Fri, 4 Feb 2000 19:13:35 -0800 (PST)
On Fri, 4 Feb 2000, Just van Rossum wrote:
> At 4:27 PM -0500 04-02-2000, Barry A. Warsaw wrote:
> >What Greg wanted, and what I implemented, was trimming the diffs if
> >they were larger than x lines, where x currently == 1000. In that
> >case, we trim the diffs to 40 lines. This doesn't help you if there
> >are tons of smaller diffs in a digest. I'm happy to tweak the numbers
> >if people want, but I'm not going to do any significant work on the
> >script (no time).
> It seems today's situation proved that this is not the right approach...
> Wouldn't it make much more sense to not send *any* diffs, but install
> view-cvs.py or whatever it's called on *.python.org?
As Barry mentioned, it is installed.
> Or otherwise: remove the feature, and reinstate it when someone has gotten
> 'round to implement a sane version.
Having the diff *actively* put in front of your face is much better than
needing to go look it up through a different mechanism (e.g. web browser
or the CVS command line). When a person gets a mail message with a notice
of change *and* the change, then they can review it right then and there.
I reviewed Ka-Ping's first change. Haven't done the second so far. I can
basically guarantee that wouldn't happen if I had to go and use a separate
tool on all those files.
Actively posting diffs is a great thing. It can definitely assist in
knowing that people will be reviewing the change. That size is needed for
a good review. More eyeballs, better software.
IMO, if the size bothers you, then unsubscribe. ViewCVS will show you when
things last changed. You can always go and look to see what has changed
within the past N days/weeks/whatever.
Greg Stein, http://www.lyra.org/