Christophe Rhodes wrote: > > Away from the CLD, and again since I've been pointed at this thread, > here are some more thoughts. "Lisp needs labour, not praise", yes, > but I don't think that work for work's sake is of any use to anyone. > Lisp will grow sustainably if people use it, learn to use it better, > write stuff in it and about it, release cool applications or useful > libraries, and so on. I think Paolo realises this, but I don't doubt > his labour, and I address as he addresses those who haven't yet lifted > a finger: if there is something that interests you, excites you, or > even is a work project, then if you can, don't just keep silent, > trying to solve every problem yourself: make the problems you run up > against known. If you've evaluated XML libraries recently and found > them all wanting for your purposes, say so. If you've tried to write > bindings to Apple's MIDI interface and failed because of Apple's > woeful documentation, then shout: someone else may have hacked > something up, even if it's not release-quality. > Since you mentioned it, it's ages since there has been a tarball release of mcclim. Is not this an area where cl-gardeners could help? Much testing is needed prior to a release, and having a pool of volunteers willing to bang on the contents of the latest and greatest CVS-of-the-day can't hurt.
I know mcclim is experimental code, but it's also widely distributed experimental code, with many distros packaging it; the failure of it to compile with current versions of sbcl is an embarrasment. It's not just mcclim that has these issues, it's CLX and slime that also have the "latest version is CVS" mentality, and I think this might be something of a barrier to someone merely curious investigating lisp - no tarballs means no packages downstream - why should downstream distro users have to leave apt-get/emerge/rpm or whatever behind in order to try lisp? I'm not sure how serious a barrier to adoption it is, but I do know I've had to learn the hard way that things like emerge cl-mcclim will not "just work" as they do for multitudes of other packages. As for "lisp needs labour, not praise". I don't understand why valuable projects like CLOCC or the Common Lisp Cookbook seem to wither on the vine. In my case, I never finish anything. I'm just a grashopper. I have on my hard drive in various stages of completion, a software renderer, a patched version of port, a mud client and a library that takes scm repository urls and fetches them into a given directory. If I ever actually get to finish one of these, and find a few people willing to test, then yes, I'll stick them into a ball and put them on cliki. Another frustration is abandoned projects on common-lisp.net - something like cl-ncurses, which I'm presently using, could do with quite a few minor patches - docstrings, the odd missing function, and could be greatly improved in utility by the accumulation of patches, but the maintainer seems to be absent and there is no mechanism to reclaim such a project. Anyway, time for bed. > Cheers, > > Christophe > > _______________________________________________ > Gardeners mailing list > [email protected] > http://www.lispniks.com/mailman/listinfo/gardeners > _______________________________________________ Gardeners mailing list [email protected] http://www.lispniks.com/mailman/listinfo/gardeners
