On 01/17/2013 02:59 PM, Julian Foad wrote:
> Ben and I recently started producing a functional and UI spec for local
> moves:
>
> <https://wiki.apache.org/subversion/LocalMoves>
>
> We need this in order to test the sub-features against a known goal, and
> we need it in order to produce software that has consistency.
>
> Please join in with the thinking and writing.
>
> There's quite a big conceptual change to be addressed, because now that
> we have local moves we can no longer assume that every path refers to "a
> node" and that every node has "a path"; now some nodes live at two paths
> in the WC (at one path "now" and at another path "before"). After trying
> to decide what various subcommands should do when given the path of one
> side of the move, and seeing that we have no consistent plan there yet, I
> wrote in the Wiki: "We need to go beyond simple path-based addressing now
> that we have 'move', because it affects two paths so we can't continue to
> assume a model where a path simply addresses 'the node' and a node has 'a
> path'. Idea: 'to' path identifies the move; 'from' path identifies any
> replacement at that path???"
>
> - Julian
Obviously, working against a well-defined goal is wise. And I suspect that
there are plenty of us devs around willing to help think though what that
goal should be. Certainly the move problem isn't new, nor has it only
affected just a few of us over the years. :-)
But in the interest of sheer practicality, I'd like to step back and ask a
tangentially related metaquestion:
Should we be further delaying Subversion 1.8 for this work?
What started back in, what, November or so (?) as what was believed to be
about a month's effort has continued for 2+ months and still seems to be
growing in scope without bounds. Now, granted, its precisely "bounds" that
you're trying to get out of this exercise -- and again, thumbs up on doing
so, how can I help, etc. -- but this is and has been for a little bit now
pretty much the only bit of coding of any reasonable degree of complexity
that remains on the critical path for an already prolonged release.
I propose that this work be deferred until 1.9 if it looks to drag on for
more months still.
--
C. Michael Pilato <cmpilato_at_collab.net>
CollabNet <> www.collab.net <> Enterprise Cloud Development
Received on 2013-01-17 23:09:55 CET