Re: tag REL8_2_6 not present in file xyz - Mailing list pgsql-hackers

From Magnus Hagander
Subject Re: tag REL8_2_6 not present in file xyz
Date
Msg-id 20080110155952.GG10183@svr2.hagander.net
Whole thread Raw
In response to Re: tag REL8_2_6 not present in file xyz  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Thu, Jan 10, 2008 at 10:36:24AM -0500, Tom Lane wrote:
> Alvaro Herrera <alvherre@commandprompt.com> writes:
> > I'm generating a diff between 8.2.5 and 8.2.6 and I get this:
> 
> > $ cvs diff -r REL8_2_5 -r REL8_2_6 > /tmp/8.2.5-8.2.6.patch
> > cvs diff: tag REL8_2_6 is not in file contrib/adminpack/README.adminpack
> > cvs diff: tag REL8_2_6 is not in file contrib/btree_gist/README.btree_gist
> 
> > Hmm, if I try it on non-anonymous checkouts, the messages are not
> > shown.  Perhaps it is a problem with the mirroring script?
> 
> Apparently.  They're definitely there in the master CVS, for instance
> "cvs log" shows me for README.btree_gist:
> 
> symbolic names:
>         REL7_3_21: 1.1
>         REL7_4_19: 1.3
>         REL8_0_15: 1.5
>         REL8_1_11: 1.5
>         REL8_2_6: 1.5
>         REL8_3_BETA2: 1.5
>         REL8_3_BETA1: 1.5
>         REL7_3_20: 1.1
> ...
> 
> I noticed during one of the betas that it's still true that a "cvs tag"
> operation doesn't propagate to the mirror at all (for instance, you
> can't see the tags in cvsweb) until someone does a commit of another
> type.  Magnus or someone thought he'd fixed that, buut apparently not.

Yeah, I broke something in the process, so Marc backed my fix out. I've
been planning to set up a sandbox to try it out properly befor eI puti t in
the next time, but I havent' had the time.


> Hmm, now I see the "missing" tags for README.btree_gist on the anon
> server too.  Did someone just sync this up?

We force a repository sync once a day, but the last time the sync ran was
at 02:52UTC, which is some 13 hours ago...

//Magnus


pgsql-hackers by date:

Previous
From: Gregory Stark
Date:
Subject: Re: flex/bison output wrongly created in the source directory
Next
From: Andrew Dunstan
Date:
Subject: Re: flex/bison output wrongly created in the source directory