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

Re: Regression: Branch from working copy with modified files

From: Norbert Unterberg <nunterberg_at_gmail.com>
Date: 2007-03-24 08:44:25 CET

[Just sending a ping on this issue]

What do you think about replacing the
   previous=current-1
revision logic by something like
   previous=ancestor(current)
to fix the problem I reported (and maybe other possible problems too)?

Norbert

2007/3/19, Norbert Unterberg <nunterberg@gmail.com>:
> 2007/3/18, Simon Large <simon@skirridsystems.co.uk>:
>
> > > C:\xx\wc>svn log -v -r4 file:///c:/xx/repo/branches/test
> > > ------------------------------------------------------------------------
> > > r4 | Norbert | 2007-03-17 16:13:54 +0100 (Sa, 17 Mrz 2007) | 1 line
> > > Changed paths:
> > > A /branches/test (from /trunk:2)
> > > M /branches/test/dir1/test2.txt
> > >
> > > You see, the changeset before r4 is not r3 but r2, and subversion has
> > > actually recorded that information. Do you think you can somehow use
> > > that information to improve TortoiseSVN's log and merge functions?
> > >
> > > If I manually set the "From Revision" in the merge dialog to 2 then
> > > the merge also succeeds in TSVN. But the problem remains in the Log
> > > dialog because I can not manually select the "from revision" when
> > > double clicking on a line with a changed file.
> >
> > This may be related to SVN issue 2496 which appeared in 1.3.x. ATM none
> > of the SVN developers has the time or inclination to look into it.
> >
> > http://subversion.tigris.org/issues/show_bug.cgi?id=2496
>
> Interesting reading...
> I believe TSVN might be affected my that issue too, but i nmy case the
> "fixed" history tracing code in SVN works correct. The point is that
> test2.txt really did not exist in r3. it was created in branch@4 from
> trunk@2. trunk@3 already contained a modified version of that file
> that has nothing to do with the branch.
>
> I think the correct fix for all these cases would be to forget the "-1
> assumption":
> When you go into the revision log and select "show differences" on
> revision n, then TSVN tries to diff [n-1:n]. This assumes that n-1 is
> the ancestor n. But in some cases this assumtion is wrong. The way
> that should always work would be to ask SVN about the ancestor of n
> and diff against that: diff [ancestor(n):n].
>
> Norbert
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tortoisesvn.tigris.org
For additional commands, e-mail: dev-help@tortoisesvn.tigris.org
Received on Sat Mar 24 08:44:43 2007

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

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.