Re: [ogsa-wg] INFO a hole???

[Hiro Kishimoto <hiro.kishimoto@xxxxxxxxxxxxxx>]

Thanks Steve for your notes,

I very much appreciate your efforts to liaise INFOD to OGSA again.

My comments in line <HK>.
----
Hiro Kishimoto

Steve Fisher wrote:
> Hi,
>
> I have heard a number of people complaining this week, at GGF, that
> "Information is a hole". It was also on the slides shown by Dave
> Snelling on Thursday morning.
<HK>
We are *not* complaining, instead asking for help from experts.
</HK>

> Previously Abdeslem Djaoui liaised between the INFOD-WG and OGSA-WG
> but now we have no direct means of communication and I suspect that
> some people in OGSA have not completely understood the potential role
> of INFOD, nor is INFOD sure of the size or shape of the hole it might
> fill within OGSA.
<HK>
We lose Abdeslem, he was an "information service" design-team lead and
liaison to INFOD WG.

We believe we do not underestimate role and importance of INFOD.
However, from our OGSA architecture point of view, information
service should also cover higher level services, e.g. discovery,
directory, and registory which will use INFOD for their transport means.

At this moment, each Grid system has its own proprietary information
services, e.g. Globus MDS, GMA, UNICORE... But there are no standards
specs for them.
</HK>

> The INFOD spec has been simplified significantly - or rather we have
> decided to concentrate on a base specification before moving on to
> more advanced features. We plan to have the base spec going for public
> comment in June, but it would obviously be better to understand very
> soon how it fits in to the big picture - to avoid having a
> specification which does not meet the perceived OGSA needs. How do you
> suggest we move forward?
<HK>
We are more than happy to talk with INFOD folks on how to fit your work
in OGSA architecture. We would very much appreciate if you or your
colleage could take care of these higher level information services in
addtion to INFOD?
</HK>

Thanks,
----
Hiro Kishimoto