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

Re: Issue 730 related question about svn_ra_reporter_t interface

From: <naked_at_iki.fi>
Date: 2003-01-30 00:42:10 CET

Luke Blanshard wrote:
 cmpilato@collab.net writes:
 Checkout aren't restartable because when a checkout dies, some
 directory is left with an entries file that contains lies...

 Updates should never be in a dishonest state, though. If a
 directory doesn't get fully updated, it's revision still reflects
 what it had before the update (which was known to be complete and
 accurate)...
 
 But doesn't an interrupted update leave the WC in a dishonest
 state too? The WC claims to match its former revision, but it's
 partway through the update, and so its claim is a lie.

No - the _directory_ is still at the former revision, but some of it's
children may already be updated to the later revision - and those are
reported to the server, so the server doesn't have to send changes for
those files again. The directory revision is only bumped when all
changes have been transmitted, which makes sure that all changes have
indeed been transmitted and that all directory operations such as adds
and deletes have been carried out.

 Also, what's the state of affairs if update is interrupted while
 creating a new directory? What does that new directory's entries
 file look like? Whatever it looks like, couldn't you just drop that
 in at the top of a new checkout and then have checkout proceed as if
 it was the resumption of an interrupted update?

Right now, it looks a bit bad - so that's a bug that I intend to fix
when fixing this issue. Nevermind it for now.

-- Naked

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 14 02:24:36 2006

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