httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Veillard <veill...@rufus.w3.org>
Subject Re: W3C <-> Apache
Date Wed, 11 Feb 1998 20:21:46 GMT

> Such research is a great and good thing.  Confusing that research with the what
> is best for the next revision of the HTTP specification is not.

  I guess there is no confusion, at least for me. If there is some kind of
standardization process on HTTP-NG, this will clearly be distinct from HTTP.
 
> HTTP may need improvements, but not as desperately as it did prior to 1.1.
> 
> The complexity of these systems is pretty amazing.  Corba, COM, OLE, Active-X,
> ILU, IDL, SGML, XML, Java RPC, and Object Serialization, etc. etc.  Read the
> threading chapter in the Java specification.

  I completely agree ! And we definitely don't try to redefine completely
such a huge framework. 
 
> HTTP's success is it's simplicity and it's blessing. 

  I would say "was" it's simplicity ... Looking at the HTTP/1.1 spec, it's size
and the cost of an implementation. We actually hope that HTTP-NG specs will
be more organized and easier to implement in a clean fashion.
 
> > ... I have many good
> > reasons for why this might be a good idea but it takes time to actually be
> > able to evaluate it in practice.  ...
> 
> I'm glad to see that the W3C is looking at this problem, particularly since it
> is possible that they do appreciate the blessing of simplicity; but I doubt
> this research has much to feed into Apache 2.0

   Well, I was able to add ILU to apache 1.2 without too much trouble,
so plugin an external piece of code is always possible as long as the
interfaces and the layering are clean. Now, a "native" implementation is
somewhat different, and will certainly requires more work. I agree that
this shouldn't be considered a priority for 2.0 team, but keeping in mind
that a rather different protocol stack may have to be added in the existing
framework is a good idea. Basically if Apache-2.0 internal APIs are ready
for a MUX like multiplexing protocol, this is a solid basis for an NG
implementation, I guess.

Daniel 

-- 
Daniel.Veillard@w3.org | W3C  MIT/LCS  NE43-344  | Today's Bookmarks :
Tel : +1 617 253 5884  | 545 Technology Square   | Linux, WWW, Java,
Fax : +1 617 258 5999  | Cambridge, MA 02139 USA | badminton, Kaffe,
http://www.w3.org/People/W3Cpeople.html#Veillard | HTTP-NG and Amaya.



Mime
View raw message