On Wed, 18 Feb 2004 10:41:33 +0000, Joe Orton wrote: >On Wed, Feb 18, 2004 at 05:17:38AM -0500, Jeff Trawick wrote: >> Joe Orton wrote: >> >> >Thanks for looking at this Craig. I notice you have left he conflicting >> >matches in the case statements, e.g. >> > >> > *:OS390:*:* | *:OS/390:*:*) >> > echo s390-ibm-os390 >> >and then >> > *:OS/390:*:*) >> > echo i370-ibm-openedition >> > exit 0 ;; >> > >> >this is really the problem which needs to be fixed... >> >> Supposedly the non-standard code for zSeries (a.k.a. s/390) has now been >> removed in apr HEAD. > >OK great, I removed the OS/2 changes too, that just leaves AS/400, does >anyone build on that platform still? > >Some of these platforms are probably already broken with the new >gen-build changes since build-outputs.mk doesn't know about $(OSDIR).
Yeah, I noticed the build was broken on OS/2, tries to compile dso/unix/dso.c. So how do you test for the platform in python? Obviously gen-build.py's get_platform() needs some work... The script also needs to be smarter in that the OSDIR is optional, falling back to unix if there's no OSDIR (EG apr/memory) which doesn't happen if I simply change get_platform() to return "os2". -- ______________________________________________________________________________ | Brian Havard | "He is not the messiah! | | [EMAIL PROTECTED] | He's a very naughty boy!" - Life of Brian | ------------------------------------------------------------------------------
