[E-Lang] Deadlock-free [was what is good about E?]

[Mark S. Miller]

At 01:28 PM Tuesday 7/17/01, Ken Kahn wrote:
>I'd like to understand this deadlock issue better. [...]
>The jargon definition of deadlock seems much broader than that of Marcs. The
>example given could happen in E, right?

Right.

>Languages like E seem to have some advantage over languages like Java
>regarding deadlock but I haven't seen a clear description of what it really
>is.

The clearest I've written so far is 
http://www.eros-os.org/pipermail/e-lang/2001-May/005287.html , in response 
to a question by Alan along the same lines.  I have no doubt that something 
clearer still needs to be written.

Btw, the claim we're making here applies to all pure event loop languages, 
including Actors, FCP, Vulcan, Janus, and Toontalk.  (Joule too, but with 
some surprising qualifications.)  E is simply the first such language 
positioned for widespread use -- by having an integrated full familiar 
sequential-imperative subset.  (The insight that there's no conflict between 
sequential-imperative programming and event-loop concurrent programming is 
due to Doug Barnes, the original creator of Original-E.)


        Cheers,
        --MarkM