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

Re: [WIP] Some quirky parts of libsvn_wc/status.c:assemble_status() for retrieving revisions

From: Philip Martin <philip.martin_at_wandisco.com>
Date: Mon, 19 Apr 2010 19:12:01 +0100

Greg Stein <gstein_at_gmail.com> writes:

> If there are no changes (you're looking at just a BASE node), then you
> will *usually* have a valid revision. If the BASE node is incomplete,
> excluded, not-present, or absent, then you can't truly rely on its
> revision. (and we "should" probably set it to SVN_INVALID_REVNUM for
> those four cases; Philip?)

Excluded, not-present and absent are all really just names that are
implied by the parent being at a particular revision. One could argue
that they should have the same revision as the parent, but INVALID
works just as well, if not better. (I tried to make not-present nodes
have SVN_INVALID_REVNUM a couple of weeks ago, but it proved difficult
as some part of code spotted the INVALID value and "helpfully" filled
it with a value obtained from somewhere else.)

For incomplete I'm not so sure. In 1.6 an incomplete node has a
distinct revision and non-trivial content. As I understand the plan
1.7 is going to add a complete directory with incomplete children and
then convert each child in one step from incomplete to complete. I'm
not sure how that is going to work out in practice, whether it relies
on editor v2, how it will work with older servers.

Received on 2010-04-19 20:12:35 CEST

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