I would really like to become maintainer for the J source. However, I don't have the time to devote to this task right now. It's on my list of programming projects I want to do (somewhere above write my own OS and write my own text editor, I suppose). Maybe some weekend I will look at the code and try to sort out what the errors in our current tests mean.
There's a lot that can be added to J: modern features like parallelism and GPU processing would improve the language a lot without complicating it. Numerous bugs have shown up in the forums that could be fixed. I think I am the only one that has expressed immediate interest in deep changes like these, but I find it hard to beleive that there aren't other people who would help. It would be really useful to have a guide to how the J source works: something that explains the data structures and control flow, and how to do something like add a builtin or add a case to ^:_1 . If we have the appropriate documentation, I can almost certainly "sit at the top of the pull tree" and check that changes to the code work or offer help to people with questions about the source. This discussion belongs in Source, if anyone is going to continue it. Marshall On Wed, Jan 16, 2013 at 02:49:55PM -0500, Raul Miller wrote: > On Wed, Jan 16, 2013 at 1:56 PM, Dan Bron <j...@bron.us> wrote: > > Or maybe there is, now, a magical one-click build for J? > > Not yet. I'd like to arrange for some kind of one, at least for the > interpreter core - currently I'm trying to remember how to test > compiled results, and what likely issues are for tests that typically > fail. > > > PS: When are we going to have a serious discussion about making an curated, > > promoted and official open-source version of J? I think, for this to work, > > it has to be independent of the JSoftware version: we cannot keep the burden > > on them. That means one of us needs to sit at the top of the pull tree. > > I'm not ready to think about this kind of issue. Building versions of > the interpeter which pass all the tests (or until we can show that the > tests themselves need fixing) has much higher priority, from my point > of view. > > That said, I see nothing wrong with the JSoftware version, other than > the fact that I am struggling to build instances that work properly. > > -- > Raul > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
