Re: Etag-on-write, draft -04

On Dec 8, 2006, at 6:06 PM, Lisa Dusseault wrote:

> Just one question from all this conversation I just caught up on  
> (didn't realize Wilfredo had posted to the main list).
> If a server doesn't return an ETag at all in response to PUT ,  
> isn't the client likely to assume that the server wants it to use  
> the Last-Modified value instead?  In which case today's clients  
> will all assume they already have the content (what they PUT)  and  
> never download what the server actually stored.

I hope not, since the Last-Modified is an entity-header and refers
only to the content of the response message.  ETag is a response
header because it is resource metadata, not message metadata.


Received on Saturday, 9 December 2006 02:48:02 UTC