On Sunday, 19 November 2017 9:41:58 PM AEDT Adam D. Barratt wrote: > > Section 5.5.1 of the above seemed to indicate that I should do it > > that way. > > Did I misunderstand it or does the documentation need improving? > > Some combination. :-) > > You used reportbug to file the report - did it not ask for a debdiff?
I can't remember. > > I've attached such a debdiff. NB It has one thing that is not > > required (but > > is still handy) that is a build-conflicts against too-new versions of > > the SE > > Linux tools. This prevents anyone from accidentally building it on > > Testing or > > Unstable (which will be unusable). Obviously the package will work > > OK without > > such a build-conflict, unless you build it with the wrong packages > > installed. > > Technically, it's version-constrained build-dependencies, rather than a > build-conflict. Is that ok? > In any case, the diff you supplied has: > > +refpolicy (2:2.20161023.1-10) unstable; urgency=medium > > which obviously isn't what you're proposing using for an upload to > stable. I realise I said "a package", but the implication was that it > be a package that you could simply upload "as-is" if the diff was OKed. So if I upload the same diff with a better version number it will be OK?
