Re: ELisp futures and continuations/coroutines

[Ted Zlatanov]

On Mon, 23 May 2011 11:59:02 -0300 Stefan Monnier <address@hidden> wrote: 

>> Stefan, Chong, anyone else, please let me know if you have objections to
>> the attached future.el package.  I will add error handling and add it to
>> Emacs tonight if there are no objections.  It's pretty simple code.

SM> IIUC this will only be used by URL, at least at first.  Is there any
SM> other package which is likely to use it in the foreseeable "future"?
SM> If not, then I'd recommend to name it "url-future" for now.

The idea from the beginning was that this is a true data type, not an
auxiliary library.  At least deferred.el and fsm.el will use it, and I
am in favor of adding deferred.el to Emacs as I mentioned.

I would rather not isolate its scope to just URL interactions because
rewriting url-future-* to future-* later will be annoying and because it
makes the package look like it has narrow utility.

SM> PS: I'd expect future-finish and future-errored to signal an error is
SM> the future is already done-p.  And future-call should return the
SM> future-value.

Thanks, I made those changes, see attached.

Ted

future.el
Description: application/emacs-lisp