• To: "'Mark Baker'" <distobj@a...>
• Subject: RE: Using RDDL as a Distributed Registry Architecture
• From: Bill de hÓra <bill.dehora@p...>
• Date: Tue, 8 Oct 2002 15:32:23 +0100
• Cc: "'Roger L. Costello'" <costello@m...>, <xml-dev@l...>
• Importance: Normal
• In-reply-to: <20021008094336.J1778@w...>
• Organization: Propylon
• Reply-to: <bill.dehora@p...>

> From: Mark Baker [mailto:distobj@a...] 
>
> Incorrect.  It's a layered architecture, meaning that any 
> intermediary in a message path can respond with the 
> "document".  That's how caching works.

No, correct. It remains a fact that a resource representation comes from
an origin server in the first instance. The RDDL Roger's talking about
will come from an origin server. I had a qualifier about HTTP1.1 caching
in my first post, but cut it out, because caching is an optimization of
sorts. And web caching doesn't work well enough to counteract the
Slashdot effect or when software stops working because Netscape pull a
document of their server. 

Bill de hÓra 
--
Propylon
www.propylon.com 

 

• References:
  • Re: Using RDDL as a Distributed Registry Architecture
    • From: Mark Baker <distobj@a...>

• Prev by Date: Re: heritage (was Re: SGML on the Web)
• Next by Date: Re: RE: heritage (was Re: SGML on the Web)
• Previous by thread: Re: Using RDDL as a Distributed Registry Architecture
• Next by thread: Re: Using RDDL as a Distributed Registry Architecture
• Index(es):
  • Date
  • Thread