Re: [RFC] UI to show a summary of problems after merge etc. (incl. tree conflicts)

[Julian Foad <ju...@btopenworld.com>]

On Tue, 2008-10-21 at 16:35 +0200, Neels J. Hofmeyr wrote:
> 
> Julian Foad wrote:
> > On Mon, 2008-10-20 at 22:08 +0200, Daniel Shahaf wrote:
> >> Julian Foad wrote on Mon, 20 Oct 2008 at 15:46 +0100:
> >>> [[[
> >>> $ svn st
> >>> C       A/D/G
> >>> ?     C A/D/G/rho
> >>> D     C A/D/G/pi
> >>>
> >>> $ svn st --conflicts)
> >>> ? C (D>M) A/D/G/rho
> >>> D C (M>D) A/D/G/pi
> 
> Wow, I didn't actually know we have this! That's very nice.
> 
> >>> ]]]
> >>>
> >> What additional information does the 'C' column provide?  (It's already
> >> clear that a node is in tree conflict because it has the little arrow.)
> > 
> > In this case, it doesn't provide more info. The idea is that the first
> > three colums repeat the "Text status, Prop status, Tree conflict status"
> > fields from a normal "svn status", and then the part in parentheses
> > gives more info.
> > 
> > Maybe the tree-conflict 'C' column is redundant, or maybe in the general
> > case where different kinds of conflict shown it would be useful.
> > 
> > - Julian
> > 
> > 
> >>> Here, "(D>M)" means "an incoming Delete onto an existing Modified node".
> >>>
> >>> In this patch, this indication is coded on the assumption that it's a
> >>> tree conflict, but I feel the command ought to show all kinds of problem
> >>> mentioned above.
> 
> What would normal conflicts look like? Let's have a mixed example:
> 
> 
> $ svn st --conflicts
> C         A/D/zeta
> ? C (D>M) A/D/G/rho
> D C (M>D) A/D/G/pi
>  C        A/D/G
> 
> Well, it looks like the third C is really redundant, since tree-conflicts
> have the whole bracket thing while others don't.
> 
> How about compressing it to this, gaining four columns of space for path
> strings:
> 
> $ svn st --conflicts
> C     A/D/zeta
> ? D>M A/D/G/rho
> D M>D A/D/G/pi
>  C    A/D/G

See my other reply just now in this thread.


> Btw, isn't this one the wrong way around:
> 
> ? D>M A/D/G/rho
> 
> The ? says rho is unversioned (?), so I guess the D took place in the local
> working copy. I guess, `update' wanted to modify it. So shouldn't that be
> M>D?

No, that's an incoming Delete onto a local Modification, shown as local
state '?' because... I dunno, this is probably a broken example. It's
just printing what the callback says, but maybe the meanings aren't what
they seem.

>  Then again, what's this:
> 
> D M>D A/D/G/pi

This one's scheduled for deletion, and there's an incoming modification
that we can't apply. Looks fine to me.

- Julian



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org