[cvsnt] Re: Increasing release with "commit -r" fails

Chuck Kirschman Chuck.Kirschman at bentley.com
Thu Aug 12 13:22:46 BST 2004


Changing revision numbers is convenient when you make a major change in
product.  Imagine that between V9 and V10 you rearchitected a large
chunk of your product and change your file format.  So just prior to V10
development (V9 branch created) you revision everything up to 10.0.  Now
when you are reading the log you can easily tell where the product break
occurred just by the revision number.  You don't have to wade through
hundreds of tags to determine where the split was.  This is even more
valuable when your up at version 12, and you are checking the history or
annotation of the file.  It is obvious which of the 2 file formats any
given change applies to, and where the break occurred.  We also used the
new base revision when we converted from PVCS to CVS; by setting all the
files to a consistent new version number we knew quickly whether the
issues we were seeing could be related to conversion problems.

That being said, changing the revision numbers is just a convenience,
and shouldn't be relied on for anything.  As Tony says, tags specify
releases, not version numbers.

The following command works in 2.0.41; I don't have the newer version
installed anywhere.

	cvs commit -f -r2.0 -m"Upping to rev2" PSSEARCH.MAC

chuck

-----Original Message-----
From: cvsnt-bounces at cvsnt.org [mailto:cvsnt-bounces at cvsnt.org] On Behalf
Of Tony Hoyle
Sent: Thursday, August 12, 2004 5:16 AM
To: cvsnt at cvsnt.org
Subject: [cvsnt] Re: Increasing release with "commit -r" fails


Axel Eckenberger wrote:

> number part of the revision number (i.e. from 1.x to 2.x). If the -r 
> switch is not used for this, what is it good for? Also the comments in

> commit.c

-r is IMO counterproductive and I've been tempted to remove it on more 
than on occasion.  It has no real use, but some people seem to like it.

> working on do not have this mechanism. In certain cases it migt be 
> appealing to bring the source up to a common revision number to show 
> that there is a

It might be appealing but it's still falling into the trap of thinking 
revision numbers are something that you should care about.

Tony

_______________________________________________
cvsnt mailing list
cvsnt at cvsnt.org http://www.cvsnt.org/cgi-bin/mailman/listinfo/cvsnt



More information about the cvsnt mailing list