[Interest] Bounties?

[Bernhard]

> Correct me if I'm wrong, but the only successful bounty systems I'm aware
of
> are for finding (security related) bugs, not fixing bugs or developing
features.

Well, those sites where not bound to a project. I think it makes a big
difference if such a bounty service is directly attached to (and maybe
controlled by) Digia. I experienced big difference in acceptance between
"yet another" service and the "official" service.

Also I could think of such official bounties more as a "super-vote". Being
an incentive for Digia itself to fix the bugs with higher priority. An
enterprise license is probably too expensive for me. But I could easily
invest some bucks into the most annoying issues to make them rise in
priority. This also would prevent "unhealthy competition" and "secrecy"
because the official Issue tracker could be used to coordinate efforts (by
assigning it to some person willing to resolve the issue and setting the
state to "In Progress" so no other people start working on it in parallel).
Still the community would be involved and could decide to fix the bug taking
the money as a bonus. 

As you said the current votes are pretty much worthless. But maybe such
"super-votes" aka bounties could be more useful. Actually I can't see much
harm in trying it. The worst thing that could happen is that it isn't used
enough. On the other hand it could fill the giant leap between non-customers
and enterprise customers.

-- 
Kind Regards
Bernhard Lindner


> -----Ursprüngliche Nachricht-----
> Von: Koehne Kai [mailto:Kai.Koehne at theqtcompany.com]
> Gesendet: Freitag, 13. März 2015 09:41
> An: Bernhard; interest at qt-project.org
> Betreff: RE: [Interest] Bounties?
> 
> > -----Original Message-----
> > From: interest-bounces+kai.koehne=theqtcompany.com at qt-project.org
> > [...]
> > In fact this thread was intended to talk about Bounties. Not for
> complaining.
> > Do you see a chance for bug tracker integrated bounties?
> 
> 
> The devil with bounty systems is in the details: Who decides when a bounty
> should be paid out? How do you ensure it's actually paid? How do you avoid
> races between multiple competitors for a bounty? What if multiple people
> are involved in a fix? Etc.
> 
> It also strikes me as something that might introduce unhealthy competition
> and secrecy in a process that should be as inclusive and cooperative as
> possible.
> 
> For some discussion about (shortcomings of) bounty systems see also
> 
> https://snowdrift.coop/p/snowdrift/w/en/status-quo-floss#bounties
> 
> That said, I agree that the current 'voting' system in JIRA is of somewhat
> limited use. Quite naturally, annoyances that  a lot of people hit (and
that are
> easy to search for) are up-voted, while other, more serious issues that
> however affect only a limited number of people might not get this
attention.
> If you've anything meaningful to add to a ticket it's IMO better to
comment
> on it.
> 
> Regards
> 
> Kai