On 11.02.2014 11:12, Ingo Weinhold wrote:
On 02/11/2014 09:38 AM, Axel Dörfler wrote:Just to repeat Ingo, I'd say we should determine *exactly* what we want to be in the release that isn't there yet. IMAP is not a good point, as we do have IMAP support -- it's just buggy (so "beta" ready) :-)This should be about the spirit of the "no new features" clause, not about it's exact wording. What we do want to avoid is adding/changing large code parts during a phase where we want to stabilize the code base. And AFAIU you're basically rewriting large parts of the IMAP implementation.
It adds a layer of complexity that cannot even be answered within the community. Aside from having to decide whether something is a bug versus a missing feature, what is really going on is that you now have to decide whether "addressing an issue" means to write much or little code. You could even take this a step further and investigate whether "writing some code" could affect other parts of the system in a substantial way, or whether that code is pretty self-contained. Actually, a one line change can affect the whole system substantially while a completely new, large component may not affect the system as a whole at all.
I don't see how given these conclusions we can actually determine (within even the limited group of committers) whether we can go into "bug fixing" mode. For every "issue", be it bug or missing feature, someone would have to know whether addressing it means it could potentially affect system stability. How would that even be possible? If you know beforehand that something has that affect, you would only ever write perfect code...
Best regards, -Stephan
