Re: Status of Bugzilla Bug 10, Round-tripping various information in properties

Geoffrey M Clemm wrote:

> [...] I have not yet heard a "good" alternative to solving this 
> problem other than recommending that servers preserve namespace mappings.

The above captures my take on the issue as well, and there seems to be 
rough consensus around this recommendation. AFAICT, the only thing 
remaining to be decided is whether the spec says SHOULD or MAY -- My 
vote is for SHOULD, because it will be more beneficial to clients if 
more servers tend to implement this feature in the long run. It is my 
opinion that a MAY level requirement in this situation will lead to more 
special casing for client writers (i.e. 'am I talking to a server of 
type A or a server of type B?') and, hence, interoperability issues.

Geoff supports the requirement but is indifferent wrt SHOULD vs. MAY.
Lisa is OK with SHOULD, but would be fine with MAY.
Jim is fine with SHOULD.
Wilfredo also seems to be okay with SHOULD, as it still leaves some 
wiggle room for server implementers.

Given the above I'll move to declare rough consensus on this issue and 
for the inclusion of the latest proposed text in the next draft.


Cheers,
Elias

Received on Friday, 23 December 2005 20:50:07 UTC