On Sun, Apr 06, 2003 at 09:40:59PM -0400, Peter S Galbraith wrote: > 6 weeks ago, Julian Gilbey <jdg@polya.uklinux.net> wrote: > > > As things stand with the buildds, the -Indep fields are almost > > useless, and it may actually be worth dumping the -Indep field > > altogether. tomcat, tomcat4, bigloo, bochs, dutch, gcc-avr, > > grub-installer, gstreamer, httrack, hylafax, latex2rtf, libgcrypt, > > libgnome, libgnomecanvas, libgnomeprintui, libgnomeui, librep, > > libwnck, lilypond, ncbi-tools6, plex86, remstats, sawfish, sysstat, > > yaboot-installer are the only packages in sid which are not > > Architecture: all and which have a Build-Depends-Indep field. > > gri has had it for a long time. [Hmm, missed that discussion when it happened] As has dancer-ircd, and it's working fine. I don't know where Julian got that list from. Phasing out B-D-I would involve installing about 50+Mb of sgml crud on every buildd for every build, which I don't really want to do. Policy currently has a "should" directive for the build target; I ignore this completely, and have build as an alias for build-arch, since that's what actually makes things work on the buildds. Why do we still have a 'build' target in policy anyway? Isn't it about time we ditched it, or at least made it optional (yes, buildds, they don't really need to run it afaict)? -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | Dept. of Computing, `. `' | Imperial College, `- -><- | London, UK
Attachment:
pgpQK9yYTJDpt.pgp
Description: PGP signature