RE: Variant-ID proposal

I have strengthened my objection: I don't think it works for caches to
generate their own variant-ids.

Consider:
1. cache C assigns a variant-id V for entity E and passes it to a client
2. entity E is tossed out of the cache C
3. Client requests entity E, passing in V
4. Cache C passes the request to the origin server
5. V is meaningless to the origin server, because it didn't create it.
Worse, it might duplicate one that it did assign, and cause incorrect
results.

>----------
>From: 	Paul Leach
>Sent: 	Tuesday, April 16, 1996 1:31 PM
>To: 	'koen@win.tue.nl'
>Cc: 	'fielding@avron.ICS.UCI.EDU'; 'http-caching@pa.dec.com';
>'jg@w3.org'
>Subject: 	RE: Variant-ID proposal
>
>I object to caches assigning variant-ids. If they're needed for
>pre-emptive content negotiation by caches based on what is the
>Alternates header, then the origin-server should generate variant-ids
>for each alternate and return them in the alternates header.
>
>It'll be a lot simpler that way.
>
>>----------
>>From: 	koen@win.tue.nl[SMTP:koen@win.tue.nl]
>>Sent: 	Tuesday, April 16, 1996 12:31 AM
>>To: 	Paul Leach
>>Cc: 	fielding@avron.ICS.UCI.EDU; koen@win.tue.nl;
>>http-caching@pa.dec.com; jg@w3.org
>>Subject: 	Re: Variant-ID proposal
>>
>>Paul Leach:
>>  [Koen Holtman:]
>>>>If you have two orthogonal mechanisms, you can't have them share the
>>>>same space in the Cval: header.
>>
>>>This sounds to me like you're confusing protocol mechanisms and
>>>implementation mechanisms.  There are two orthogonal protocol mechanisms
>>>here, but in any server, I would imagine only one implementation
>>           ^^^^^^^^^^^^^                  ^^^^^^^^     
>>>mechanism, which would _indeed_ share the same variant-id scheme. Hence,
>>>I don't agree that the design principle you invoke applies to this case.
>>
>>I think the response I just sent to Roy explains why this reasoning is
>>wrong: there will _not_ be one implementation mechanism, placed in the
>>origin server, assigning variant-ids for responses of a particular
>>resource.  Such mechanisms can also be present in all proxies in front
>>of the origin server, because these will want to do preemptive
>>negotiation on behalf of the origin server.
>>
>>If only the server assigned variant-ids, we could allow it to mix the
>>data of the alternates-part and the variant-part into an opaque token
>>in an undefined way.  But this is not the case.
>>
>>>Paul
>>
>>Koen.
>>
>

Received on Tuesday, 16 April 1996 23:31:10 UTC