[illumos-Developer] Fwd: Bug ID's in hg commit comments

[Garrett D'Amore]

Btw, I made an executive decision on this.  As we don't have access to
regular code drops from Oracle anymore, Oracle's bug IDs matter less to
us.  Also, it turns out we can't easily fix the existing committed
entries in hg.

So, we will use untagged bug ids until we reach the point of collision
with Oracle CR numbers (i.e at 1 million bug ids).  When we hit that
point, we'll figure out a solution to the collision that may occur.  For
now it is sufficiently far off that I'm not worried about it.

	- Garrett

On Sat, 2010-08-14 at 17:09 +0300, Cyril Plisko wrote:
> On Sat, Aug 14, 2010 at 4:25 PM, Michael <michaelsprivate at gmail.com> wrote:
> > oops, sent to OP only by accident....
> >
> >
> > ---------- Forwarded message ----------
> > From: Michael <michaelsprivate at gmail.com>
> > Date: Sat, Aug 14, 2010 at 3:24 PM
> > Subject: Re: [illumos-Developer] Bug ID's in hg commit comments
> > To: Mike Gerdts <mgerdts at gmail.com>
> >
> >
> > On Sat, Aug 14, 2010 at 1:53 PM, Mike Gerdts <mgerdts at gmail.com> wrote:
> >
> >> Would it make sense to have commit messages that look like:
> >>
> >> io:12 libc breaks lex
> >> Reviewed by: matt at greenviolet.net
> >> Approved by: garrett at nexenta.com
> >>
> >> The added "io:" is shorthand for "illumos.org".
> >
> > good idea, but wouldn't the chance of confusing "io" with "I/O" be a
> > little too high? I'd suggest using "illumos" - just 5 letters more :-)
> >
> 
> ILLUMOS:XXXXXX Would be very easy to parse unambiguously at the
> expense of a couple of extra bytes.
> (speakng out of genunix.org experience)
> That scheme is also very easy to expand to other issue id authorities
> if need will arise.
> 
>