[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn commit: rev 1562 - trunk/subversion/include trunk/subversion/libsvn_wc trunk/subversion/libsvn_subr trunk/subversion/libsvn_client trunk/subversion/svnlook trunk/subversion/libsvn_repos

From: Karl Fogel <kfogel_at_newton.ch.collab.net>
Date: 2002-03-20 04:35:27 CET

Greg Stein <gstein@lyra.org> writes:
> > All the suggestions are good, but this basically boils down to "Today
> > is config day, not wc day." :-). I'll address a few of them below,
> > that are directly related to the svn_io_check_path() change. That
> > change was made because I was writing new calls to svn_io_check_path
> > in the config code, and wanted to avoid generating *new* stringbuf
> > lossage. Old lossage is still lossage, of course, but at least no
> > more than was there before.
>
> Sure. But a commit is still a reviewable commit :-)

Absolutely! And I certainly don't disagree with anything in your
mail. Just wanted to point out that I hadn't any current plans to
address those issues. (Okay, and I'll confess, also was pointing out
that I was aware of the issues already. :-) ).

> > `npath' is used in other places in that function (and not just for
> > npath->data). Sure, they could all be changed, but... See above. :)
>
> They should be changed. Come on, Karl. I did an inspection and saw that
> 'npath' wasn't needed. And it was pretty clear that it wasn't used. If you
> want to challenge even the most obvious changes, then why should I continue
> with suggestions?

Huh? How is this change different from the other stuff you pointed
out?

`npath' is used as an argument to a recursive call (this is
svn_wc__ensure_directory). So if I change it, I have to change the
prototype of that function, which means changing all the other callers
of that function, which is precisely the sort of change propagation I
wanted to avoid (an avoidance with which you agreed). What am I
missing here?

I don't understand what you mean by this:

> I did an inspection and saw that 'npath' wasn't needed. And it was
> pretty clear that it wasn't used.

But at any rate, it's not an accurate description of the code.

There's nothing "obvious" about this change. It has the same costs as
the other changes we're discussing. So what's up? :)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Mar 20 04:27:15 2002

This is an archived mail posted to the Subversion Dev mailing list.