Re: What's happening with codeville?

Current status is that it's in a "works well enough" stage. At least for me.
I'd like to find more time to work on it, but finding a chunk of time
recently has not been easy. There are some bugfixes which have been done
since the last release, so it's about time to push out a new one.

My plan for when I do find time is to do some internal code cleanup and move
away from berkeley db as the primary data store (it works fine as indexes on
the primary data and will probably continue being used). That should pave the
way for using merge algorithms.

The current merge algorithm works quite well for me in practice. I do a lot
of merging between a lot of branches at work and the only times you really
get into trouble are when you try to reorder code or move things between
files. My primary interest in playing with other merge algorithms is in the
presentation of conflicts. Codeville, in its current state, could be made to
present better conflict info in some situations. However, something like
precise merge could be coaxed into prividing virtual ancestors on a per-
confict basis. Kinda like individual 3-way merges for each conflict section.

Ross

On Sun, Mar 25, 2007 at 12:43:44AM -0400, Paul Chiusano wrote:
>    Hi,
> 
>    What is the status of codeville? The mailing list is dead, the last update
>    on the website was from over a year ago, and there's still a long todo
>    list... what's going on? Is the project still being worked on actively or
>    has it been sidelined for some reason? Has the precise codeville merge
>    been integrated into the existing implementation? (Also, are there any
>    other projects that use this merge algorithm?)
> 
>    Paul