As I'm currently unable to beat you to anything due to the large geographical distance and other people might have the same problem, I tried to fix the issue and found this patch: https://issues.apache.org/jira/secure/attachment/12469996/0001-Fix-COUCHDB-1018.patch
- Klaus On Tue, 2011-01-25 at 11:03 -0500, Paul Davis wrote: > Yep. Looks like a bug. I'll get to it eventually if no one beats me to > it. Hint hint. > > On Tue, Jan 25, 2011 at 2:43 AM, Martin Hilbig <[email protected]> wrote: > > On 24.01.2011 22:30, Paul Davis wrote: > >> > >> I think more likely its that 12-12345 fails to parse as a correct > >> revision and the error gets confused as it bubbles up the stack. > >> > >> To verify, try replacing 12-12345 with a modified version of the > >> previous revision, ie, so it matches the (\d+)-([a-z0-9]{32}) regular > >> expression. > >> > > > > i changed the last bit of the revision and still get: > > > > {"error":"{not_found,missing}","reason":"{2,<<37,251,126,178,120,66,237,17,97,127,78,134,86,96,12,85>>}"} > > > > also klaus is right, i just stumbled upon COUCHDB-1018. > > > > have fun > > martin > >
