On Tue, Mar 12, 2013 at 03:31:14PM -0500, Todd T. Fries wrote: > Penned by Landry Breuil on 20130312 15:16.38, we have: > | On Tue, Mar 12, 2013 at 10:27:26AM +0100, LEVAI Daniel wrote: > | > On sze, m?rc 06, 2013 at 20:29:44 +0100, Landry Breuil wrote: > | > [...] > | > > That said, i'll recheck with the regular unflavored bitlbee, and with > | > > your diff. > | > > | > Have you had a chance to do some testing? > | > FWIW, upstream don't give a crap about it, so it's likely that this > | > would stay in the ports tree for a while (if accepted). > | > | I've done some testing, and yes it seems to work fine - with the > | non-flavoured libpurple i dont need to specify the server, whereas > | without your diff i'm pretty sure it failed to find it; I see the dns > | request for _xmmp-client._tcp.foo.com via tcpdump. > | > | So i'd be happy to take this diff.. iff upstream was more receptive. > | File a new bug and harass them :) > | Or as stuart says we could add the needed funcs to asr.. > | > | Tests and opinions from other bitlbee users out there ? > > I also need to manually specify the servers without this diff. > > Seems uber annoying.
Did you test the -libpurple FLAVOR ? afaict, it doesnt exhibit this problem. One would need to check how libpurple does the SRV resolution. > If adding the needed funcs to asr is possible, that sounds like > a clear winner in my opinion. If they're part of a spec.. otherwise, you know the rule :) Landry
