(resending to the bug)
Sean Whitton writes ("Re: Bug#886625: push-source should be usable no matter
the state of the working tree"):
> On Mon, Jan 08 2018, Ian Jackson wrote:
> > I think we should probably retain the current checks that the working
> > tree is up to date, to avoid accidents. But would could avoid running
> > clean. So we should check there are no un-ignored untracked files,
> > but not mind ignored ones.
>
> Right, makes sense to keep these because `dgit push{,-source}` is more
> expensive than `git push` in general, but there ought to be a switch to
> override it.
Exactly.
> >> -- though we need to think about whether `dgit build-source` too
> >> should operate only on HEAD, rather than the working tree as at
> >> present.
> >
> > Maybe. I think build-* are for the old package-ish workflow. (Now
> > that push-source exists I'm not sure why you would build-source,
> > really.)
>
> Maybe we should drop/deprecate it, for simplicity's sake?
Well, a note next to the build-source docs saying "you should consider
push-source" would be helpful.
Hrm, I have a counterargument to the position I just advanced above
:-). As follows:
dgit build-source is a useful tool for getting a .dsc you can feed to
sbuild or lintian or other tools that don't understand git. (I was
using it that way myself yesterday.)
One often wants to do that with a dirty tree. It would be useful to
be able to do it for refs other than head. (I think this is true of
push-source too.)
But maybe that should be "build-dsc" or "export-dsc" or something?
After all for these purposes one doesn't need a .changes.
Ian.
--
Ian Jackson <[email protected]> These opinions are my own.
If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.
