GGF12 OGSA Outreach session =========================== Sept. 22 (Wed), 9-10:30am, QC note: Hiro Kishimoto * Early discussion by Hiro GGF IP policy, sign-in sheet * Hiro goes over slides on WG status, v1 document comments summary, and v2 process * Ian goes over slides on roadmap document. Malcom: What does each green box mean on this "A Perspective on SOA Specifications"? Jay: This slide is presented at the last OGSA F2F meeting in SF bay area. It is one of our member's view and I have much mature version of this slide. Tom: Picture will become a little bit different if you put dependencies among specs. Jay: I already did a little about dependencies. E.g. WSRF and WSN sit a side each other. Question is; which is core for our activities? Tom: Ws-address is now submitted to W3C. Jay: Let's create short texts then stop there since nobody knows the answer. Let's create observational document in addition to roadmap document. Malcolm: Thinking about a path from architecture to specifications, some standards may become manure too rapidly and they can be totally useless. Ian: Another specs like CIM is also important for us. Tom: WS-I profile is also important. (The profile is already out but adoption curve is rather lengthy. Ian: An adoption curves are different in science and in industry (industry must be slow). Tom: We should make investment to stability. Jay: I would like to ask another SDOs what they think about these dependencies. We should also ask ourselves if we should have such dependencies? We can focus on low level and concrete specs on early stages. But they do not have not much functionality and are not sufficient for high level grid requirements. Ian: It is a discussion of "top down approach" v.s. "bottom up approach." Malcolm: I see a few waiting things. Cyclic dependency, Naming, etc. Ian: Priority is important. Tom: I think a sub-group should work on dependency/priority. Jay; WSRF and WSN are foundation for other functions (they are like instruction sets). And concept of OGSI are being preserved. C: GGF is the top of specs pyramid. Jay: GGF is pioneer. But since this space is highly politicalized, GGF cannot have over arching role here. A federation of SDOs will do this architecture things and GGF should strongly support this federation. Jay: OGSA-WG proposing 1) OGSA core, 2) design team, 3) WGs formation. How do you feel? Dane: Top-down and bottom-up will have gaps when they met. Identify them then fill and fix them. Malcolm: I think framework is manure. Big problem of the infrastructure is that definition of terms is different (e.g. name of terms). This is quite a distractive. It will slow down everything. Ian: How to co-work to go farther? Malcolm: Rename terms will take very long time and will ask DAIS to reorganize them. Jay: Does OGSA preclude or not preclude any term definition? Malcolm: I prefer waiting for completion of naming discussion. David Snelling: OGSA is facing a risk of stepping another one's toes. If we step on some toes, then we should simply get step back. Dave Berry: At the first time, data design team took time to understand each other, you can do this to make agreement. Jay: If you want to creating new WG, you should point to any activities relevant to grid activities. Dave Snelling: All activities should not be standards one. Propitiatory developments is enough, given bringing them to SDOs later. Hiro; GGF is usecase driven and it is our strength. Dave Snelling: High level SDO collaborations are necessary. So far SDOs collaborations are p2P relations only. Now process is started, things are happening. Jay: About common management model, it is shifting GGF focused model to generic one. Fred: Since CMM-WG does not want to do any duplicate efforts, thus we have merged with WSDM TC. Later today we will have a cross WG session. Jay: At the last F2F, WSDM has expressed their interest to naming. Andrea: WSDM is model agnostic. But CGS-WG has done very good job input to DMTF CIM 2.8.