On 2013-11-04, jpacner wrote: > Hi Holger, > > you're entirely right with my misuse of 'high-quality'. I should have > quoted it. The submitter himself would be responsible for the quality. > The point of this suggestion is that patches would be incorporated > faster, but on the other hand they could be much faster abandoned > (because the users will provide almost instant feedback about the issues > they experience with these fast-incorporated patches).
That assumes that a significant number of users re-build mutt after every patch, and that they use mutt in such a way that reveals issues with the patch, and that they can attribute an issue to a particular patch. In my experience, none of that happens, and certainly not all together. There is a tendency among some developers, especially those who write buggy code, to claim that a bug can't possibly be theirs because they don't think they did anything to affect the now-buggy behavior. That means that someone else has to take it upon themselves to debug the problem and determine who caused it. In the mean time, all the developers who are trying to be good citizens by regularly re-building mutt are stuck with versions of mutt that don't work unless they revert to the previous working version. It's better to verify the quality of the patch first, even if that slows patch adoption. Regards, Gary
