"David Lichteblau" <[email protected]> wrote in message news:20091114080932.ga2...@radon... > Quoting Pixel // pinterface ([email protected]): > > patch available at > > http://web.kepibu.org/tmp/add-asdf.dpatch > > Thanks... > > > The patch only makes ASDF available like any other clbuild system. I > > leave worrying about ensuring clbuild's copy of ASDF gets used when > > running `clbuild slime` or `clbuild lisp`--and whether such a thing > > would even be a good idea--for others. > > ... although that is the essential part.
I disagree that it's the only bit which makes including ASDF interesting, though that may be due to the quirks of my particular setup. (A single clbuild source tree is shared across a heterogeneous set of machines and operating systems, so I can't generally use binaries built under clbuild.) > We already have code that downloads ASDF for the Lisp which doesn't > include it (namely clisp). If you change that CVS invocation, I could > see the benefit of this proposal. Ha! I seem to have missed that. Didn't think to look for it in the shell script. Something more like this, perhaps? http://web.kepibu.org/tmp/asdf-try-2.dpatch There's a bit of trickery at play, because getting an implementation to use a project requires ensuring that project is available before functions like update_project exist,* and I'm not comfortable enough yet with clbuild internals to attempt rearranging things. The patch continues the current assumption that, if a lisp bundles a copy of ASDF, that's the preferred copy. * I understand now why asdf is not in the projects file, and why the CVS invocations were hardcoded into the clbuild script itself. _______________________________________________ clbuild-devel mailing list [email protected] http://common-lisp.net/cgi-bin/mailman/listinfo/clbuild-devel
