Web lists-archives.com

Re: Auto-update for sid? Auto-backport?




On Thu, Nov 16, 2017 at 07:06:39PM +0000, Simon McVittie wrote:
> On Thu, 16 Nov 2017 at 17:02:00 +0000, Holger Levsen wrote:
> > On Thu, Nov 16, 2017 at 05:53:40PM +0100, Wouter Verhelst wrote:
> > > Yes; and semver.org is a formalized system for version numbering stuff.
> > > If upstream has committed to it (and does not make mistakes), then the c
> > > versions in the above example MUST (in the RFC definition of that word)
> > > only contain bugfixes, and no interface changes.
> >  
> > oh shiny! thanks for pointing out semver.org!
> 
> Semantic versioning is not a panacea: it assumes that a developer can
> know in advance whether changes are a compatibility break or not. In
> simple cases where a bug fix or a new feature never causes any unintended
> regressions, that's easy; but if there were no regressions, then we'd
> all run unstable on servers.

There are certainly problems with assigning the correct
semver-compatible version number in some environments, as you point out.
That's also part of why there are multiple revisions of semver itself, I
would assume.

However, those are upstream's problems, not ours; and given that
upstream tries to do the right thing, we can infer from a version number
which changes are expected, and whether it's a good idea.  I could
imagine a policy which would do something along the following lines,
given a X.Y.Z semver-compatible version number:

- if X changes, automatically file a bug but don't do anything
- if Y changes, try to rebuild the package and run any tests (if
  available); post the result to the BTS, but don't upload.
- if Z changes, try to rebuild the package. If autopkgtests exist, run
  them; if they don't fail, upload. If no tests exist, build, but don't
  upload. If uploading, file an "automatic upload happened" bug in the
  BTS, marked release-critical, so that the package won't auto-migrate
  unless someone has at least tested the package in a real-world
  scenario.

There will still be risks; there are always risks. In this context, and
with that policy, however, I think that they are minimized and
acceptable.

(I could imagine that if the package was involved in a transition, the
release managers would also not like it if the package was
auto-uploaded; having the script that implements the above look at the
transitions configuration before trying any upload might be helpful to
avoid that)

-- 
Could you people please use IRC like normal people?!?

  -- Amaya Rodrigo Sastre, trying to quiet down the buzz in the DebConf 2008
     Hacklab