Sorry to be so off the cuff about this. What
I had in mind was the following:
follow our usual tracking process most of the time with changes shown
using word tracking.
once in a while we decide that enough changes have accumulated and decide
to make a clean copy. This is a revision number by itself that just takes
a spec (say version 1.35) and checks in the same spec with all changes
accepted (as version 1.36), stating in the row for 1.36 that this was a
clean copy revision. Which means the content was unchanged from 1.35.
each revision we have a new column that shows the clean copy revision that
it started from. Thus 1.37, 1.38 and 1.39 might all show 1.36 as the
origin clean copy.
Is that better?
From: Prasad Yendluri [mailto:email@example.com]
Sent: Thursday, September 02, 2004
To: Satish Thatte
Cc: Alex Yiu; Diane Jordan; bpel
Subject: Re: Fw: [wsbpel-spec-edit]
Done with changes mentioned in the last editors conf call
Satish Thatte wrote:
I have no problem with doing the accept-all-changes now if you want
to do it Alex. A clean copy would be nice. We have the history in
the previous versions. And we can make a note of that in the change log
in the document for the clean version.
If we archive the copy with the changes tracked as a specific milestone
then we can create a new copy, right? We can even create a separate new
clean copy sequence numbers log in the spec if we need to.
Sorry to be thick but,
can you be more specific on the process above?
On accepting changes, each version is archived separately in CVS but, we have
been distributing copies with change tracking in, so that people can see the
changes made. Now if we accept all changes that should be done on a version
that the TC has seen on voted up on (and accepted etc.), so that change
tracking is effective from that reference point. If we accept all changes at
this point would it confuse people? as we changed things substantially since
the last accepted version. Just wondering..