Christian: of course you will be running the Chicago 2010 marathon as your next fall marathon. Not only are Paris and Chicago jumel e (i.e. sister cities), but the Dollar is so conveniently falling vis- -vis the Euro that you'd get all this done for about a Euro and a half by the time the race happens. Moreover, 2010 is going to be Debian's US year anyway. I crammed my two European races into two four day trips which I won't exactly recommend, but you do not need ten days either. Lastly, I cannot promise you many interesting cheeses but there are at least a few pretty tasty beers to sample. bient t Chicago!

Photography is all about being in the right place, at the right time

Freaking out the freaks on Omegle.

8 GB of storage will more than suffice for a nice and clean Debian installation ;-) If you live in germany, i absolutely recommend http://www.pulster.de/ for all free phone accessories!

I wasn't going to write about this since I thought that Chris's post covered pretty much everything I would have said, but after reading Scott's entry on how people would have interpreted Mark's remarks differently if he'd said "We'll have less trouble explaining to boys what we actually do" instead I realised that people are still confused about the fundamental issue here.

The assumption that Scott's making is that "girls" and "boys" are semantically equivalent in this case. They're not. There's various ways in which the symmetry is broken, but the most basic one is that Mark's a straight man. When the overwhelming stereotype is that "we" as a community are heterosexual males, using "we" as a shorthand for "People who are straight men" is unfortunate because it supports that stereotype. Using "we" as a shorthand for "People who are attracted to men" doesn't. Unsurprisingly, this results in a fairly significant change in who's going to be offended.

Whatever his intentions (and I could easily believe that it was a slip of the tongue), Mark managed to imply that the Linux community is entirely made up of straight men. This is possible because straight men do make up the majority of the Linux community. In contrast, Scott's version doesn't succeed in implying that everyone in the Linux community is attracted to men because it's blatantly obviously not the case, so we know that Scott is using "we" in a different manner. Context is important, and unless you can invert everything else about the situation as well then simply replacing the word "girls" with "boys" doesn't give you any meaningful insight into whether or not people are justifiably offended.

In a more general sense, I'm saddened by this case because I think it's a clear case where the Ubuntu code of conduct could have been used to good effect. "Be excellent to each other"[1] ought to include accepting that you've offended other people without meaning to and making appropriate restitution. If the offence was unintended, an apology should be cheap. Whatever the reality of the situation, failing to provide that apology gives people the impression that either the offence was intended or that Mark doesn't care about those who were offended. That's not a good way to build an inclusive community.

[1] Mako's original summary of the code of conduct

So, now that I completed my 2009 program by running the Cologne marathon 6 months after my first one in Paris, I now plan the 2010 "season"..:-) (one point for you guys who told me that once you start you don't stop) The next step is definitely Paris. The race is close to my home, this is a great race in a great city at a great time. I really can't miss it. So, that one is registered. The point is "what about 'the other one'". I plan to run another marathon during that year, preferably during autumn. Actually, I would probably be able to run at least one more but, well, many people urge me to avoid exxagerating and that includes my beloved wife..:-) So, one more. Which one? At this very moment, I have about 4 options:

• Cologne again. Pros: this is a nice race, human-size, in a welcoming place, also an oppotunity to run again with local friends. Cons: what the hell would we do in Cologne during the few days we would stay there, now that we've seen most of the city (one cathedral, 12 roman churches)...and that would mean the same program than 2009
• Berlin. Pros: the top german marathon, one of the "Big Ones", in a city that deserves a few days sightseeing and which Elizabeth really wants us to visit. Cons: a little bit early in September...and Germany again..:-)
• Donostia (San Sebastian). Pros: this is an open invitation by a relative of my brother in law, this is in Basque country which I enjoy, probably with a quite warm crowd. Cons: late in November and I fear the hills to be quite hard over there
• Chicago. Pros: would be a great opportunity to met with Dirk and even though we would not run together, I would enjoy this. Also a nice place to visit and p ssible to arrange a visit to my wife's family in Kansas City neighbourhood. Cons: arrangign a 10-days trip tp USA in early October is very very hard to fit with our professionnal schedule...and fairly expensive...:)

So, at this moment, Berlin and Donostia have the lead. A good compromise could be running a marathon in Berlin and the 20km Behobia-San Sebastian run which is very popular in Basque country. We'll see..

I have not been a frequent fixture on my own blog as I am writing what is called an Annual Review of Anthropology on digital media and ethnography. Truth be told it is killing me as there is a 6000 word limit and 100-150 works one must mention and entertain (usually by throwing in some categorical statement that makes sense for 10+ works). One thing is clear: the literature on digital media by anthropologists is switching from trickle to steady and very interesting stream. Even if I Epically Fail, I have already learned a lot, which is what I keep telling myself as I struggle through the writing stage of the article. But if you want a taste of some recent work, there are some blog entries you can check out: Daniel Miller who was one of the first anthropologists to venture in this area (and kick-started the first program in digital anthro) at UCL has written a nice review of various books recently published. And for the same blog, I wrote an overview
of my work on hacking, liberalism, and pleasure. So if you want a short introduction to the books being published by anthropologists on digital media, I highly recommend checking Daniel Miller s post.

I have not been a frequent fixture on my own blog as I am writing what is called an Annual Review of Anthropology on digital media and ethnography. Truth be told it is killing me as there is a 6000 word limit and 100-150 works one must mention and entertain (usually by throwing in some categorical statement that makes sense for 10+ works). One thing is clear: the literature on digital media by anthropologists is switching from trickle to steady and very interesting stream. Even if I Epically Fail, I have already learned a lot, which is what I keep telling myself as I struggle through the writing stage of the article. But if you want a taste of some recent work, there are some blog entries you can check out: Daniel Miller who was one of the first anthropologists to venture in this area (and kick-started the first program in digital anthro) at UCL has written a nice review of various books recently published. And for the same blog, I wrote an overview
of my work on hacking, liberalism, and pleasure. So if you want a short introduction to the books being published by anthropologists on digital media, I highly recommend checking Daniel Miller s post.

Earlier today the 33rd Chicago Marathon was held. I will organise my comments by pointing out three epic fails. First, of course, was the Chicago weather. After two successive marathon in excessive heat --- the 2007 race I completed, poorly, with its thousands of runners forced to abandon when the race was cancelled due to excessive heat, and the 2008 version which I skipped as I ran Berlin that year just weeks before Chicago, this year had forecasts of temperatures in the thirties and possibly snow the night before. Well, the weather cleared up -- but with those clear skies we still got a severe weather alert for the area due to frost! So at the (now earlier by 30 minutes) race start, it was very nippy and in the higher 30s, improving steadily under sunny skies. Overall, a little chilly and hence not exactly ideal, but not too bad in the grand scheme of things. A little windy coming up Michigan Avenue. But hey, Wanjiru still finished with a course record though still well off the world record. So maybe not a fail after all. The second fail, though, was my GPS which I had just blogged about yesterday (albeit indirectly). I had had my issues with the previous (much simpler and older) Forerunner 205 which lost satellite tracking when running downtown amidst the skyscrapers. I had high hopes that this newer model would do better. But no, not only did it loose track within the first few miles, it even managed to outdo the older model by turning its GPS tracking off in the later stages. Now that's a fail! The third and final fail, unfortunately, concerns my run. Training had gone well enough to let me hope for another decent race. However, a latent cold during the last two weeks had left me somewhat afraid I might not do well. Things felt better yesterday, and I ended up running a decent first eighteen miles to the waterstop the rest of my household was working. And thereafter: well let's just say that the wheels came off. I ended up with a 3:25:40 (or around 7:51 min/mile and 4:53 min/km) which is not terrible but also not one of my better races. I updated the two race data geekery charts shown here before to illustrate the fail. As I am running out of shades of blue for Chicago, I show the 2009 race in purple. A decent, flat chat indicating a reasonably steady pace throughout most of the race ... followed by one of the very worst finished. And as discussed above, I cannot really blame the weather either. I also updated my 'performance by race type through time' chart which suggests that I may be getting slower for marathons (if we allow for a non-linear effect). Eek.

Angie has my old laptop that some might remember from a couple Ubuntu developer summits. Adam Conrad used to refer to it as the surfboard: a lovely 15" widescreen laptop. It's my test machine for Ubuntu updates and so far I'm loving Karmic. This machine has been continually upgraded since either Hoary or Breezy without a reinstall.

Jaunty was miserable. Bugs in the Intel drivers meant that DVD playing stopped working, sounds would occasionally lock up, and things were just painfully slow. Karmic to the rescue. I did the update at beta time like I usually do, and without any further effort:

* Flash just worked.
* DVD playing in totem just worked.
* Boot time was slightly quicker
* nethack-gnome continues to work (despite being removed from the archive a few releases ago)
* Suspend/resume continues to just work.
* Atheros wireless continues to just work with the free kernel drivers.

I then added the dev channel of Google Chrome, which imported all the Firefox bookmarks and started to remove some optional things that Angie doesn't use anymore and don't run well on the now 5 year-old laptop. I took out openoffice, a pile of -dev packages that were leftover from when I used to build glibc on this thing (This laptop had the Hurd running on it at one point... *g*), and then discovered the System->Admin->Package Cleanup (Or at least, it's called something like that, the laptop is in French) where I could start to remove unused packages. I did a few rounds of that, some apt-get autoremove. I took universe and multiverse out of the packages list and did some more removals, making sure to leave in things like deborphan and the gstreamer packages that are in universe and multiverse. Lastly, I did a round of deborphan to take out some pieces that were just hanging out.

As a final step I've updated to grub2. I don't have a separate /boot partition on this machine, so I haven't been brave enough to convert it to ext4 yet.

The boottime at this point still feels a little long for what this laptop does. While at Canonical, I played with some appliance configs and had from kernel load to running firefox in a dozen seconds or some. This laptop doesn't have anything I'd see as unusual for daemons - there's no mysqld, sshd, or anything unneeded like that. But to have everything Just Work(tm) at this stage is amazing.

Nice job. =)

update: Angie's asked me to figure out why the screen saver kicks in after 30 seconds instead of 10 minutes like it's set to. But still, if this is the worst that goes wrong with this release, then it's still a wild success already at Beta.

So, Waldi filed #545032, arguing the udev rules shipped in devicekit-disks are broken. He even added a conflict in dmsetup because of that. Whenever you ask him about this issue, the only things he answers are:

• The rules are broken and they need to be removed.
• The information is already available in other variables.

It would be nice if all of this was true, but:

• There is no explanation about how and why these rules could actually break anything.
• The information is not available in other variables, apart from a couple of trivial things. Or maybe it is, but since all that udev crap has absolutely zero documentation, there is no way to know how it is available.

The result: GNOME is currently not installable on a system with LVM. There is no foreseeable possibility to fix that. Great. Thanks.

Hey all! I m sorry I haven t been very active with my packages recently. I all-of-a-sudden started grad school and have been swamped with studying. I also started a contract and have been busy trying to learn a new codebase while contributing something other than snark. I promise I ll get back to packaging IronRuby and IronPython on Mono for Debian as soon as things start settling down. Getting an A in the class is higher priority, though, sorry Don t worry. I haven t forgotten about you ;) Cheers, C.J. PS, I am implementing a Perl library to exercise my understanding of the class. You can follow along at the search.cpan.org page for Lingua::HPSG or by cloning the git repo: $ git clone git://karma.colliertech.org/colliertech/langparser

Some tens or hundreds of thousands of T-Mobile USA customers probably just lost the contacts, photos, and notes "on their phone" forever. Those data are primarily stored on servers run by Danger, a subsidiary of Microsoft; rather, they were, until data loss destroyed them. Many customers' phones do not have a complete copy of their data; some have no data at all. T-Mobile texted the affected customers with a link to this message:

Personal information stored on your device - such as contacts, calendar entries, to-do lists or photos - that is no longer on your Sidekick almost certainly has been lost as a result of a server failure at Microsoft/Danger.

Long story short: No backups.

The story is a sad one, told many times Microsoft chose to make no backups, reports Sidekick news website Hiptop3.com. It seems the team at Microsoft in charge of user data storage allowed an upgrade to the storage system with no backup in place. When the upgrade failed, the data was toast. This sad story of "no backups" is something that sysadmins -- myself included -- find ourselves in from time to time. Toward the end of 2006, the "Leafycaust" (disk failure with no backups at leafyhost.com) destroyed lots of data for Students for Free Culture. I had my own catastrophe in mid-2007 when two disks connected in RAID-1 (which is not a backup system) failed within two days of each other. (I now run nightly off-site backups for all the systems I maintain.) People who rely on computers ought to make sure the people responsible are doing those backups, and that the backups are actually usable. Students for Free Culture was a Leafyhost customer; T-Mobile is a Microsoft customer. It's somewhat humorous that Microsoft and Leafyhost provide the same level of assurance.

"im about to cry!" In the Hiptop3.com article above, users tell their own sad stories and give advice.

• "Since then, I ve cried twice."
• "Fortunately, through Facebook, I ve been able to contact one of my brothers who gave me my parents numbers."
• "im about to cry!"
• "For those of you out there who think your information is safe. Learn from this please. Store information online, on paper, on a phone, everywhere. Invest the time!"

Invest the time. It's good advice. Everyone: Please make backups. Don't trust any one thing - your IT staff, your personal computer, or your paid service provider, to keep your data safe. (For more technical readers: Don't rely purely on your DBMS, your filesystem, or one particular location. You know the drill; now practice it.) I'll conclude with a quote. This one is about the Leafycaust, but it could easily be about Microsoft/Danger:

Essentially, these notjobs burn through several thousand dollars of people s money for the service of deleting all their critical (and personal) data.

Marvell made a new version of U-Boot for the SheevaPlug available last month. This version, 3.4.19, integrates SD card support and improves support for FAT32 USB partitions (although USB support in general is still a bit flaky). This page describes how to install this new version of U-Boot on your SheevaPlug. First of all, download the SheevaPlug U-Boot 3.4.19. Unpack the zip file and copy the file u-boot-3.4.19/u-boot-rd88f6281Sheevaplug_400db_nand.bin to your TFTP server. I suggest you call the file u-boot.bin. Then start your SheevaPlug and type the following: The upgrade process will ask whether you want to change the environment: Make sure to answer "n" here. We also need to make a small change to the u-boot configuration because this version of u-boot no longer accepts the console parameter we have been using. Thanks to Tom Rinehart for pointing this out. Instead of console, we'll therefore add a new bootargs_console and update bootcmd to make use of it: You'll see that bootcmd contains $(console). You have to set a new bootcmd and change the console parameter to $(bootargs_console). In my case, I'd type: Make sure you don't forget the quotes! Finally, save the environment and restart your machine:

This entry was originally posted in slightly different form to Server Fault There are several ways to run Tomcat applications. You can either run tomcat direcly on port 80, or you can put a webserver in front of tomcat and proxy connections to it. I would highly recommend using Apache as a front end. The main reason for this suggestion is that Apache is more flexible than tomcat. Apache has many modules that would require you to code support yourself in Tomcat. For example, while Tomcat can do gzip compression, it's a single switch; enabled or disabled. Sadly you can not compress CSS or javascript for Internet Explorer 6. This is easy to support in Apache, but impossible to do in Tomcat. Things like caching are also easier to do in Apache. Having decided to use Apache to front Tomcat, you need to decide how to connect them. There are several choices: mod_proxy ( more accurately, mod_proxy_http in Apache 2.2, but I'll refer to this as mod_proxy), mod_jk and mod_jk2. Mod_jk2 is not under active development and should not be used. This leaves us with mod_proxy or mod_jk. Both methods forward requests from apache to tomcat. mod_proxy uses the HTTP that we all know an love. mod_jk uses a binary protocol AJP. The main advantages of mod_jk are:

• AJP is a binary protocol, so is slightly quicker for both ends to deal with and uses slightly less overhead compared to HTTP, but this is minimal.
• AJP includes information like original host name, the remote host and the SSL connection. This means that ServletRequest.isSecure() works as expected, and that you know who is connecting to you and allows you to do some sort of virtualhosting in your code.

A slight disadvantage is that AJP is based on fixed sized chunks, and can break with long headers, particularly request URLs with long list of parameters, but you should rarely be in a position of having 8K of URL parameters. (It would suggest you were doing it wrong. :) ) It used to be the case that mod_jk provided basic load balancing between two tomcats, which mod_proxy couldn't do, but with the new mod_proxy_balancer in Apache 2.2, this is no longer a reason to choose between them. The position is slightly complicated by the existence of mod_proxy_ajp. Between them, mod_jk is the more mature of the two, but mod_proxy_ajp works in the same framework as the other mod_proxy modules. I have not yet used mod_proxy_ajp, but would consider doing so in the future, as mod_proxy_ajp is part of Apche and mod_jk involves additional configuration outside of Apache. Given a choice, I would prefer a AJP based connector, mostly due to my second stated advantage, more than the performance aspect. Of course, if your application vendor doesn't support anything other than mod_proxy_http, that does tie your hands somewhat. You could use an alternative webserver like lighttpd, which does have an AJP module. Sadly, my prefered lightweight HTTP server, nginx, does not support AJP and is unlike ever to do so, due to the design of its proxying system.

Yesterday I did what Free Software authors supposedly don t do. I wrote documentation. In fact, I had a relatively detailed README already, but I thought this would be a good opportunity to create a more elaborate documentation, using the ubiquitous DocBook. You can read the HTML documentation for arbtt online, where it s automatically updated when I push to the darcs repository. You can see, I use the same CSS file that most Haskell-related DocBook documentation seems to use.One motivation to use DocBook was that I can extract manpages from it, which should be present if I package arbtt for Debian. I was about to complain that references from the manpages to other part of the documentation can not work sensibly, but using the profiling feature of docbook-xsl, I can replace them by a textual reference to the user s manual if the file is processed for manpage output. Have a look at my changes for that if you want to know how it works.For some reason, the <refentry>-tags that make up the manpages are split to separate files when generating chunked HTML output, but they do not appear in the table of contents. You have to find a spot in the text where they are linked, as they are now in the section Program references , to find them. This is unfortunate, I expect that a few readers might miss this important part of the documentation. Am I doing something wrong?For the configuration file, I put an EBNF-style grammar description in the documentation. There is a special DocBook module for that, but I m not satisfied with it. On the right hand side of production rules, it has only special support for nonterminals, but no tags to semantically mark up choice, repetition etc. I followed the lead found at the YAML specification and put <quote>-Tags around literal parts of the grammar, but this is not valid according to the DocBook DTD. Speaking of the DTD: I guess since the EBNF-stuff is just a module, the tag <productionset> is not a valid content of <figure>. So, if I choose to use the EBNF module in a sensible way, I will not have a valid DocBook file.All in all, I am satisified with the result. Especially that nobody can say any more I like your program, but I can t contribute, because I don t know Haskell . Just improve the docs! :-)