"Paul A. Steckler" <[email protected]> writes: >> > That helps a bit for some issues, but I think most of the bugs in the >> > async branch mostly relate to maintaining unstated or unknown >> > invariants in the implementation. >> >> Umm, I'm not sure I share that view, I'd dare to say that for a start the >> new protocol would allow to drop 90% of the code, and after all I don't >> see a reason things wouldn't work as long as you stay within the >> supported feature set. > > I hope that's so!
Note that the protocol supported by SerAPI in the async case is a strict subset of what the XML protocol offers; this is so as we have prioritized robustness vs advanced features. But even if the "async" feature set is not so complete, IMHO it should provide quite some speed advantages over the sequential model. Cheers, E. _______________________________________________ ProofGeneral-devel mailing list [email protected] http://lists.inf.ed.ac.uk/mailman/listinfo/proofgeneral-devel
