[lnkForumImage]
TotalShareware - Download Free Software
Confronta i prezzi di migliaia di prodotti.
Asp Forum
 Home | Login | Register | Search 


 

Forums >

comp.lang.lisp

web serving: fork a lisp interpreter for each client?

polymorph self

8/5/2015 9:40:00 AM

Is this what paul graham did in viaweb when he fkaed continuations using closures and let lisp 'have a conversation' with the browser client?

This would then take care of utilizing 16 cpu machines automatically?

poof the consurrency problem gone at a slight memory cost?

Do any lisp web framworks or libs use this approach?
5 Answers

Marco Antoniotti

8/5/2015 12:42:00 PM

0

I think Graham forked 42 CL interpreters per client.

You can Google 42 to understand why.

Cheers
--
MA

polymorph self

8/7/2015 5:02:00 AM

0

On Wednesday, August 5, 2015 at 8:42:25 AM UTC-4, Marco Antoniotti wrote:
> I think Graham forked 42 CL interpreters per client.
>
> You can Google 42 to understand why.
>
> Cheers
> --
> MA

I read the hitchhker guide books, good stuff

Seriously though what do lispers do to handle many web clients? liek haskell happstack claims great performance, and google go does as well with goroutines, and erlang as well wit lightweight processes [actor model?], I saw some lisp libs for concurrency and some for parallelism but am not sure they can be used to handle many web requests at same time on say 8 cpu machines.....I noticed antiweb, teepeeedee2 and hunchentoot are able to?

polymorph self

8/21/2015 4:44:00 AM

0

On Friday, August 7, 2015 at 1:01:37 AM UTC-4, polymorph self wrote:
> On Wednesday, August 5, 2015 at 8:42:25 AM UTC-4, Marco Antoniotti wrote:
> > I think Graham forked 42 CL interpreters per client.
> >
> > You can Google 42 to understand why.
> >
> > Cheers
> > --
> > MA
>
> I read the hitchhker guide books, good stuff
>
> Seriously though what do lispers do to handle many web clients? liek haskell happstack claims great performance, and google go does as well with goroutines, and erlang as well wit lightweight processes [actor model?], I saw some lisp libs for concurrency and some for parallelism but am not sure they can be used to handle many web requests at same time on say 8 cpu machines.....I noticed antiweb, teepeeedee2 and hunchentoot are able to?

No one in about this one?
Forking an interpreter per client and then that lisp having a conversation with the client seems brilliant!! Paul Graham kicks ass!
Isn't this what ruby rails does?
I am still excited about cl-prevayler and then the above idea for scalable web apps of all sorts.
Paul Graham also mentions faking continuations with closures...What does that mean?

polymorph self

8/21/2015 4:48:00 AM

0

On Friday, August 21, 2015 at 12:43:51 AM UTC-4, polymorph self wrote:
> On Friday, August 7, 2015 at 1:01:37 AM UTC-4, polymorph self wrote:
> > On Wednesday, August 5, 2015 at 8:42:25 AM UTC-4, Marco Antoniotti wrote:
> > > I think Graham forked 42 CL interpreters per client.
> > >
> > > You can Google 42 to understand why.
> > >
> > > Cheers
> > > --
> > > MA
> >
> > I read the hitchhker guide books, good stuff
> >
> > Seriously though what do lispers do to handle many web clients? liek haskell happstack claims great performance, and google go does as well with goroutines, and erlang as well wit lightweight processes [actor model?], I saw some lisp libs for concurrency and some for parallelism but am not sure they can be used to handle many web requests at same time on say 8 cpu machines.....I noticed antiweb, teepeeedee2 and hunchentoot are able to?
>
> No one in about this one?
> Forking an interpreter per client and then that lisp having a conversation with the client seems brilliant!! Paul Graham kicks ass!
> Isn't this what ruby rails does?
> I am still excited about cl-prevayler and then the above idea for scalable web apps of all sorts.
> Paul Graham also mentions faking continuations with closures...What does that mean?

http://www.paulgraham.com/...
Did you use real continuations to save state?

No, we used macros to fake them in Common Lisp, as described in On Lisp.

Sebastian Christ

8/21/2015 9:30:00 AM

0

>>>>> On Thu, 20 Aug 2015 21:43:45 -0700 (PDT), polymorph self <jackvanc3@gmail.com> said:

> Paul Graham also mentions faking continuations with closures...What
> does that mean?

See On Lisp[1], chapter 20 and 21.

[1] http://www.paulgraham.com/o...