Since about two days I’m proud owner of a Lenovo X61s notebook. It will become my main and new developer box for grml, featuring all that nifty stuff like 64bit support, hardware virtualization and so on… I managed to use all the important features with a very current version of grml. I’m using kernel 2.6.22-grml on a setup with LVM, cryptsetup-luks,… everything works fine so far (including Suspend-To-RAM - even in grml live-mode when booting from USB pen - thanks for the excellent hibernate package, madduck!). As you won’t find useful and approriate information about Linux on the X61s yet (either pretty incomplete or wrong) I decided to put all relevant information on a webpage in the grml-wiki:

• Lenovo Thinkpad X61s @ grml-wiki

Short version: the X61s rocks. I’ll provide information to thinkwiki as soon I checked out all the details…

/me on the Debian Maintainers GR After reading on Planet some opinions against the DM GR I made up my mind: I want to post in favour of the GR (and I've already voted accordingly). I swear I'll (try to) be brief. My main point is: this is not an ugly hack to address the long unsolved problem of our sloppy NM queue, this another way of letting people contribute to Debian and acknowledge their work. Rationale follows. Maybe it's just me, but in several year of DD-ness I've met on the net a bunch of people which are "just" interested in maintaining some packages in Debian. I've been training some of those people and of course sponsored their uploads. But if they aren't willing to enroll the NM queue, if they aren't interested in the other social/political aspects of the Debian project I can't blame them, it's their call. I'm in Debian as a DD because I do care about other stuff than plain package maintenance, but I can't ask every hacker around there to share my motivations. Nonetheless I don't see any valid reason to reject and/or put extra artificial burden (i.e. life-time sponsoring) on the good technical work that such people can contribute to Debian. It's just a matter of identifying a new status for contributors that states "I'm contributing my package maintenance skills to Debian, and I'm proud of it", ... without implying full Debian zealot-ness The GR precisely addresses this need: establish this new status, the Debian Maintainer status. (We would need a whole bunch of similar new status for other kind of contributions, like translations and artworks, but this is an entirely different story, unfortunately not addressed by the DM GR) And regarding the fear of decreasing package quality, the sponsoring process has already proven to be bugged in this respect: when they really want, sucking packages can enter the archive, they will find some sponsor. There's no silver bullet, our best hope is to establish social practices which encourage people producing (directly or not) good packages and discourage the others. In this respect the ability to stop accepting advocacies proposed in the GR sounds good to me.

First off, this guy is walking around handing out free fudge. Neat idea, but I can't quite figure out why!

Bill Hilf from Microsoft spoke today. He started by saying "I guarantee you won't agree with a lot of things I say today."

Nat introduced him as "our man on the inside."

Microsoft is submitting their Shared Source license for OSI approval. I don't know if this means that MS will have to actually make any modifications.

New website today: http://www.microsoft.com/opensource.

Bill invited feedback from anyone at billhilf at microsoft dot com.

He got some fairly pointed questions from Nat about software patents. Bill said two things about that: 1) Microsoft flubbed the communication with the press and that their intent isn't to sue, and 2) Microsoft is learning and he's working to help them learn.

Rick Falkvinge from the Pirate Party

Funny and informative talk about why they've started a Pirate Party in Sweden.

Originally copyright only impacted public places, but now it impacts people's private lives as well.

Copyright is protecting the entertainment industry -- funded by luxury purchases -- at the expense of privacy and individual freedoms.

"The Norwegian Liberal Party forked our platform into Norwegian"

www.piratpartiet.se/donate

"Political donations in Sweden are not regulated", so you all can contributed

Steve Yegge (Google)

"Google will probably fire me for this talk"

Neither Steve's mic nor his laptop appear to work... after his line about being fired, someone shouted "Google is very powerful!"

"A brand, in geek terms, is a pointer."

GTE was, for a time, the most reviled brand in the US. Then they invested in improvements and ran commercials basically saying "we don't suck anymore", but nobody believed it.

"The single biggest branding problem in Open Source is the name 'Open Source'."

Last year I spent several months living on one side of Melbourne and working on the other and travelling by train to work. Every day I had to catch two trains each way with an average wait of 5 to 10 minutes for each train to arrive giving a total of at least half an hour a day spent waiting for trains. The obvious solution to this problem is to have trains not go back and forth on one line but instead go from one side of the city to the other. This map shows that there are 14 train lines out of the city with about 5 major lines. The smart thing to do would be to have every major line have one train every 5 minutes during peak hours and to have every train go back out on a different line. Then if you are on one of the major lines and want to travel out on one of the other major lines then every 20 minutes there would have a train that would take you straight there without changing trains! Currently we don’t even have such frequent trains, during peak hour the trains on most lines run no more often than 5 per hour, the Sydenham line has trains 4 times per hour during peak hours and the trains are crammed full before they get half-way to the city. If the trains ran more frequently and were routed through the city then commuters who travel through the city would save 20+ minutes per day without going to any effort and 30+ minutes a day if they chose to start their journey at a time to avoid changing trains. This would be a significant incentive for catching the train instead of driving! For the commuters who travel to work via a single journey then having trains run every 5 minutes at peak times would mean that an average of 2.5 minutes was spent waiting for a train each way (an average of 5 minutes per day) instead of the current situation of 15 minutes per day or more. This would mean triple the number of trains on the Sydenham line which may sound excessive. However the trains are currently so crowded that there could be twice as many trains and all seats would still be full. If there were three times as many trains then I expect that more people would catch the train (surely some people would be convinced to drive to work by the idea of spending 20 minutes with barely room to stand), it’s not inconceivable that there could be three times as many trains and all seats could still be full! The next issue I have been considering is the time taken for a tram ride to/from the central city areas in peak hours. Peak hour trams stop at every stop because there are always people getting on and off. If a tram could stop less frequently then it could make a slightly higher average speed. One way of achieving this would be for the peak hour trams to stop at every second stop outside the center of the city. On the way in half the trams would accept passengers at each stop (each tram would be designated as either odd or even and labelled as such - the tram stops are already numbered). But if you have twice as many trams then the average wait would be the same while the duration of the trip would be reduced. On the way out of the city the tram driver would announce that after stop 10 (to pick a random number that might work) the tram would only allow passengers to get on or off at even/odd stops. If you knew that your stop was on an even number and the tram was an odd-numbered tram then you would change trams to an even tram. The small delay in changing trams would be made up by a faster trip overall. Politicians are always talking about ways to alleviate the water shortage caused by climate change and to improve the economy. Having people spend an extra 10 minutes a day working because of saved time on the trains would help the economy. Encouraging people to catch the trains via more frequent and efficient service as well as less overcrowding would help reduce climate change - which is the best way of improving our water supply and the only way of helping the farmers long-term!

Share This

Peter Brown replied:

"Thanks for the message. Here are some comments on the issues you raise:

"> - On the Radio New Internationalist show "Up in Smoke", during an interview with Peter Brown of FSF (about 43 minutes in), I was surprised to hear the following revelation that FSF uses Skype and the failure to challenge the claim that Skype is free software:"

The FSF doesn't use Skype. I would appreciate if you would clarify that on your blog post. The interviewer used Skype to telephone me on an FSF telephone. I agree that from what I said, it would be easy to infer that FSF uses Skype, but it doesn't. The interview lasted about an hour and we covered a lot of ground. Also, in this exchange I didn't clarify for the listener that when the interviewer remarked that "Skype was free", they meant as in price, but not as in freedom.

"> - I was disappointed not to find much on the FSF web site about the need for free VoIP software, building on GNU oSIP and other free software, instead of Skype's proprietary software, which has contained spyware and worms already. Will you be adding such information soon? What is the official FSF view of Skype and similar proprietary VoIP systems?"

Skype is proprietary and we don't use it. In that section of the interview we were discussing telecoms monopoly and net neutrality, and I lost the focus on the free software issue - my bad. As far as adding an article about VOIP on fsf.org I think that's a great idea. Unfortunately, I don't think we can cover that immediately, as we have some campaigns we have other campaign activities to focus on. We would be glad to take a contribution though, if you would like to write such an article - we might also want to think about using such an article as a basis for a campaign.

"> - Nevertheless, well done for covering DRM, net neutrality and the privacy problems of Google, Yahoo! and Microsoft in a relatively short interview."

Thanks. I actually used to work at the New Internationalist in Oxford, and they are a great coop. I have been working with them now for a couple of years, tying to encourage them to cover free software related issues for their audience. Have you seen the article Bruce Byfield (Newsforge) wrote that I got placed in the November issue of the magazine? all the best peter"

Thanks for the reply! To my shame, I'd completely forgotten that Skype can also call ordinary telephones and the interview didn't remind me. I guess that might be because my main awareness of Skype is when it's a problem, when someone gives a skype: address instead of a sip: one or a real phone number. However, I'm not sure a called person would usually say that they were being called over Skype, rather than being telephoned. FSF people are smarties, though, so can hear the artefacts, or maybe the interviewer had mentioned it beforehand. Anyone else want to summarise the benefits of free software VoIP compared to Skype systems? I won't get time before next month. kris commented:

"There's openwengo, it has more features than skype (i.e. video), is encrypted, open and free. Enjoy."

I keep meaning to try openwengo, but each time I look at it, I fail to find the source tarball for the latest release. It also seems to be Qt-only, which I also try to avoid installing (as it'd be a fourth set of desktop widgets). At the moment, I'm using linphone from the command-line. I sometimes consider replacing it with a stand-alone telephone adapter. I wonder: are any running free software? (2007-07-27: 7 comments)

Retiring from Debian In response to the inactive developer ping, I have decided to retire from the Debian project. Between my job and the new baby, my time for hacking on other projects is nonexistent, or possibly negative. I have already passed Audacity maintainership to the debian-multimedia team. I still use and like Debian, and I hope someday I'll return from "emeritus" status to become an active developer again. Cyclone While browsing through some very old backup files, I rediscovered one of my first real software projects, an IRC client for BeOS. This was a neat collaborative effort, but we weren't really on the ball with project hosting or version control. This is from a file I wrote documenting what was left of the project after it stopped:

Cyclone was started in spring 1998 by a group of developers who hung out in the #bedev channel on EFNet. The original Cyclone team included Jason Gosnell (Avatar), Jeff Hamilton (Pyrus), Matt Lewinski (mattl), John Wiggins (prok), and Matt Brubeck (mgb), with contributions from several other #bedev regulars. The Cyclone project produced a small, elegant IRC client. Unfortunately, the team dissolved as several members moved around the country to go to new schools or jobs at the end of summer 1998. Along with the loss of some important pieces of source code, this prevented Cyclone from ever seeing a public release.

I also found some poetry I wrote in middle school, but it's compressed using a version of StuffIt that does not seem to be compatible with any open source tools. This is probably a good thing.

In 1992 Severn Cullis-Suzuki (David Suzuki’s daughter) who was 12 years old gave a talk to the UN’s Earth Summit in Rio on behalf of . She gave a really good talk, see the below Youtube video. The best quote is “If you don’t know how to fix it, please stop breaking it!”. Unfortunately they haven’t stopped breaking things yet. http://www.youtube.com/watch?v=uZsDliXzyAY. Here is some background information on Several Cullis-Suzuki.

Share This

I am still looking for a solution that allows me to tag my mails similar to what Gmail offers. Having thought about this for quite a bit now, I can see three solutions, implemented at different levels in the mail flow: as filesystem, patch to the IMAP server (Dovecot), or patch to mutt. Since this post is about the filesystem, I'll concentrate on that, but I'll also touch upon the other two solutions briefly.

I just sent this: Dear FSF, On the Radio New Internationalist show "Up in Smoke", during an interview with Peter Brown of FSF (about 43 minutes in), I was surprised to hear the following revelation that FSF uses Skype and the failure to challenge the claim that Skype is free software:

"Peter Brown: ... This interview today is being transmitted through Skype. We're talking through Skype. Now, that obviously is a direct threat to the entrenched telecoms and they would like to restrict that, to lower the quality of the voice connection and they want to do that with lots of other types of transmission. Rachel Maher: And can they do that? Because Skype of course is free and there's a really direct benefit for organisations like ours, which is a non-profit organisation using Skype technology. Will they be able to do that? Peter Brown: Yes, so what they can do is easily identify types of information being passed, so what they can typically do is downgrade the service that you're using, so that - for instance, voice communication - those packets of data can be slowed down, effectively making voice communciation more difficult. Now this can only really be achieved if they're able to manipulate legislation. Unfortunately, they have a long history of being very successful at this. I mentioned earlier on that Digital Restrictions Management are a threat to our freedoms because it's allowing them to erect gateways and to control what it is that we can view and do with our computers. ..."

I was disappointed not to find much on the FSF web site about the need for free VoIP software, building on GNU oSIP and other free software, instead of Skype's proprietary software, which has contained spyware and worms already. Will you be adding such information soon? What is the official FSF view of Skype and similar proprietary VoIP systems? Nevertheless, well done for covering DRM, net neutrality and the privacy problems of Google, Yahoo! and Microsoft in a relatively short interview. Regards, MJR

The dpkg team has a nice history of changing VCS over time. At the beginning, Ian Jackson simply uploaded new tarballs, then CVS was used during a few years, then Arch got used and up to now Subversion was used. When the subversion repository got created, the arch history has not been integrated as somehow the conversion tools didn’t work. Now we’re likely to move over git for various reasons and we wanted to get back the various bits of history stored in the different VCS. Unfortunately we lost the arch repository. So we have disjoints bits of history and we want to put them all in a single nice git branch… git comes with git-cvsimport, git-archimport and git-svnimport, so converting CVS/SVN/Arch repositories is relatively easy. But you end up with several repositories and several branches. Git comes with a nice feature called “git rebase” which is able to replay history over another branch, but for this to work you need to have a common ancestor in the branch used for the rebase. That’s not the case… so let’s try to create that common ancestor! Extracting the first tree from the newest branch and committing it on top on the oldest branch will give that common ancestor because two identical trees will have the same identifier. Using git_load_dirs you can easily load a tree in your git repository, and “git archive” will let you extract the first tree too. In short, let’s see how I attach the “master” branch of my “git-svn” repository to the “master” branch of my “git-cvs” repository:

$ cd git-svn
$ git-rev-list --all   tail -1
0d6ec86c5d05f7e60a484c68d37fb5fc31146c40
$ git-archive --prefix=dpkg-1.13.11/ 0d6ec86c5d05f7e60a484c68d37fb5fc31146c40   (cd /tmp && tar xf -)
$ cd ../git-cvs
$ git checkout master
$ git_load_dirs -L"Fake commit to link SVN to older CVS history" /tmp/dpkg-1.13.11
[...]
$ git fetch ../git-svn master:svn
$ git checkout svn
$ git rebase master

That’s it, your svn branch now contains the old cvs history. Repeat as many times as necessary…

In a default configuration of Xen there will be a virtual Ethernet device created for each interface which will be associated with a bridge. A previous post documented how to configure a bridge named xenbr0. The basic configuration of Xen that most people use is to have a single virtual Ethernet port for each Xen instance and have them all connected to the one bridge, and then the Dom0 will have an IP address on the bridge interface that is used for routing packets to the outside world. This works really well if you have a subnet that you are using for all Xen DomU IP addresses, if you are using NAT for communication, or if the DomU needs no communication outside the Dom0 and other DomU’s on the same machine (a common case for testing). But if you have a collection of servers that you want to consolidate on a single piece of hardware then you end up using a single sub-net that spans some physical machines, some Xen Dom0’s, and some DomU’s. The solution to this is to use bridged networking. Unfortunately most documentation of bridged networking is really confusing, and non of my google searches turned up the most relevant fact: When setting up a bridge on the local Ethernet you must make your physical ethernet device (eth0 or whatever) be strictly a slave to the bridge and then assign the IP address used for the physical network to the bridge.

ifconfig eth0 up

brctl addif xenbr0 eth0

For example if you have 10.0.0.42 being the IP address used by the Dom0 on the local Ethernet via device eth0 and you want to use bridging for DomU’s then you simply make eth0 owned by xenbr0 (the typical name for the Xen bridge) with the above commands in your script to configure the xenbr0 device. Then treat xenbr0 in the same way that you treated eth0 before enabling bridging. Also there’s nothing stopping you from having one bridge for DomU’s that can talk directly to the physical Ethernet and another for DomU’s that are only to use routed networking, see my previous post about using multiple ethernet devices in Xen for more background information.

Share This

My brother Ted and his partner Nabil had a great commitment ceremony yesterday afternoon in Dolores Park in San Francisco. It was a beautiful day, and they had lots of friends who came out for them. I came up from Los Altos Hills with our parents, and my brothers Nate and Andy came with their friends and/or families, too. Ted and Nabil march to a different drummer, to put it mildly. Their ceremony was remarkably like a Quaker wedding, although I don't think they planned it that way. They had a nice picnic and tea party in the park for a few hours, let the guests get to know each other, and then called everyone together to listen to their promises to each other. (Very sweet, by the way). They then exchanged bejewelled carabiners -- those funky little mountain-climbing metal loops that people also use as keyrings. "We may lose necklaces or rings, but we're not going to lose our keys." The carabiners were decorated with the California state gem, benitoite (Benito-ite), which is rarer than diamonds. Brilliantly, their friends who designed the carabiners also managed to include a USB thumbdrive, on which they're going to copy all the pictures, stories, movies, and other digital paraphernalia from the ceremony itself. Ted and Nabil will be carrying around those pictures as long as they have the carabiners. Clever! After they'd exchanged the carabiners, they hooked them to their belts, and then hooked them together. "And now... we're HITCHED!" Thunderous applause. They then invited up anyone who wanted to make a toast. My brother Nate went first, and girt in a bastard sword (for unclear reasons), he made a great toast. A dozen other people, including my Dad, came up to speak. By then, Ted and Nabil were falling over with fatigue, so we had mercy on them and stopped coming up to say nice things about them. Everyone dug in to champagne and cake, and Amita June and her two cousins Elena and Tessa ran around on the huge playground in the park. Definitely a good, good wedding-like event. tags: ted nabil wedding carabiners commitment ceremony

Long day This morning we had a Skype conference call relatively early in the morning, and then a brunch with Nabil's family (who are visiting from Virginia and Philadelphia). It was really nice -- my folks decided to build a huge patio in their backyard, and it's great for having kids and family running around, as well as for outdoor brunches. (We haven't had a meal inside since we got here, actually.) This afternoon we drove up to Novato to visit Maj's grandmother Jeannette and her Uncle Kevin and Aunt Robin. Also present were Jeannette's lifelong friend Roseanne and Kevin's highschool pal Jim. Amita June loved running around their big backyard and playing with their sweet old dog Holly. She also got a lot of mileage out of their porch swing. Maj and Robin swapped cooking tips and renovation horror stories, and I got to listen to Jeannette and Roseanne tell gambling stories from various Southern California Indian casinos. Jim, meanwhile, was on his cell phone in another corner of the yard, negotiating with another gold prospector with a hand injury to work the friend's claim. Quite an evening. After a dinner of barbecued salmon, caprese salad, and corn-on-the-cob, Kevin and Robin's neighbors came over, and we talked more about statistical analysis, sheep-farming in Marin County and driving through Austria. A long drive back to my folks', but Amita slept the entire time, thankfully. Good for us all. tags: kevin robin novato jeannette indiancasinos porchswing amitajune

Voice jail I had to call Air Canada a few days ago to get our tickets to come out here to California, and I had a terribly frustrating time working my way through their automated systems to get to a human being that I needed to talk to. And I thought to myself, "Why do I hate these automated voice systems so?" I never call a 1-800 number or another voice system if my problem could in any way be handled automatically. I always go to the Web first, and only try to get on the phone if the particular special-purpose request I have hasn't been, or can't be, built into the Web interface. But the economics of call centres makes the company want to shunt me into one or another automated voice systems in hopes that I'll be satisfied and not bother a real human being. My only goal with a phone call is to talk to an intelligent human being who can handle my exceptional situation. But the interface is designed precisely to deny me that goal. Can that interface ever be anything but frustrating? tags: phones voicejail frustration interface

Least likely to succeed Ross Mayfield posted a nice pic of me offering to run his company in exchange for a hot meal. It's a good shot of me, actually. In other linking issues, my Technorati "Authority" value is now 101, putting me squarely into B-list Bloglebrity territory. Hooray for the second string! tags: b-list socialtext technorati rossmayfield bloglebrity

For a while now I've wanted a standalone internet radio; something that only required power and had wifi and would then stream audio from my desktop machine. A bit like the Slim Squeezebox but without the need to plug it into a HiFi setup. I'd found the range of devices based on the Reciva Barracuda module, but they all seemed a bit too expensive to impulse buy. Currys are now doing the Logik IR100, which uses this module, for £49.99, at which point I decided it was worth a look. I'd already done some investigation about the module. It runs Linux, Reciva do the Right Thing and make their code available, and it can have its firmware upgraded. There used to be a wiki of info, but it's disappeared recently. Richard Kalton has a blog of his progress - he's got a shell, but doesn't seem to have found any serial console pinouts and seems to have gained the info required to get a shell (which he doesn't disclose) from reading the flash off the board itself. 'bill888' also has a general Logik IR100 page. The radio itself seems quite consumer friendly. You plug it in, press the on button and tell it to search for networks. It then shows a list of ESSIDs which you can select from. It'll let you enter a WEP/WPA key if necessary, but I setup an unencrypted wireless network for testing. It then downloads a station list and you can chose what you want to listen to. Seems to do exactly what it says on the box. It doesn't support Oggs though. And all my music is Ogged. So I want in. As a first approach I'm trying not to take it apart - it doesn't seem like anyone's got serial access, or confirmed JTAG as working, so there doesn't seem to be anything to be gained by opening it up yet. So I decided to watch network traffic, in particular what it does when you ask it to check for new firmware. The initial request is for: http://copper.reciva.com/cgi-bin/service-pack.pl?serial=<serialno>;sp=<current service pack>&hw=<hardware id>&sv=<serial firmware>&check=1 This then sends a redirect to: reciva://copper.reciva.com:6000/service-pack/sp-wrap-cache/sp.255-c-158.tar Now reciva:// is an encrypted custom protocol it seems, that passes the serial number to the server, which then returns a challenge, which the device then hashes up and generates a session key, passes back to the server and asks for the file. You can see this in lib/rtp.c from the curl tarball on the Reciva GPL site. Unfortunately there are bits missing - several files (rtp_encrypt.h, rtp_encrypt.c, sernum.c) and it appears to want to talk to some sort of sernum daemon. As such I haven't been able to get a copy of the firmware upgrade tarball. :( However, if I could get one and have a look to see the format (is it a full file system image, or just changed files? Is it relative or absolute paths?) then I think it should be possible to build a new firmware image, hijack the connections to copper.reciva.com on my local network, and redirect to a http:// URL rather than a reciva:// URL, thus avoiding having to do the encryption stuff. Assuming, of course, that the upgrade isn't itself signed or encrypted. More prodding required...

It's our second wedding anniversary on Monday, and because Sarah's not been to Napa Valley yet, and because this is probably the last opportunity to spend a bit of quality time away together before the pitter-patter of little feet, we've lashed out for a nice weekend away. We're staying two nights at the Gaia Napa Valley "hotel". It's really nice, but a lot more like a motel in most respects. I really like the environmentally-conscious initiatives they've taken though. I can highly recommend it, and the day spa attached is also excellent. Oh, and they have free wireless, which whilst surprising given how much they charge for things like in-house movies, is also a bonus. It was good enough to make laptop-based VoIP calls with. Today, we drove around aimlessly after our massages, in search of wine tastings (it was handy having a a ready-made designated driver). It seems the wineries here are all into charging for tastings, which in my opinion is a bit of a deterrent (in some cases) to purchasing. The charge varies between $5 and $15. Some wineries seem to refund the cost of the tasting if you make a purchase (which seems like the way to go in my opinion). We stopped for a late lunch at the Rutherford Grill, where I think I've had the most awesomely delicious rack of barbecue pork ribs I've ever had in this country. I cannot recommend that place enough. Apparently their rotisserie chicken is also to die for. We picked up thirteen bottles of wine in our travels this afternoon, so for the first time in ages, we'll actually have something in the wine rack at home. Tomorrow we're going on the Napa Valley Wine Train, which looks like it'll be fun. So now Sarah can say she's seen Napa.

As posted earlier on the rather recently announced CRANberries feed (that is generated and hosted on my box) covering CRAN package updates, my main server bit the dust. The hard disk no longer wants to talk to any of its lvm partitions, leaving me without /usr, /var/, /var/local/, /home, /srv. Fortunately, most of this was backed-up but now I have to go through reconfiguring the replacement machine I already set up. Not fun. If anybody has tips on recovering the lvm partitions, I'm all ears. /etc/lvm/ seems fine if that helps as a starting point.

I read with interest Erik Moeller's note about sources on Wikipedia: Wikipedia s core problem is not expertise, it s self-selection. He quotes the following paragraph from the article about Mitt Romney: Romney was sworn in as the 70th governor of Massachusetts on January 2, 2003, along with Lieutenant Governor Kerry Healey. Within one year of taking office, Romney eliminated a 3 billion dollar budget deficit. During this time he did not raise taxes or debt. He also proceeded to end his term with a 1 billion dollar surplus as well as lower taxes and a lower unemployment rate. Erik pointed out that the cited source was the Mitt Romney for president campaign, and then worried about the self-selection of Wikipedia contributors. I think he's on the wrong track. I don't think the sources of Wikipedia articles matter that much, if the text is readable, true, neutral and verifiable. The part that really made me cringe, in the above paragraph, was the awkward construction He also proceeded to end his term... Awful. I've streamlined the sentence to He ended his term with a 1 billion dollar surplus, lower taxes and a lower unemployment rate. As I tried to explained in my unfinished essay on Readers and writers on Wikipedia, I think that as Wikipedians we concentrate far too much on the agenda and rights of contributors, and far too little on the readability and usefulness of the articles themselves. I think WP is going to become more professional when more Wikipedians practice their writing skills than their rule-making skills. We should concentrate on the quality of the output, not on the source of the input. tags: wiki wikipedia erikmoeller mittromney writing

On Wednesday I had a viva for my Master by Research thesis on “Text classification by quality”. After a 15 minute presentation and almost two hours of questions from two of my examiners, I was told that I will receive my Masters degree (after minor corrections to the thesis). It is a great relief and a great reason to celebrate! In order of celebration I will embrace consumerism and get me two things: World of Warcraft: Burning Crusade expansion and a cat. Next week I will be going around animal shelters and cat owners in Riga in a search for a 1-2 months old black, male kitten. It has been my dream to have a cat again since my last cat died of old age 3 years ago. Unfortunately I lived in dormitories where they did not allow pets or in UK. But now, I can finally get one. I am very exited about that

Since quite a while gettext has support for a feature which I was waiting for since ages: --previous. It means that when a string gets marked fuzzy it adds the previous original string as a comment so one is able to see the difference which led to the fuzzy marking. This makes translation of po files much much more comfortable because one can easily find out now what has changed. Let me give you a reallife example:

#: data/core/help.cfg:331
#, fuzzy
#  msgid "race^Mechanical"
msgid "trait^Mechanical"
msgstr "Kriegsgerät"

With out the previous string (in the "# " line) you would propably be puzzled about what has changed because the string gets stripped off the everything before the ^ for displaying it in this case. Though, not all translation tools behave correctly. One of them is unfortunately poedit. It moves the previous string to a wrong place and thus breaks the parsing of the file. A bugreport has been filed about it—but unfortunately it also means that upstream projects that already support (like Wesnoth) it are considering disabling it again because of this. So I wrote quickly a perl script that is able to fix that for you: pofix.pl. It's quite simple, reads from stdin and writes to stdout.
Go and support --previous before your competitors do it and hijack all your translators! ;)