Tim Bray wrote:
> [snip]
> I think we can agree that the server you describe is engaging in bad
> practice.  Are we discussing whether it's worthwhile putting language in
> the spec explicitly to discourage it?  -Tim
> 
> 

-1.  The server is free to do whatever it wants with the data inside the
entry.  A server dropping an unknown/undesired extension element from a
posted entry is not necessarily a bad practice.  The server may simply
not be capable of retaining extensions it does not understand.  That's
life.  All that we should be concerned about here is on determine
whether or not an entry was created or not (regardless of it's specific
content).  If the server chooses to reject the entry, it should do so
with a proper http status code (likely some 4xx code) rather than return
some 2xx code for everything.

- James

Reply via email to