Re: ACTION-928: progress registration of the X-Device-*

We have been there already:

I thought we had already decided not to define and 
register yet another set of HTTP fields because nobody
(developers, vendors of proxies, operators) is asking for
them, and their introduction raises serious problems -- for
doubtful benefits.

Registering header fields and then explicitly stating in the
CTG that they must not be used or deployed is a 
possibility, but then how does one justify registering fields
whose utilization is prohibited in the corresponding 
standard? This is highly questionable -- especially from the
viewpoint that this is a technique to squatter the name
space of HTTP and hinders other organizations that might
have legimitate claims to rely upon that name space for
their own standardization activities.



Received on Friday, 10 July 2009 19:34:09 UTC