From: "Erik Bågfors" Date: 2002-01-31T07:51:12+09:00 Subject: Re: Cardinal - project looking for a leader On Tue, 2002-01-29 at 00:19, Phil Tomson wrote: > A few weeks back I proposed the Cardinal project - basically a project to > create a Ruby frontend for Parrot. I really think this is a critical > project for the future of Ruby. My plan was to get my other projects into > a state where I could leave them for a while and work on Cardinal starting > in early February... > > However, after releasing the latest version of TaskMaster last week, I've > been talking to some folks involved in areas like Linux Clusters and > software testing and the feedback I'm getting is that with a bit of > improvment and a lot of added docs and examples TaskMaster could become > really useful in those niches (and it could even lead to some consulting > gigs - it would be nice to make a bit of money again ;-) So, my plan now > is to focus on improving and documenting TaskMaster for the next few > months because in the short term I think I can have a greater impact > there than if I move on to working on Cardinal. > > So, that leaves Cardinal 'leaderless'. Would anyone out there be willing > to startup Cardinal (get a page on Sourceforge or Savannah)? It really is > an important project for the Ruby community. From talking to the parrot-people (well, dan anyway), trying parrot and reading the perl6-internal-mailinglist, I believe that parrot is not really ready for ruby yet. It is still _very_ much under development (more than I thought). However, this is our chance to get our feedback into the parrot-code so I think it's important that we work with them from now on. Since I'm really interested in this project but I don't want to be a leader I'd like to say what I think are important and what a leader should need to think about. This is all from my point of view. The way I see it, there are four important areas to work with. 1) The ruby in ruby parser 2) The compiler to parrot bytecode 3) The ruby-classes that needs to go into parrot 4) Changes to parrot for nr 1) there are lot's of people working with this for different reasons, some not related to parrot. I think this is great and I also think that we should use as much of that knowledge and information as possible but we should try to drive them in a "parrot-direction" just yet. The people doing this is doing a great job I think. nr 2) This should take the information from the parser and generate parrot bytecode (of perhaps parrot assemby). For this to work nr 3 have to be started. I really don't see any reason to work with this just yet because things are too much in development right now. nr 3) This is what I've been working with a little. I really think that parrot is great and I don't think there will be any problems doing this once parrot is in a state that make's doable. Implementing classes is not at all hard and considering that parrot needs to provide regexp's, IO-handling and other low level stuff for all languages, we shouldn't have to deal to much with stuff like that (of course, they will not mind our help :) ). nr 4) This is were we should need someone who knows ruby by heart. I know it's not me! :) We need someone who is good at low level stuff and that can look into parrot and see if it's missing something that's important for ruby or if something simply is not right for ruby. This should be done as soon as possible. So. The question is how many people are interested in working with this. I know I am but I'm not ready to take any kind of leader roll, and my knowledge is not good enough to work with some parts of this. I really think we need people for nr 4 and nr 3 now! (nr 1 is already being worked on independently of ruby). Anyway, I'm really really exited about this project! /Erik -- Erik B�gfors | http://www.ardendo.se/ Erik.Bagfors@ardendo.se | GSM +46 733 279 273 fingerprint: 6666 A85B 95D3 D26B 296B 6C60 4F32 2C0B 693D 6E32