The architecture for Wave is really difficult to get a handle on, though.
Relative ease of use is a major benefit of ShareJS, though I can't fully vouch for its algorithmic correctness. I've been meaning to contribute some additional tests to the project.
I know - but having developed an OT library and read the ShareJS code thoroughly, I think there MAY be 2 edge cases where conflict resolution is not symmetrical for some data types. Or I could be totally wrong.
As I mentioned, I hope to contribute tests to show that everything's peachy. Not trying to talk trash about this great library.
I'd be keen to know what those edge cases are. For JSON OT in particular I'm not entirely convinced it's correct (I wrote it), but it IS highly tested and fuzzed.
Comments
See also: Google Wave, open-sourced as Apache Wave.
http://incubator.apache.org/wave/
The architecture for Wave is really difficult to get a handle on, though.
Relative ease of use is a major benefit of ShareJS, though I can't fully vouch for its algorithmic correctness. I've been meaning to contribute some additional tests to the project.
The engineer behind ShareJS worked on Google Wave for two years.
I know - but having developed an OT library and read the ShareJS code thoroughly, I think there MAY be 2 edge cases where conflict resolution is not symmetrical for some data types. Or I could be totally wrong.
As I mentioned, I hope to contribute tests to show that everything's peachy. Not trying to talk trash about this great library.
I'd be keen to know what those edge cases are. For JSON OT in particular I'm not entirely convinced it's correct (I wrote it), but it IS highly tested and fuzzed.