Re: [hybi] Issue tracker openness

[Maciej Stachowiak <[e-mail]>]

On Oct 6, 2010, at 11:24 AM, Ian Fette (イアンフェッティ) wrote:

> On Wed, Oct 6, 2010 at 11:19 AM, Willy Tarreau <[e-mail]> wrote:
> On Wed, Oct 06, 2010 at 10:58:22AM -0700, Ian Fette (????????????????????????) wrote:
> > Currently, only a subset of people have access to open a new issue on the
> > issue tracker. I personally find the issue tracker a very nice tool to use
> > so that we don't lose track of things, especially when it comes to requests
> > like "Please merge X to Y" or "Please add the following text to the spec".
> > Would anyone be opposed to opening up the issue tracker such that anyone can
> > file new issues?
> 
> While it may be useful, it could also end up as we sometimes see on some
> issue trackers, where discussions only happen there instead of on the list.
> The advantage of the list is that people must defend their arguments.
> Sometimes it may become tempting to open a bug when you feel like nobody
> understands your concerns on the list.
> 
> So I'm not completely convinced by the of advantages vs drawbacks.
> 
> Regards,
> Willy
> 
> 
> I don't think that opening an issue obviates you of the requirement to defend your arguments. It's possible that some discussion may move from the list to the issue tracker, and whether that's desirable is an open question (that we could perhaps try to correct for), but I don't see why the fact that the discussion is happening on an issue would mean that there is a lower bar to meet in terms of justification. Frankly, in all the projects I've ever worked on, having an open bug tracker where discussion happens on that bug tracker makes the discussion much easier to follow.

I have had positive experiences with open access to issue/bug trackers for both software projects and standards projects. Allowing anyone to track an issues seems to significantly reduce the risk of comments getting overlooked. It does take some social norms and monitoring to avoid bad patterns, but then, so does a mailing list. It also works much better, in my opinion, for simple editorial comments or reports of trivial errors - it is highly desirable to have those tracked more formally than just email, but not worth the time of the Chairs to enter each one in the tracker, or of the WG to discuss each one in detail.

It seems not to be the norm in the IETF community to let anyone raise issues, but perhaps it would be worth an experiment in this WG.

Regards,
Maciej