Re: [hybi] Serf support for BWTP

Justin,

Wow! cool!   Working code rules!!!

However, what I'd like to do with BWTP is to update it so that
it can be easily and efficiently sent over websocket (using
binary framing) or sent over it's own connection (more or less
as currently specified).

But currently I'm spending most of my time on Jetty's websocket
implementation (and using it in cometd), to make sure that we
have the best websocket implementation that there is... and
then we can still say that we hate it and that it is not because
we don't support it :)

cheers

Justin Erenkrantz wrote:
> Hi hybi@,
> 
> Out of some stuff that's been done with CREST [1], I got pointed at
> BWTP and decided to implement it in Serf [2].  A colleague of mine at
> UC Irvine is implementing BWTP with PLT Scheme for some CREST stuff
> he's doing.  (In short, CREST would benefit from having an async
> protocol.)
> 
> If ya'll haven't followed the Twitter conversation between me and
> Hixie over the last few days, I've pretty much come to the same
> conclusions regarding "Web Sockets" RFC that Greg Wilkins did much
> earlier.  (Kudos to Greg!)  I guess that Hixie has tried to write his
> RFC drafts in a way that stresses one particular way of writing a
> client - but for a pure async client library like Serf, the
> description in the RFC is pretty impenetrable...and, in my mind, not
> very helpful when it comes to writing an independent implementation.
> So, I found BWTP a lot easier to implement...and so I did.
> 
> The BWTP implementation in Serf isn't perfect and there's probably
> more to do, but it'll suffice for now.  Though there do seem to be
> some errors in the BWTP RFC, but I'm not sure if Greg is updating the
> BWTP RFC drafts or not.  If so, I'm happy to post about those apparent
> errors as well.
> 
> Anyway, Hixie asked that I subscribe/post to the IETF list, so here I
> am.  =)  -- justin
> 
> 1. http://www.erenkrantz.com/CREST/
> 2. http://code.google.com/p/serf/
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi