branching in packages?

[Ilya N. Golubev]

> Semantic content: "I don't feel like reading the patch discussion, so
> put information that belongs with the patch discussion in text that I
> do read."

Unfortunately, the <m3u07pftkj.fsf at jerrypc.cs.usu.edu> message pointed
by cvs log message of `comint.el' sync does not mention cvs tags at
all, neither it contains

pre-JJ-comint-sync
post-JJ-comint-sync


And if their names were there, do they <belong> *only* there?  New cvs
tags are change in cvs itself, and done for purposes of some big
change in the same cvs.  So why pointers to them do not <belong> in
description of that change put in cvs, why *must* refer to mailing
list message?  The latter may be unaccessible for some time, because
all one has is local copy of cvs repository, and no network access at
all.  <The big ben-mule-21-5 check-in!> of 2002/03/13 08:52:41 shows
example of less error prone practice: stating cvs tags - not in change
log entry, but in cvs commit message.

After all, these pointers are certainly far smaller than detailed
description of code changes being committed - which are universally
considered belonging in cvs.


As for looking for these tags in other messages (except
<m3bqrovyq4.fsf at jerrypc.cs.usu.edu>) - it basically amounts to vgrep
in mailing list archives, and just impossible.  <xemacs-patches>, in
particular, contains information about all, even proposed, patches,
not only about ones already committed.