Re: regarding the resolution of issue 431

On Thursday, Jul 10, 2003, at 03:31 US/Eastern, Jacek Kopecky wrote:
>
> Granted, the optimization preservation across intermediaries is just an
> 'e.g.' in issue 431, but I like concrete issues, therefore I proposed 
> we
> close issue 431 by saying what seems to be the consent of the group,
> i.e. that intermediaries can indeed change what is optimized.
>
> Are you (and maybe others) saying that you disagree with the line above
> tentatively and wish to wait until we have the requirements?
>
I think its premature to close 431 until we have agreed on the 
requirements. We discussed some potential requirements on the call 
(e.g. preservation of attachment order) that have implications for 
intermediaries.

> Or are you saying that there may be a bigger issue we will see after we
> have the requirements? If this is the case, let's close 431, wait for
> the requirements and open issues stemming from them.
>
If you want to separate the example in 431 from the more general issue 
of intermediary semantics that's fine, but even then I don't think we 
can close the new issue for the example until we have an agreed set of 
requirements - how can we know that its OK for intermediaries to shift 
data between attachments and inline until we know the set of 
requirements we are trying to meet ?

Marc.

>
> On Thu, 2003-07-10 at 01:17, Mark Nottingham wrote:
>> To complete my AI;
>>
>> I think that some people still have doubt about this issue because we 
>> have
>> not completely identified the requirements that we're attempting to
>> address with MTOM. Specifically, it may be that MTOM enables 
>> optimisation
>> in with a variety of qualities, and therefore will need to be modeled 
>> as
>> properties which can then be used by intermediaries as input to
>> optimisation decisions.
>>
>> As a result, people are not comfortable with closing issue 431.
>>
>> --
>> Mark Nottingham
>
>
--
Marc Hadley <marc.hadley@sun.com>
Web Technologies and Standards, Sun Microsystems.

Received on Thursday, 10 July 2003 17:29:25 UTC