Ben Collins-Sussman <sussman@collab.net> writes:
> Chia-Liang Kao wrote:
>
>
> >>So am I missing anything obvious on why BASE:1 is a better choice?
> >
> >
> > I think BASE:1 is still the correct behavior from the perspective of wc,
> > though I find it a bit annoying myself too.
>
> There's been lots of debate about this in the past, you can search the
> list archives to read about it.
>
> The rationale for the current behavior is consistency. For any
> subcommand X, running 'svn X wc-path' will operate on the version of
> wc-path sitting in your working copy, i.e. the BASE revision. So when
> you run 'svn log foo.c', you're asking, "give me the change history of
> the BASE revision of foo.c".
It was more than just consistency -- in fact, consistency was a
benefit (in some folks' eyes) that just fell out of the real reason we
made the change, which was to solve the bug about the inability to run
'svn log' on something in your working copy which had been removed
from HEAD. By 'svn log' on the thing you asked for (the exact thing
in your working copy), that thing is guaranteed to exist.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Feb 18 14:06:16 2004