Matthias Urlichs <email@example.com> writes: >> No he is right. There is benefit in testing whether a package >> builds in an environment that is not minimal. > > We already get that by virtue of the maintainer, who (presumably...) > builds and tests the package before uploading it. > >> And if we can get >> that with no effort at all, then why not. > > ... but can we? I value consistency. Fine, but the buildds are not identical. > I especially would value the fact that, barring architectural > differences, an i386 package is built by exactly the same process > that's used for an ia64 and s390 and m68k package. Thus, when somebody > reports a bug on those architectures, there is significantly less work > involved when I check whether that bug is reproducible, arch-specific, > or whatever. Not true. -- I'm sick of being the guy who eats insects and gets the funny syphilis.
Description: PGP signature