[Python-apps-team] Bug#494889: fetching from static-http://... broken

Thomas Arendsen Hein thomas at intevation.de
Wed Aug 13 06:55:03 UTC 2008


* Jeroen van Wolffelaar <jeroen at wolffelaar.nl> [20080812 23:46]:
> Using etch (version 0.9.1-1+etch1):
> 
> [jeroen at mordor]/tmp$ hg clone static-http://buildd.debian.org/~jeroen/status
> requesting all changes
> adding changesets
> adding manifests
> adding file changes
> added 16 changesets with 30 changes to 10 files
> 10 files updated, 0 files merged, 0 files removed, 0 files unresolved
> [jeroen at mordor]/tmp$ 
> 
> Using lenny (version 1.0.1-3)
> [jeroen at eee]/tmp$ hg clone static-http://buildd.debian.org/~jeroen/status
> abort: 'http://buildd.debian.org/~jeroen/status' does not appear to be an hg repository!
> [jeroen at eee]/tmp$ 
> 
> This is not documented in the Debian changelog (the upstream changelog
> is missing), and even if this feature is intentionally discontinued, the
> hg command should report so and not fail with the "no repo found" error
> message.

The feature is not discontinued, there is "just" a bug.

The automatic tests did not catch this problem, as it only appears
when the server uses the old repository layout without the store
directory, compare e.g.
http://hg.intevation.org/kolab/kuudecode/.hg/ (old layout)
with http://hg.intevation.org/argh/.hg/ (new layout)

"hg clone static-http://hg.intevation.org/kolab/kuudecode/" works
while "hg clone static-http://hg.intevation.org/argh/" yields the
same error as you reported.

A quick workaround for you is to clone your repository and move this
clone back in the old place. Of course old hg clients won't be able
to clone the new layout via static-http, only the usual network
protocol stayed compatible.

Regards,
Thomas

-- 
thomas at intevation.de - http://intevation.de/~thomas/ - OpenPGP key: 0x5816791A
Intevation GmbH, Osnabrueck - Register: Amtsgericht Osnabrueck, HR B 18998
Geschaeftsfuehrer: Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner





More information about the Python-apps-team mailing list