status of Milestone 8 (Subversion 0.8)

[Karl Fogel <kf...@newton.ch.collab.net>]

At this point, M8 consists of:

   Issue #518:  network layer needs to use diffs in both directions
   Issue #524:  eol translation and keyword expansion

Greg Stein is working on issue #518; I think he's pretty close, if not
done, but he can give more details on that (Greg?).

As for eol translation and keyword expansion, they are "finished" and
now in testing.  Since the potential for data corruption is unusually
high here, we're spending extra time on testing *before* making the
feature available in default client builds, and before upgrading the
server. :-)

Therefore, Ben Collins-Sussman and Mike Pilato are now doing intense
manual testing, while I'm working on the automated regression tests
(see trans_tests.py).  Please feel free to join in either activity!
And if you do, make sure to uncomment the

   #define SVN_TRANSLATE

in libsvn_subr/io.c for testing.  (When the feature is deemed safe,
we'll get rid of that #define altogether.)

See notes/line-endings-and-keywords.txt for brief descriptions of how
the features should behave.

-Karl

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: status of Milestone 8 (Subversion 0.8)

[Greg Stein <gs...@lyra.org>]

On Mon, Jan 14, 2002 at 06:16:03PM -0600, Karl Fogel wrote:
> At this point, M8 consists of:
> 
>    Issue #518:  network layer needs to use diffs in both directions
>    Issue #524:  eol translation and keyword expansion
> 
> Greg Stein is working on issue #518; I think he's pretty close, if not
> done, but he can give more details on that (Greg?).

I'm doing final testing now. I've got the code written on both sides of the
network and am just working through a couple bugs. I'm now getting something
about an error about "unable to completely write the svndiff to the parser
stream"; whatever that is about.

But: the client *does* send the right data, the server sees it and sends
back an svndiff properly labeled, the client detects that and tries to parse
the svndiff, then it blows up :-)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org