Re: Tree-conflicts: HEADS UP: update failure with tc-siblings :0

Quoting Julian Foad <julianfoad_at_btopenworld.com>:

> Do you agree that the functions svn_wc__add/del_tree_conflict, as
> currently implemented, are bad and should go away (because they take an
> existing adm_access baton and return with it out of sync)?

I don't think so, because as soon as we find a tree conflict
during merge, we have to record it immediately. If we record
the conflict loggily, we have to run the log before the
callback returns.

On looking at the libsvn_wc functions called during merge
(svn_wc_add3, etc.), where files and dirs are actually added,
deleted, or changed, I see quite a few places where an entry
is written immediately or logs are written and run before
returning.

The adm_access is supplied by the callback driver (e.g.,
repos_diff), and is set up via svn_wc_adm_retrieve(), as in
the update code. In other words, merge doesn't refresh the
adm_access data.

The merge code runs logs and writes entries immediately
because the log-accumulator stuff is internal to the working
copy library.

I suspect the let-close-directory-run-the-logs policy in the
working copy library is a performance enhancement.

Steve

-- 
Stephen Butler | Software Developer
elego Software Solutions GmbH
Gustav-Meyer-Allee 25 | 13355 Berlin | Germany
fon: +49 30 2345 8696 | mobile: +49 163 25 45 015
fax: +49 30 2345 8695 | http://www.elegosoft.com
Geschäftsführer: Olaf Wagner | Sitz der Gesellschaft: Berlin
Amtsgericht Charlottenburg HRB 77719 | USt-IdNr: DE163214194
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org