httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Richards <>
Subject Re: $Id$ considered harmful
Date Thu, 26 Sep 1996 16:07:42 GMT
Paul Sutton <> writes:

> I'm afraid I cannot see any benefit from having that information in the
> file. The only versions that are put on public release are those with
> clearly defined Apache version numbers (1.1.1, 1.1b3, 1.2b1), which will
> be tagged in the repository. I've dealt with quite a few bugs and other

That's not true. Only a small minority of people who use the
inter-release versions have access to cvs.

> problems, and in none of those cases would having an internal version id
> in a file would have helped.

Well, I suspect you haven't yet had to deal with someone who has a
development copy from some date X that has a bug in it. The first
thing you want to do is find out what version they're running.

> On the downside, besides the CVS problems, the id's make every commit
> bigger and create unnecessary differences between files of differenet
> versions which hides the real code changes.

Ehh? We're talking a handfull of bytes here? If you want to remove
them then fine but I think it's a loss of usefull information and the
only reason for doing so would appear to be to accomodate third party
companies who are maintaining their own cvs repositories and are
having problems with importing Apache's $Id$ strings.

  Paul Richards. Originative Solutions Ltd.  (Netcraft Ltd. contractor)
  Elsevier Science TIS online journal project.
  Phone: 0370 462071 (Mobile), +44 (0)1865 843155

View raw message