[ogsa-rss-wg] OGSA-RSS Agenda Topic

[Philipp Wieder]

Hi Donal,

Donal K. Fellows wrote:
> Mathias Dalheimer wrote:
>> As Philipp already stated: This is a twofold presentation. The 
>> architecture topics are covered in GSA. In this presentation, they 
>> will give us more information about their interfaces and interaction. 
>> I think this will take approx. 25 minutes.
>>
>> But nevertheless, we have a timing problem. Next time, we should ask 
>> for two sessions. Can you tell me which specific topics you want to 
>> discuss? Then I can volunteer to create an agenda. The agenda so far 
>> is (just collecting, no order intended):
>>
>> (1) Service description milestone discussion
>> (2) Presentation: Scheduler Interfaces in the ASKALON Grid Environment
>> (3) Candidate Ordering Language discussion
> 
> I think the presentation has to go at the end. We're already one GGF
> behind our schedule (from our charter) and that means we *really* need
> to get on with it. Presentations to us are not our core activity; we're
> a working group and not a research group. As such, we can only really
> have time for show-and-tell stuff if we've covered our core business
> first. I'm sorry to be hard on this, but I think it's part of being a
> co-chair.
you convinced me. Looking at the topics you listed below I share your 
view that it will be even tough to discuss these. But there won't be 
neither a chance to introduce the group and its activities to newbies 
nor to have presentations like the one we are discussing. Therefore the 
remaining question is why the group did not apply for two sessions?

Best regards, Philipp.

> As for agenda, I think we've got quite a few things to discuss. I hope
> to have an abstract specification of a CSG (using my idea of CSG as
> itself an abstract concept) done by the time we go to Athens, and that
> should stimulate quite a bit of discussion. We then need to discuss:
> 
>  * Whether splitting CSG/EPS the way I'm thinking of is a good idea.
> 
>  * Ways of concretizing into WSRF, WS-Transfer (et al) and/or REST.
>    Especially, how many of these do we need to do?
> 
>  * What is the nature of a candidate document? This is *very* important
>    and I'm not convinced any more that at the abstract level we can
>    mandate WS-Agreement since that spec is bound tightly to WSRF.
> 
>  * Possibly some things about the COL, but my idea there is that that is
>    just a language that can be plugged in in an extensibility point in
>    the core specification, and not that it is the only possible thing.
> 
> I also expect many of these discussions to take longer than I expect. OK
> that sounds backwards, but it certainly fits my experience of standards
> work... :-/
> 
> Donal.
>