Currently several things are in flux, such that they will not be
resolved by tomorrow, or even if they were, we wouldn't have time to
really test:
- APR's i18n support on Windows is still undergoing changes.
- Subversion's use of APR's i18n support still needs a kink or two
worked out, or so I gather from Blair's mail :-).
- Path canonicalization and case canonicalization are not
completed. I have an outstanding patch, but it has been shelved
for several days while we dealt with the commit bug fallout
(issue #797).
- Mike Pilato has started revamping the URI-escaping throughout
Subversion; it turns out to be a bigger task than we thought
because of the many places where a component is joined to a
"path" that is really a URI. (The component therefore has to be
URI-escaped at that time.)
All of these, except possibly the first, are pretty trivial code
changes; the URI stuff affects a lot of code, but it's all fairly
straightforward. The real issue here is having time to test. We
don't want to be "code complete" and then, bang!, just roll a tarball
from whatever we've got :-). We want to be code complete, and then
spend a day or two testing candidate tarballs.
So after talking with Greg Stein, we've put Alpha at Tuesday, with an
interim tarball today or tomorrow morning. Brane et al, who have been
burning the midnight oil to finish stuff by Thursday, I hope you are
not put out by this -- perhaps the breathing room is welcome to you
too (?).
That's all. Imagine where we'd be if Kevin Pilch-Bisson hadn't hit
Ctrl-C when he did... :-) (Answer: we'd be at Alpha, with a latent
data corruption bug that some new user would encounter and post to
Slashdot.)
-K
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Jul 17 22:01:46 2002