[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: cvs and gnats link?

From: Jesus Manuel NAVARRO LOPEZ
Subject: Re: cvs and gnats link?
Date: Tue, 18 Sep 2001 10:25:34 +0200

Hi, Kaz:

Kaz Kylheku wrote:
> >tracking system.   I've been looking at GNATS and would like to know if CVS
> >and GNATS can be linked.  From a developer standpoint, it would be most
> >convenient and less error prone to type the same information once rather
> >than twice.
> What would you ever enter twice? Bug reports go into GNATS, code goes
> into CVS.
> In your bug tracking system, you can make certain references to versions.
> For example, a bug can be marked as having been discovered in
> release 'toaster_1_3' and laster marked as having been fixed in
> 'toaster_1_7'. Not by coincidence, these would correspond to release
> tags in CVS.

Well... that's exactly what I think the original poster ment.
With your approach, *if* a bug is *really* discovered at toaster_1_3 and
fixed in toaster_1_7 *will be by coincidence*... since you cannot
enforce to be entered equal at CVS *and* GNATS.  The idea would be to
automagically tie, at least one way, GNATS to CVS, or CVS to GNATS, so,
for instance, if ticket for bug 1234 is meant to be tied to (procedural
assertion), let's say branch br_bugfix_1234, this data is either
automagically created, or created at GNATS side and then the branch
automagically created, or the branch created and through its name GNATS
database automagically updated.  The same would go with, for instance,
checkin comments and the like, so the report at GNATS side for ticket X
will be *really* tied to what it's meant for.  Once you ask a human
operator to intro the same info twice (or more), if you really end with
that info properly introduced will be by coincidence, you it is worthly
to avoid that situation as much as possible.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]