Frank Heckenbach wrote:
Hi Daniel,I think it might be a good idea to get the view of the upstream developers on this patch,Do you know if this patch has been accepted upstream?No, I haven't reported it upstream, only via Debian.Sure. Feel free to do so. I've reported the bug according to http://www.debian.org/Bugs/Reporting which says: : Don't file bugs upstream : : If you file a bug in Debian, don't send a copy to the upstream : software maintainers yourself, as it is possible that the bug : exists only in Debian. If necessary, the maintainer of the package : will forward the bug upstream.
Following Debian Procedure is a Good Thing. The Reality is that DDs vary in the amount of time and inspiration they have to follow up these issues. The fix may be simple or it may involve additional complications, either way, it can take a frustratingly long time for the Maintainer to pass the patch on to upstream, for it to get accepted and then cycled back into a usable repository. This particular package is maintained by the Debian Multimedia Team, so we could move this discussion on to the appropriate mailing list if need be.
because as I understand it, jack_diplomat is undocumented and unmaintained. I don't think it even has a web page.I don't see how this is relevant. This is not about anything specific that jack_diplomat does, any program that calls jack_client_open() twice with different server_name (optional parameter) will experience the bug. BTW, please let's not play guilt games here. jack_diplomat, even if it may be unmaintained ATM which I don't know, works well. (And, AFAIK, there's no requirement for a program to have a web site. ;-) It's libjack that has the bug, so that's where it should be fixed.
Chill, it was clearly a misunderstanding. There _is_ a requirement for applications to have available source code, which needs to have a url of some kind. That usually works out to being a website.
And BTW, the problem and the fix IMHO are rather trivial. All the patch does it the end is to get rid of a misapplied static buffer.
It may be trivial, but the next step is to find out upstream's opinion on the matter, as Daniel says. It could be useful to confirm that the behaviour is the same with other, more familiar clients and also with the testing version (0.103.x). I would expect that to be the case. The main potential stumbling block is whether upstream agrees that this is a bug, libjack0 not allowing multiple connections may well be considered a feature, in which case a "won't fix" is likely.
cheers, tim