Michael Sinz wrote:
> While I agree that the .svn directories should still be there and
> contain at least some format and linkage information, I have a problem
> with the statement that the root would not change.
I mean the root of the WC. (sorry :P)
>
> If the linkage information is relative to the .svn directory it is in,
> then yes, I agree with this statement.
That's what I meant.
> However, if the linkage information is an absolute path, then I strongly
> disagree. I many times end up having things be renamed (moved) from one
> location to another. And in networked environments, what is
> /home/mks/project on one machine may be /nfs/fox/mks/project on another.
I agree. There's nothing certain about the environment outside the WC.
>
> (While I like and very much has used the current severability of the WC,
> I also really want server-controlled configuration and inherited
> properties, both of which end up acting against severability.)
Like you said. I think it's a cool feature. But I have nothing against
firing a single command before severing :-).
I wanted to say that, because severability isn't possible anymore, the
structure of the WC is static. Which means that if you're 6 levels deep
inside the WC, you can still assume that the root of the WC is 6 levels up.
So you can take advantage of this by recording (at least) that you're 6
levels deep (or with a relative path pointing 6 levels up) to point to
the root of the WC.
This wasn't possible before because of (the possibility of implicit)
severability.
It's even possible to use this as some sort of transition phase before
working with a WC without the '.svn' directories in each directory.
Although I don't think that will be very useful :P
I hope I made myself clear...
Danny
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Jan 26 22:53:17 2007