Re: [INTERM PATCH] Issue 571: A+D should be R in update

On 9/13/05, Malcolm Rowe <malcolm-svn-dev@farside.org.uk> wrote:> On Tue, Sep 13, 2005 at 03:39:33PM +0100, Max Bowsher wrote:> > Do we really need a hash at all? Are not all D+A combinations immediately> > consecutive, meaning that all we have to do is to queue exactly one delete> > notification at any one time, in case the next notification is an add for> > the same path?> > No - check the doc-comments for svn_delta_editor_t. All that's promised is> that a delete will occur at some point prior to an add.> > I've not really looked at this, but one things I was wondering: instead of> just adding a 'replace' notification type to svn_wc_notify_action_t and> generating it artificially from the individual D+A changes (as per the> current patch), would it make sense to also rev svn_delta_editor_t and> add something like replace_entry, so that a tree editor can indicate a> replace operation directly?
Yes, ofcourse, but that's a longer-term goal.
This issue is just about what's reported, there's another plan (maybenot in an issue?) called 'implement true renames' which includes yoursuggestion.

bye,

Erik.
Received on Tue Sep 13 17:13:56 2005