[saga-rg] Design Team meeting plans..

Tom Goodale goodale at cct.lsu.edu
Thu Nov 18 10:31:20 CST 2004


Hi,

In the call for volunteers we had

- day 1, 2:    general design questions and decide the scope of the strawman API
- day 3, 4, 5: define strawman API with above scope

and at the face to face at supercomputing (minutes to come soon) we looked 
over the use cases and it seemed that the most pressing parts of the API 
were job startup and secure communications between processes, with file 
transfer coming third.

I'd suggest that on the first two days we identify all possibilities from 
the use cases, turn these into a set of subsystems and objects and 
brainstorm how they interact, and over the next two days work out the 
details (e.g. methods/functions/...) associated with the top two 
subsystems.  Then use the final day for a review and to check that we can 
still satisfy the interactions identified on the first two days.

Does that work ?  Is there a better procedure ?

We should probably hold another phone call next week and discuss this in 
detail.

One thing which also came up in the face to face was that it would be very 
good to have some input from security experts on issues we may face with 
the API and its interaction with security systems, so we may want to have 
some discussion in the first couple of days, or on the last day, of such 
issues, although I think at this stage we should be concentrating on the 
functionality we need to support our user communities and not getting 
hung-up on other issues.  It's not clear if we need to explicitly deal 
with security issues at the level this API is aimed at;  at the F2F we had 
a suggestion that the phone call after the next one (i.e. the 1st of 
December) should be dedicated to a discussion on security.


On the accommodation side, we are trying to block a set of rooms at the 
faculty club here from the 12th to the 18th of December.  I should have 
details for you later today.

Cheers,

Tom

On Wed, 17 Nov 2004, Craig Lee wrote:

>
> All,
>
> A draft agenda will be essential.  We can bash it when we get started.  Also, 
> we should have:
>
> 1) A list of issues/topics on the table to be decided at the meeting
> 2) Homework to be done before the meeting, e.g., reviewing all the use cases 
> and other existing APIs
>
> If everybody does their homework and if we know the major issues on the 
> table,
> then we can use our face-time much more effectively!
>
> Also, Shantenu mentioned to me that he was going to pull in the relevant use 
> cases
> from OGSA.  That should help.
>
> --Craig
>
>
>
>
> At 02:42 PM 11/17/2004, Christopher Smith wrote:
>> Ok ... thanks for the clarification Tom. Do we have some kind of agenda 
>> yet?
>> Or at least an idea as to what the format of the meeting will be?
>> 
>> -- Chris
>> 
>> 
>> On 17/11/04 10:10, "Tom Goodale" <goodale at cct.lsu.edu> wrote:
>> 
>> >
>> > Hi,
>> >
>> > the intent was to meet from Monday to Friday.  If people arrive on Monday
>> > and leave on Friday we'll only have a three day meeting, which is 
>> probably
>> > not going to be very effective..
>> >
>> > Cheers,
>> >
>> > Tom
>> >
>> > On Tue, 16 Nov 2004, Christopher Smith wrote:
>> >
>> >> I plan to fly in on Monday and leave Friday ... is this what everyone 
>> else
>> >> has in mind? I'm assuming the meeting will start on Tuesday morning. I 
>> need
>> >> to nail this down real soon.
>> >>
>> >> -- Chris
>> >>
>> >>
>> >> On 11/11/04 07:41, "Shantenu Jha" <s.jha at ucl.ac.uk> wrote:
>> >>
>> >>> All: Mark your calendars - The design meeting is confirmed for the 
>> 13-17th
>> >>> December at LSU.
>> >>>
>> >>> Tom and I will send a more detailed note about what transpired at the
>> >>> SAGA-F2F (though only three of us were able to make it).  There will be 
>> a
>> >>> couple of conference calls in the run-up to the design meeting for 
>> agenda
>> >>> bashing etc.
>> >>>
>> >>> Tom/LSU folks: Who should people contact for help with local 
>> arrangements?
>> >>>
>> >>> More on this thread soon.
>> >>> Shantenu
>> >>>
>> >>>
>> >>>
>> >>>> So is there some kind of proposed agenda for this meeting (and is it 
>> still
>> >>>> planned for LSU)? I mostly need to start planning travel, and need to 
>> know
>> >>>> what days to fly in, etc.
>> >>>>
>> >>>> -- Chris
>> >>>>
>> >>>>
>> >>>> On 1/11/04 16:30, "Andre Merzky" <andre at merzky.net> wrote:
>> >>>>
>> >>>>>
>> >>>>>
>> >>>>> Quoting [Keith Jackson] (Oct 29 2004):
>> >>>>>>
>> >>>>>> I can do that week. Unless we hear otherwise by Monday afternoon
>> >>>>>> california time, can we just plan on this?
>> >>>>>
>> >>>>> Yes, fixing the date would be very helpful.  Works for me, in
>> >>>>> case I didn't say that, yet ;-)
>> >>>>>
>> >>>>> Andre.
>> >>>>>
>> >>>>>
>> >>>>>> --keith
>> >>>>>>
>> >>>>>> On Oct 29, 2004, at 2:03 AM, Andre Merzky wrote:
>> >>>>>>
>> >>>>>>> Hi all again ;-)
>> >>>>>>>
>> >>>>>>> There have been a number of date proposals for the planned design
>> >>>>>>> team meeting.  I am not really sure if there is any agreement by
>> >>>>>>> now, but it seems that December 13-17th is doable after all.
>> >>>>>>>
>> >>>>>>> Could we try to come to a final decision soon? :-)
>> >>>>>>>
>> >>>>>>> However, we would like to propose Berlin (aka AEI Potsdam) as the
>> >>>>>>> European meeting place (AG node) - Hartmut volonteered to organize
>> >>>>>>> and host the meeting.
>> >>>>>>>
>> >>>>>>> Are there other options for Europe?  Cardiff has been in
>> >>>>>>> discussion as well IIRC.
>> >>>>>>>
>> >>>>>>> Best regards,
>> >>>>>>>
>> >>>>>>>  Andre.
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> --
>> >>>>>>> +-----------------------------------------------------------------+
>> >>>>>>> | Andre Merzky                      | phon: +31 - 20 - 444 - 7759 |
>> >>>>>>> | Vrije Universiteit Amsterdam (VU) | fax : +31 - 20 - 444 - 7653 |
>> >>>>>>> | Dept. of Computer Science         | mail: merzky at cs.vu.nl       |
>> >>>>>>> | De Boelelaan 1083a                | www:  http://www.merzky.net |
>> >>>>>>> | 1081 HV Amsterdam, Netherlands    |                             |
>> >>>>>>> +-----------------------------------------------------------------+
>> >>>>>
>> >>>>>
>> >>>>
>> >>>>
>> >>
>





More information about the saga-rg mailing list