Re: Java and HTML and well known socket numbers

Paul Burchard (burchard@cs.princeton.edu)
Sat, 1 Jun 96 22:36:55 -0400


Message-Id: <9606020236.AA21654@cs>
From: Paul Burchard <burchard@cs.princeton.edu>
Date: Sat,  1 Jun 96 22:36:55 -0400
To: MegaZone <megazone@livingston.com>
Subject: Re: Java and HTML and well known socket numbers
Cc: www-html@w3.org, www-talk@w3.org

MegaZone <megazone@livingston.com> writes:
> The majority of the designs anticipated using Java are
> more properly handled when viewing Java as a new resource
> type instead of linked hard with http/html.

There is no "hard link" between Java and HTTP -- or for that  
matter, between HTML and HTTP.  The whole point of a generic  
transfer protocol like HTTP is to handle "new resource types"  
without requiring a completely new protocol for every application.

It's true that currently HTTP is suboptimal for transferring large  
numbers of small files (this is *not* a Java-specific problem), but  
HTTP/1.1 will alleviate the problem by standardizing persistent  
connections and pipelined requests.

> Most critically the concept of clients sending applets *to* a
> server to be executed on the server - this is outside the
> scope of an httpd.

So PUT and POST are outside the scope of HTTP?


Russell Holt <holtrf@destinyusa.com> writes:
> What about a java ORB?

Now *this* makes some sense.  For example, ILU will also be  
switching their "string binding handles" to URL syntax.

--------------------------------------------------------------------
Paul Burchard	<burchard@cs.princeton.edu>
``I'm still learning how to count backwards from infinity...''
--------------------------------------------------------------------