It would be great if scrivener had support for git based syncing. It would allow easier collaboration through existing popular collaboration services such as GitHub or BitBucket. It would also open possibility to use custom git server/repository.
Having a book or paper on GitHub will allow not only nice collaborative writing, but also branching of versions, review of contributions, comparison of changes, corrections from people not having access to the scrivener directly from github and much more.
The services above offer free accounts too and are easy to set up even for non-tech people. Those a bit more tech skilled might run their own repositories.
The trouble with this is that Scrivener’s UI could not handle this well at all, mainly because of the structure involved in the binder. Looking at collaboration is on the list for the future, but more as a 4.0 or 5.0 thing because of the complexity and resources involved.
+1 for collaborative features. Nothing else out there which has collaborative features approaches Scriveners awesomeness for everything else. I don’t even care if the collaborative feature set includes Git functionality. I just want to be able to collaborate with someone else in a sane manner. Real-time collaboration would be sweet, but I’d settle and be happy with just being able to save our project on Dropbox or Google Drive and then “check out” the project whenever it is my turn to work on it–and then check it back in when I finish my work. Right now collaboration with Scrivener is a drag (as it is with most programs).
As I say, it’s on the list, but not for version 3.0, it’s something we’ll revisit with version 4.0 or 5.0, off in the future. It’s an incredibly difficult and complex feature that could lead to data loss and huge problems if done incorrectly, and would take many months of work working on that and nothing else. We have things like snapshots and date stamps, which means that the main text can fairly easily be brought back in with minimum data loss, and collections could be automatically created to show users what had been changed, but it’s the other stuff that is harder: notes, synopses and such have no equivalents of snapshots, but more importantly, the binder structure makes things incredibly complex. It would be very difficult to merge two completely different hierarchies, so one would have to replace the other, which could lead to orphan files that would then need bringing back in… Scrivener 1.x and 2.x also have certain problems with the way their underlying file formats are designed, in that documents don’t use unique identifiers but numerical IDs, which works well for individual projects but not for this sort of thing. That will be changed with the next major version, which will at least remove that obstacle for future thoughts on this.
All the best,
Keith