[FFmpeg-devel] [RFC] Some ideas for roundup

[Stefano Sabatini]

On date Saturday 2010-10-09 13:15:28 +0200, Michael Niedermayer encoded:
> On Sat, Oct 09, 2010 at 11:52:10AM +0200, Stefano Sabatini wrote:
[...]
> > * level of difficulty of the issue
> >   Many bugs are trivial to fix, and may be an ideal task for wannabee
> >   contributors (think at it like a sort of "FFmpeg love" tasks).  Also
> 
> Iam not sure here, easy issues will be fixed quickly normally so age is likely
> a good indicator for difficulty already. That said iam not opposed to add such
> a field i just have a bit a doubt that people will maintain it, i think

Yet there are trivial bugs which are left untouched for years and for
various reasons (we mostly work on issues in the free time, what we
care depends on the motivation and the interest we have in the issue,
not necessarily this is related to the difficulty of the task).

The main objective is to make easy for a wannabee contributor to spot
which are the issues on which she could try to work, this would be
also useful for tutoring projects (GSOC, GCI).
 
> >   the "estimated" level of difficulty may be useful for funding issue
> >   resolutions, which may be related to the time of the issue and to
> >   the level of difficulty.
> 
> Iam strongly opposed here, this has very high abuse potential and as said
> the age is already a natural indicator for difficulty
> 
> 
> >   Note that this item may be merged with the per-issue tags, since the
> >   level of difficulty may be implemented as a set of tags.
> 
> > 
> > * per-issue funding
> >   It should be possible for an user to fund the resolution of an
> >   issue. The money may be devolved to the fondation and re-imbursed to
> >   the donor if the issue is not fixed in some variable amount of
> >   time. The money may be assigned to the fixer and in part to the
> >   foundation (e.g. 80% to the fixer and 20% to the foundation).
> > 
> >   Here things may be tricky if we want to avoid conflicts amongst
> >   developers (e.g. two devs working at the same issue, or when it is
> >   not clear which is the "right" solution", also complex issues
> >   usually require much work which may be shared amongst several
> >   developers), so we have to consider if the cons prevail over the
> >   pros.
> > 
> >   (Another consideration is that bug-funding may lead to more bugs.)
> > 
> >   And to make more fun we should accept not only money but whatever
> >   the donor like (e.g. a postcard, a peluche, some HW, a box of
> >   special beer, potatoes, whatever but possibly not illegal.)
> 
> we can add a funding topic to issues where there is funding if people want

My idea was to create a new class representing the funding, then each
issue may have various properties (the funder, the amount delivered,
expiration time, etc.).
-- 
FFmpeg = Faithless & Foolish Minimal Pitiful Exploitable Gadget