Re: Comparing source members

Probably not... the sequence number and the change date are very unique to the iSeries, and I doubt anybody in the Subversion team has considered stripping out the first characters of every line (but I might be wrong). It may be possible, but not without tweaking Subversion or its configuration on your server.

On the other hand, I have rarely found that the "false positives" are a significant issue for me.

-----Original Message-----
From: wdsci-l-bounces@xxxxxxxxxxxx [mailto:wdsci-l-bounces@xxxxxxxxxxxx] On Behalf Of Adam Glauser
Sent: Friday, December 21, 2007 4:36 PM
To: wdsci-l@xxxxxxxxxxxx
Subject: Re: [WDSCI-L] Comparing source members

Luis Colorado wrote:

Vern, you wrote:

The RSE Extensions are next best for me, but they strip
the source sequence and source change date. Sequence is not such a
big deal, but change date can be. That will be a loss, in my view,
when going to editing in IFS and using subversion, etc.

Subversion doesn't use the RSE extensions. When you use Subversion (and I assume that the same applies to CVS) you keep the sequence number and change dates. The merge is really easy to use.

I haven't been stripping the dates/sequence numbers when committing my
source to Subversion, but I had thought that I would want to if I
started versioning all my source instead of just my own work on a
particular project.

The reason I'd want to strip the date/sequence is to avoid "false
positives" when comparing two versions. I'm thinking of the sort of
thing that Vern mentioned earlier, in which I the sequence numbers or
dates of some lines change without the actual code changing. Is there
some way to tell the Subversion server to ignore the first X bytes when
calculating the diff?