Re: Etag-on-write, 2nd attempt (== IETF draft 01)

On Sep 12, 2006, at 3:49 PM, Julian Reschke wrote:

> Lisa Dusseault schrieb:
>> Actually, I'm proposing to narrow the scope by adding the  
>> requirement of working with existing deployed offline-cache  
>> clients, which will limit the number of solutions that will meet  
>> the overall set of requirements.  This doesn't require a new feature.
>> Lisa
>
> I'm really not sure what you're talking about. The draft spends a  
> lot of time explaining that ETags returned upon PUT do /not/  
> indicate the state of the entity sent by the client, but the state  
> of the entity as stored by the server. As servers are allowed to  
> (and indeed do) rewrite content, there's no generic way in HTTP to  
> avoid refetching the content, if octet-by-octet identity is  
> required (which I don't think is in many cases).

It's technically correct to say that servers are allowed to rewrite  
content.  However, existing deployed offline-cache clients assume  
that they do not.  Thus, it would be a seriously beneficial feature  
and possibly a requirement of new work to have backward-compatibility  
with that assumption, whether it was a wise assumption or not -- it's  
a deployed assumption.

Lisa