I also got plenty of congratulations on my new job at Sun. Almost everyone I talked to felt positive about this move from Sun. One individual whose experience with Sun dates back to the nineties expressed some doubt about Sun's ability to play well with open source. I can say with some confidence that, at least in the corner of Sun that I live in, they really seem to get it. As an example, read the bottom of this post from Vice President Jim Parkinson where he tells us that Ted and my roles in the open source community are not changing. It is nice to get that kind of statement from that far up the chain. For further reassurance look no further than the way Sun has handled the JRuby project.
This was the first time I met Ted Leung in person. Be sure to read his notes on PyCon 2008 as they contain much more detail than mine. Ted was hired by Sun at the same time that I was hired. His job description ranges wider than mine: He is tasked with helping Sun become a good member of the Python community. Mine of course is just to make Jython succeed, a task that is quite a bit easier to define.
I got to meet some IronPython folks, Jim Hugunin and Dino Viehland. Jim Hugunin is of course the original inventor of Jython as well as IronPython. Dino is the lead developer of IronPython. They have also managed to get Django running in their world, and they are doing some pretty cool stuff with Microsoft technologies like Silverlight. I feel for them in that they are completely forbidden from looking at even a line of CPython code, and it is fairly difficult for them to send patches to outside projects (to, for example, make some part of Django work better on IronPython). They are working with the lawyers to get them to ease up a bit, but that's tough going. I'm not sure how they can work like that... ouch.
At PyCon, Jeffrey Yasskin started an effort to get a concurrency memory model written for Python (Similar to the way Java and C++ have such models). I have high hopes that this can be accomplished as it should make it simpler to think about concurrency in Jython (Something that Jim Baker has been spending a lot of time on lately).
So on to my favorite part of PyCon: the sprinting. Last year the Jython sprint consisted of three guys at a table with me as 1/2 a sprinter while I was working with Trac. This year Jython took up a room with around 11 sprinters. Sprinters from other projects came in throughout the week asking about how to get Jython working with their projects which was very nice. A particular moment where the value of the PyCon sprints stood out to me is when brand new Jython committer Nicholas Riley was working furiously on getting parts of Twisted working on Jython while surrounded by about six Twisted core developers giving him pointers. Awesome.
A partial list of Jython sprint accomplishments:
- Start of Twisted on Jython
- New compiler/New parser work
- Better threading support for Jython
- Collaboration with SQLAlchemy to get JDBC support
- Work on writing decimal.py as a wrapper around Java's BigDecimal
- Work on porting mmap to jython
Frank,
ReplyDeleteThanks for your great summary! And for mentioning the Twisted work :-)
It was quite a blast for us to pile into the Jython sprint room and watch the goodness unfold as Twisted unit tests got running on Jython and JP offered invaluable insights :-)
Hi Duncan, that was truly great! That is pretty much what PyCon sprinting is all about (IMHO). Getting the right people in the right place at the right time.
ReplyDeletewe are eagerly awaiting jython 2.5. great work!
ReplyDelete