Re: Concurrency, again

[Stefan Monnier]

> - regexp scanning in a huge file that took so long my IM sessions would time
> out because the Emacs-based client code couldn’t respond to the server
> (which scanning also turned out not to be interruptible with ^G)

FWIW, this is an algorithmic problem.  The better to way to fix this is
to avoid backtracking in the regexp engine.

Also to avoid this problem, we'd need to spawn a new thread and since
regexp-matching is normally not expected to take a long time, that means
we'd need to spawn a new thread either "all the time" (just in case it
proves necessary), or "lazily" (i.e. if the current command takes too
long, then user can use a special handshake that spawns a new thread
and lets the current command finish asynchronously in that thread).


        Stefan