Re: RFS: webcpp (adoption, bugfix and standards/dh7)
Hello Jonathan,
On Wed, Nov 26, 2008 at 21:22, Jonathan Wiltshire
<debian@jwiltshire.org.uk> wrote:
> Hi Sandro
>
> On Wed, Nov 26, 2008 at 05:33:57PM +0100, Sandro Tosi wrote:
>> ehm, I personally like to have all the changes in the revision =
>> "current revision in debian"+1, so -6, and you have the opportunity to
>> "fix" this since.... your package FTBFS if build twice in a row. what
>> does this mean? take your source pacakge and uncompress it, then exec
>> "debuild -us -uc && debuild -us -uc", the second execution will fail
>> with
>> (snip)
>
> Hmm, I was encouraged before to ensure each upload is unique, but I
> guess it's preference.
yeah, it's preference: I don't like to have more entries for a single
"series of changes", between the version in the archive and the
proposed version. Both are technically correct, but may change from
sponsor to sponsor.
> (A direct upload to ftp-master requires uniqueness, doesn't it?)
When you upload a package to ftp-master, there are 2 possibilities,
it's ACCEPTED (hence included in the debian archive) or it's REJECTED.
In the latter case, you can reupload the same revision, since from
archice POV that revision never existed, so it's ok.
> I think it failed to build twice in a row because the paches were
> removed before distclean was called, but correct me if I'm wrong. I've
> changed it in debian/rules and it seems to be ok now. Diff attached.
mh, distclean is not a target of debian/rules; what I suggest is using
something like
clean: clean-patched unpatch
clean-patched: patch-stamp
<do the real clean stuff>
dh clean
you can take a look at [1].
[1] http://wiki.debian.org/DebianPackagingHandbook/PatchSystems
After this fast fix, all should be ok; sorry if it takes another
iteration, because I make it not clear enough from the beginning.
Kindly,
--
Sandro Tosi (aka morph, Morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi
Reply to: