(and back to list...)
---------- Forwarded message ----------
From: Phyrefly <phyrefly.phyre@gmail.com>
Date: 21-Nov-2006 16:29
Subject: Re: Slightly less simple labels (PML part 2)
To: Tony Harverson <tony.harverson@iglu.com>
> > Can I suggest that the next step up from this is: "Working
> > man's labels" or WML would be on a per-file basis, so that
> > rev 3 of files A, B, and C would hold a label, but not files
> > D, E and F? (this would probably naturally include the
> > possibility of having the same label on a different rev of a
> > different file, I can't see an implementation that would
> > include the former and not the latter) Possibly replace
> > "file" with "folder" in that description, although I'd prefer not to.
>
> In the discussion we had regarding labelling, fixing bugs, and branching, I think you said that you wanted to allow branches/1.0.1/C to replace trunk/C in the label if it had become necessary to fix C for the 1.0 label. Is this what you mean by intention with "Different Rev of a different file"?
>
> Tony
No, I'm trying to add one new piece of functionality at a time, so we
can see where it makes sense to draw the line. Although, having said
that, if I can label file B at rev 2, and file A at rev 1, then
there's no reason I can't remove the label from file A and label
tags/tag_1/A at rev whatever. At this point, the branch holds the
label though, not the trunk.
However, when we start to combine labelling and branching in the same
implementation, I'd suggest that we would not want two versions of the
same file (whether that means a different rev, or a different
copy/branch/tag) to ever have the same label.
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Tue Nov 21 17:36:28 2006