On 06/10/2013 11:28, Colin Watson wrote: > On Sat, Oct 05, 2013 at 02:53:33PM +0200, Jonas Smedegaard wrote: >> Feels to me that it would be better to fix this at its core instead of >> covering over it by use of bogus build-dependency. I am sure you've >> already considered that option, and would appreciate your elaborating >> more (I saw and tried to follow your conversation on irc, but failed to >> understand it there). > > I don't think artificial build-dependencies are a particularly > inappropriate fix. > > Any change to britney to try to have it promote only architectures that > worked would amount to accepting permanent technical debt in unstable, > which I think is very poor design. Package maintainers should generally > be trying to avoid dependency breakage in unstable where they can as > well as in testing (or, to put it another way, accepting permanent > dependency breakage in unstable makes it harder to see the wood for the > trees), and by far the simplest way to do this is to fix the packages > rather than blaming the infrastructure for pointing out a real problem.
(I have nothing to say about the general case) for the particular case of node-* packages, it is true that when test suites are or will be run, they will have to build-depend on nodejs. So adding a build-dependency on nodejs for each arch:all module doesn't seem so bad to me today. Jérémy. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
