David posted the new StatCVS scores:
We are still going strong! The Mauve test suite has seen a lot of new additions. And it takes a long time now to run through them all. Luckily Tom has setup an autotester that warns about regressions if it sees a new mauve failure when someone commits something that wasn’t properly tested against all of mauve.
Updated FreeSwingTestApps and finally tried out BeanShell. I am impressed! It is a great way to quickly test out some things. It builds out of the box with GNU Classpath. The swing gui needs some work, but you can actually try some Free Swing stuff now. And Robert made BshBot that you can see in action when you visit #classpath on irc.gnu.org (when Robert is online).
Eclipse 3.1 and OpenOffice 2.0 hit Debian unstable
Happy to see both eclipse 3.1 and OpenOffice 2.0 hit Debian main.
Fun with small devices
Found some interesting things people do with GNU Classpath these days:
Away from the internet for a week. So don’t expect me to reply to email quickly.
Tom wrote an article for Red Hat magazine Fedora and the J-word full of nice hints and tips on gcj. Havoc Pennington wrote up some observations about server programming in java. His conclusion: GNU Classpath will save us! Thanks Havoc. We try to do our best.
i2p – GCJ support is on the way
Nice liberation story from the i2p project:
In any case, this is quite kickass, as it means we’ll be able to both integrate more cleanly with other languages AND ship whereever GCJ ships (DFSG friendly!) Thanks go to the GCJ and GNU Classpath folks for their hard work!
Kickass indeed!
Andrew posted some nice results showing that we now pass more than 95% of the Jonas test suite with gcj. That prompted Christian and me to see how far we could get with JBoss on Cacao and JamVM with GNU Classpath. After some hacking 4.0.3 actually starts up:
Not very stable at the moment. But a nice start.
A couple of important new releases to complete the “GNU Classpath picture” this week:
Casey also made a GNU Crypto 2.1.0 RC1 release, but warns that people shouldn’t use this release if they aren’t willing to upgrade again in a week or two (and especially, don’t package this release for a distribution).
On developer.classpath.org you can see the results of running gjdoc on GNU Classpath CVS to generate full class documentation.
It has been a long time since I read a book in one go cover to cover. The Curious Incident Of The Dog In The Night-Time by Mark Haddon was a really nice read.
And then I heard the sound like swordfighting and the roaring of a train coming into the station and I worked out that there was a big computer somewhere and it knew where all the trains were and it sent messages to the black boxes in the little stations to say when the trains were coming, and that made me feel better because everything had an order and a plan.
Robert setup FreeSwingTestApps to have a collection of Free Software applications that we could use to test out our Free Swing implementation. I tested a couple with GNU Classpath CVS embedded in GCC/GCJ CVS and we are definitely making some nice progress:
As you can see Weka has some buttons not drawn precisely where they should. But lots of things work as the screenshots of Joao show. JTetris and Foxhunt are small games that basically just work. But JTetris can loose keyboard focus and Foxhunt redraws really slowly at the moment. Xebece is pretty cool, but needs Graphics2D which still has to be ported to cairo 1.0 so there is some instability when running it (and it needs a real RTF parser). So still lots of bug(let)s, but the basics are certainly there!
Norman Hendrich has supplied us with a stream of very high quality bug reports. Today he filed a “bug” to celebrate the hard work of Lillian, Roman, Tom F.,Tony and all other Free Swing and AWT hackers. It will be a pleasure to close this bug :)
Update: Norman said some nice things on the mailinglist:
a while ago I made myself some enemies complaining about the AWT/Swing support in gcj. Back then, I was really frustrated about obscure compiler bugs in gcj being fixed within minutes, while even small test programs would crash…
Since then, I swiched from testing gcj+classpath to jamvm+classpath. Overall, the progress with Swing is really astonishing, despite a few drawbacks due to added functionality (like the accessibility stuff, better plaf support, and the repainting refactoring).
Perhaps submitting bug reports with testcases made all the difference?
Some days during the last weeks, it really felt like I had three or four people working just for me: thanks Lillian, Roman, Tony, and Tom F. Naturally, thanks to all the other AWT and Swing hackers, too. I would submit a few bug reports in the morning from Hamburg, and fixes started to come in from all over the world the same day.
| Thanks Norman! You truly are our “gui-testing-adversary”. People seem to really like to beat you at this testing game by fixing even the most cunning test you can throw at them. |  |