On 12/01/2010 10:50 AM, ext Andrew Flegg wrote:
I'm happy for the WhizBang tester to file a bug against WhizBang - but
once it's identified it's applauncherd (either by a WhizBang developer
*or* an applauncherd developer), a minimal test case SHOULD be created
anyway to add to the automated unit test suite.
This is indeed happening, and the test case is made public in the git
repository at gitorious. The issue is what happens in the public
bugzilla. The original bug was reported in the internal bugzilla and
marked as fixed there. But in order to sell our package update to MeeGo,
we would need to create an artificial bug at BMC so that we could then
close it in order to refer to it in the Changelog. This is what I meant
by us not being treated as proper upstream. Try suggesting kernel
developers or xorg developers that they would need to file bugs at
bugs.meego.com in order for their components to be updated in MeeGo.
You'd be laughed out of the door (if you were lucky).
The response to our submitrequests has been essentially "there has to
have been a problem in the MeeGo stack for it to be accepted in." We do
not do releases for fun, there is always a reason, and those reasons are
documented in the change log in the source tree. It is extremely
demotivating to be forced to spell out those reasons yet again, just in
order to fulfill what to us only appears to be bureaucratic overhead.
Come to think of it, I think the underlying reason for all this pain is
that bugs.meego.com mixes distro bugs and some upstream bugs.
--
Pertti
_______________________________________________
MeeGo-packaging mailing list
[email protected]
http://lists.meego.com/listinfo/meego-packaging
