the real world. And I'm increasingly convinced that if these worlds don't integrate into the real world, if the ideas that have been bubbling up out of ADVENTURE and the MUCKs and the MMORPGs and such don't spread out and become something larger and more inclusive than their own little pretend world, then in the end that whole area has been nothing but pointless games. Which I hope it hasn't, but perhaps I'm getting old.. Speaking of freedom, there are levels and layers, and it's possible to build the free on top of the non-free. After all, I'm typing this on a device that is present in the er, First Life, which is not particularly free in its implementation; it has non-free firmware and BIOS, and is composed of atoms of some substances that are less easily available than air, and which follow some annoyingly inflexible physical laws, and it needs a constant input of metered electricity to run. But on top of all that is some free code that gives me a measure of freedom. Similarly, it would be possible to run free code in Second Life, and given how rare it is overall for such services to support coding at all, let alone explcitly give a program (or object's) creator the freedom to make it Free, that Second Life does so seems encouraging to me. If we get a distributed standardized free online world by 2010, I'll be happy indeed. (Disclaimer: I write text based MOOS, and have never used Second Life, though I chatted with someone there on the phone today.)

How can you go wrong with this headline: Physics students, duct tape extend faded couch's mileage Yes, that's right, by using a lawn mower engine and duct tape, they have motorized a couch.

Tonight Sarah and I went to see Cirque du Soleil's performance of Corteo at San Jose. It was excellent. Really really good. Words cannot do justice to the description of it. We were both amazed and awestruck at the performers' gymnastic abilities. The music and the song were excellent as well. Money well spent on good seats.

Last Monday, Sarah and I visited the fine institution that is the California Department of Motor Vehicles, so that we could get ourselves California drivers licences. We had to do both a written test and a behind the wheel test. The written test was your standard multiple-guess exam (36 questions, maximum of 6 incorrect answers to pass). Annoyingly, all the material on the exam was not in the California Driver Handbook, which the handbook tried to make out was the case. Fortunately it was largely a case of common sense. I got two wrong. The funniest thing was that Sarah and I both got the same exam paper, and Sarah did her test just after I finished mine, and we were allowed to keep the paper, so she could see which two I got wrong. One of the more unusual things I found about the process was that they allocated me a licence number before I'd even sat the test. Last Thursday Sarah had her behind the wheel test, and today I had mine. It was pretty easy. No reverse-parking, no emergency stopping. I just had to drive around for 15 minutes. I passed without any faults. Given it took me four attempts to get my original licence in Australia, I was pretty pleased with this. (I failed my Queensland licence a number of times due to nerves). The other thing that I was amazed to discover was that I don't get the licence for another four to six weeks! They mail it out. The DMV processes over here are just mind boggling. The fact that you can't get a licence on the spot, and that it takes something like three months to get licence plates for a new car just blows me away. It's nothing like the process in Australia.

If you’re having trouble with a natsemi ethernet controller locking up in Linux after complaining about oversized ethernet frames like this:

eth0: Oversized(?) Ethernet frame spanned multiple buffers, entry 0x00ba8b status 0xe0000bd5.

then this patch (or a kernel version which includes it) may be useful. The patch is against 2.6.9+ versions of natsemi with NAPI support but doesn’t use the NAPI support.

So my sister is working at a school where the computers are in a shocking state. All infected with viruses, spyware and malware. The machines either crawl along or are incredibly unstable. Some of them run some sort of virus protection program and I’ve seen one or two computers running two different anti-virus programs. On my travels I have seen countless infected Windows machines. It’s incredible! And the worse thing about it, is that the Free Software geeks and outfits can’t seem to capitalise enough on it. So what can I do? The Computer Scientist. Well, not a lot. For example all the Windows machines are in Korean. I can’t figure out how to change win32 language. Can you? Plus, I don’t want to re-install Windows. I want to deploy some Linux based solution. So what are the requirements? They need to :

• print timetables (via a crappy Epson inkjet) with something like Office
• be able to play macromedia based games
• use a browser

That’s it. It also can’t be more complex. No username or passwords please. We don’t want to have to train people to use a new system. Must be simple. Oh diddums. Can Director macromedia kid games be played in Linux? I don’t think so. Oh I wish these kids’ software was on a N64 or something. There is only like 10 PCs in this school and installing them with Linux would take quite a bit of my time. I need some sort of centralised control tower where I can control deployment of my own customized Ubuntu based distro. I know very well about SkoleLinux but it is simply too complicated. No need for email. Outsource to Gmail and hosted service. Must be simple. All I can think of will be complex and painful. Grrr…

Dear Lazyweb, Anyone aware of a nice blog software that

• Uses some kind of proper XML to store all data (and thus always create valid output,
• Allows me to enter stuff using some StructuredText human-writeable markup and copy HTML fragements from other pages in there.
• Does reasonable caching instead of doing XSLT transforms over and over again...
• Allows me to tag posts, and provide customized RSS feeds (e.g. all english, non-personal, XML-related posts
• Doesn't force me to use the web to blog (like WordPress)
• Doesn't use PHP or ASP or JSP and a MySQL database which I consider bloated and potentially insecure - should be nice and clean code, which PHP rarely ever is.
• Doesn't use extensive frameworks such as Ruby on Rails or TurboGears either... I don't want to have to install these.
• Doesn't bring with it another stupid user management (like WordPress and all these fancy php thingys do)
• Preferrably doesn't require CGI at all...
• Doesn't need to have comments, and lets me disable them. It's my blog, I don't want to have to care about spam and you are supposed to read it using a RSS reader (and no UI-hell Internet Exploder 7), actually. :-)
• Is OpenSource, obviously. I will certainly want to remove some things I consider stupid. I've yet to see a commercial software I'm happy with.

And I bet I forgot some additional restrictions... Thanks for any suggestions..

Recently there was a discussion on whether BitTorrent should get some encryption (or obfuscation, more likely) added. Right now, you can filter out BitTorrent traffic with the following iptables command (note: this needs iptables 1.3.4 and a recent kernel):

iptables -I OUTPUT -j DROP -m string --string "BitTorrent protocol" --algo bm --from 0 --to 100

(Note: match is not optimized, the range of 0 to 100 and algo bm is arbitrary) The nice thing is, that this match is very unlikely to match non-Bittorrent traffic. Linux users can use it to filter out all BitTorrent traffic or mark it for speed restrictions. Network operators are very interested in filtering this traffic: it slows down more important traffic (and if you are e.g. using stochastic fair queing, the large number of connections for P2P filesharing will make it get a larger share than other apps!) and can cause costs. If you have mobile users, who e.g. connect their laptop to your network, they will occasionally use BitTorrent at home, and forget to disable it when they connect to your network. That's really bad, it happens and it's not like this is intentional "malbehaviour" or "sabotage"! So back to the original topic - obfuscating BitTorrent traffic. Let me explain what network operators might do when BitTorrent can't be slowed or filtered any more (Note that this already applies to some services such as Skype or Kazaa as far as I know): First of all, they might decide to filter all connections to non-standard ports. This effectively eliminates (most of) filesharing - but also e.g. Skype and some IM and chat services. Basically any non-standard service. (Note that "professional" VoIP services using SIP may still be whitelisted). Secondly, they could decide to slow down all connections to non-standard ports. This would make filesharing, Skype, IM etc. still possible, but for example file transfers via IM will be really slow. No video via Skype, probably, depending on the limits set in traffic shaping. Thirdly, they could disable all outgoing connections and require you to use proxies for that. Then you need to enable/disable your proxy all the time when switching networks, and you'll only be able to use a limited set of services. Again, Telephony via SIP is still technically possible, using a SIP proxy or by whitelisteing major SIP providers. Skype is no longer possible. IM needs whitelisting, and will likely be blocked, too. None of these options are very favourable for the users. So please do not support "obfuscated" Protocols such as Kazaa or Skype, or an obfuscated variant of BitTorrent. Stick with stuff administrators can set up policies for - and be nice to your admins. Tunneling via non-standard ports, hiding services will only get you into trouble, and probably even make innocent other users suffer. After all, filesharing - especially illegal downloading movies and music - is not THAT important, is it? So please make the important services not suffer from it. Thank you. Oh, and please use SIP, not Skype. Because your admin can whitelist your SIP provider if you tell him you need it for work or setup a proxy for that - and with Skype, he might be afraid that you become a supernode and cause high traffic costs. And then there is a much larger software choice for SIP, including many OpenSource solutions, including PBX software such as Asterisk.

The latest burning of different consulates by outrageous and uncivilized muslims only shows again one thing: they ask for respect for his way of thinking while at the same time they don't respect at all others way of thinking. Clearly, islamic movement is being used lately as comunism was used some time ago in Europe, as a way of control people and make them not thinking freely. Watching those images on TV makes me change my mind on respect to the islamic world. And while I think that perhaps those Mohammed cartoons could be offensive for muslims, their reaction makes me think that they deserve those. But what can you say of people that behave as the year they live in: 1441.

* Temperature in Spain raised after the cold wave (mild cold wave for cold wave values anyway).
* Upgraded my Linksys running OpenWRT to White Russian RC4. This toy makes me so happy.
* Upgraded my 3Com OfficeConnect 11g to 2.06, to deal with the router being Dos by using p2p software (even though it is just a acting as an ATM bridge). This toy makes me so unhappy. I miss my 3Com OfficeConnect with no nothing that did nothing but what it was told to (a simple ATM bridge). But this old friend does not support my new DSL 2 connection. His loyalty and rock solid operation for over 4 years 24x7 non-stop will be missed.
* After all the upgrades were done, I realized that the reason my host was unreachable from the Internet was that my fixed IP address (that I pay 12 EUR a month for) was changed to IP 1. Lovely!
* My old skates were ABEC1. If you know anything about skating, you know the fucking thing does not move, which was good while I learned. I felt a bit more adventurous, and wanted to go faster! Today I found the 150 EUR skates I dreamed of for 79 EUR. God bless Credit Cards and January sales. I am now the proud owner of a new pair of K2 Exo 6. ABEC 5, aluminum-titanium alloy frame... a dream. I just hope not to break a leg in these. Wish me luck!


Yay!

Finding arm machines is not hard. In fact, everyone of you probably owns some of them. However, finding something suitable for Debian infrastructure usage is harder. For example Buildd load pattern requires not only a decent CPU, but above all good IO performance. Most ARM systems do not have a hard drive, and if they do like on personal media players like Archos pma400, the drives are slow and optimized to minimize power consumption.

Therefor, the logical place to look for machines would the NAS devices, like the already-being-Debianized Linksys nslu-2. NAS machines tend to (relatively) cheap, well available and need good IO performance. The main problem with NAS devices is the lack of RAM (to minimize costs they tend to have 32MB or something soldered to mainboard), so I was more than happy to spot a NAS with 256MB of ram:

Linksys NSLU2	Allnet ALL6500
266Mhz* ixp420	400/600Mhz IOP80219
32MB ram	256MB ram
USB 2.0	2xSATA & USB 2.0
100Mbit ethernet	2x 1000Mbit ethernet
95	350

*overclocked

The real shocker came when reading "Manual englisch": (arrows by me)

This bloody thing seems to have a DIMM socket!

GPL sources seem to be well available. Can someone from .de confirm if the memory is really there? (Why does so many cool hardware never end OEM'ed this north?) IOP80219 supports up to 1GB so it really seems quite promising :)