From: "Cliff Woolley" <[EMAIL PROTECTED]>
Sent: Friday, July 20, 2001 4:05 PM
> On Fri, 20 Jul 2001, Roy T. Fielding wrote:
>
> > Hey, what's going on with the retagging? There exists on developer dist:
> >
> > -rw-r--r-- rbb httpd 5383847 Jul 19 16:59 httpd-2_0_21-alpha.tar.Z
> > -rw-r--r-- rbb httpd 466 Jul 19 16:59 httpd-2_0_21-alpha.tar.Z.asc
> > -rw-r--r-- rbb httpd 3226188 Jul 19 16:59 httpd-2_0_21-alpha.tar.gz
> > -rw-r--r-- rbb httpd 466 Jul 19 16:59 httpd-2_0_21-alpha.tar.gz.asc
> > -rw-r--r-- wrowe httpd 4305440 Jul 20 13:20 httpd-2_0_21-alpha.zip
> > -rw-r--r-- wrowe httpd 477 Jul 20 13:16 httpd-2_0_21-alpha.zip.asc
> >
> > which means that there are now two different versions of the source being
> > called 2.0.21? That is bad.
>
> I agree. Ditch 2.0.21 and reroll 2.0.22 from scratch, IMO.
>From HEAD?
Cmon... we will never get to a point of stability until we adopt a completely
more thorough model. That means taking the 'good bits', or regressing the 'bad
bits', from the prior release.
If the difference between .21 -beta and .21 -release is three lines, but we have
a week of intervening patches to the code, then what's the point of falling all
the way back to an alpha.
Look at jakarta-tomcat. Four beta cycles on the same code base, barest changes
possible for stability, and they have a release.
I agree the lack of a subversion is a problem. That's what I think we should be
addressing. My theory would look something like this:
tag APACHE_2_0_x_alpha_1
stop look and listen. Is it a 'respectable' candidate? If so...
co -r tag APACHE_2_0_x_alpha_1
tag -b APACHE_2_0_x
since it's earned a release branch (no need to waste one if we are too far from a
given release). Now apply the sorts of small changes over the course of a week or two.
The group consensus and RM decides together if it should be ditched (more code changes
than are worthwhile), or can be promoted to -beta, -release.
cvs co -r APACHE_2_0_x httpd-2.0
would always give you the most recent patches to that version.
Thoughts?
Bill
