DAIS Telcon Minutes 12/02/04 ---------------------------- Minutes by: Mario Antonioletti (mario@epcc.ed.ac.uk). Attendees: Allen Luniewski (IBM) Simon Laws (IBM) Brian Collins (IBM) Mario Antonioletti (EPCC) Norman Paton (University of Manchester) Savas Parastatidis (University of Newcastle) Susan Malaika (IBM) Chris Kantarjiev (Oracle) Dave Pearson (Oracle) Vladimir Beytser (SDSC) Sastry Mallady (Oracle) Agenda ====== Review of actions arising from the Redwood face-2-face meeting Sematics and Metadata update ---- o Review of actions arising from the Redwood face-2-face meeting o Sematics and Metadata update Simon: have been doing a bit of work on this with Norman and Mario. Have been driving out the terms for service semantics. Have a few and also have a few that are orthogonal to this that manage these operations. State at the moment is that we have some slides that are not ready for circulation. Plan to document where we got to with the semantics in written form to be included in the spec while in parallel we will produce more coherent slides that can be circulated later. That's pretty much where we are. Are people happy with this? Norman: are people happy with this or do people want the slides to be circulated to the list? The disadvantage is that the slides are not self explanatory. Dave: circulating the slides might ensure people are happy with the concepts before writing anything down. Simon: they are very light at the moment ... people may not have enought time to go through them. Dave: do not want people to do lots of iterative work Norman: options are to bat out the slides as they are now for folks to comment or we wait a week and a textual version will then be available - we can then update the slides for GGF10. Dave: decision should be down to the people that are working on it my comment is on the work involved. Simon: my suggestion is that we carry on as we are doing now. Norman: Metadata ... means susan. Susan: what aspect of metadata? Norman: ... where things are with CGS? Susan: one of the topics dicussed at Redwoods was our interaction with CGS at the DMTF to try and describe databases. Databases were also described in the CGS working group. Try to get them to incorporate the bits of metadata that DAIS needs in the CGS metadata model. Have produced a draft document that a couple of people will look at before distributing more widely. Norman: has it changed much since we worked on this? Susan: sent out a draft to Larry and you (Norman) ... gave this higher priority than the face-to-face minutes .... combined your material with mine ... have 10 driving scenarios to encourage the CGS WG to prodcue metadata for DAIS .... proposed as something that they might want to work on .. Norman: what is the proposed process as this is going as an informational document that is going through to GGF. Susan: I think so ... pass it on to Larry so that they can start the work on extending the CIM database model to describe databases.... Larry can then start work on this.... the document is already in the right format. Norman: thanks for that ... anyone particularly interested to have a look at that? Dave: I would like to see it... Susan: you have have been sent a copy of this .... Savas: Was wondering whether simon was going to use this story for metadta ....perhaps better waiting until you publish it before reading it. Norman: there is also quite a lot of metadata in the core specs and you could use that for the mappings. You could used that for mappings in the WSRF context. An informational doc will arise that will be produced to motivate discussion. o Progress on spec documents Mario: Amy and I have been doing some work on the xml spec ... have been iterating through this ... the latest version is up on Grid Forge so we encourage everyone to go there, have a look and comment. Have also made some modifications to the relational spec in order to get it consistent with with the XML spec. I have passed this on to Amy and Susan to have a look at - should coordinate between Susan and Amy as to who has the edit token for the next iteration - I have also been doing some modifications to the core spec which I hope to pass on to Norman next week in order to get the content on semantics included. Susan: can write some more on the stored procedure for the relational spec. Mario: Amy will also be having a look at the relational spec. Dave: there is a difference in the xml and relational specs with respect to the way they edge into data management... there is some data management in the xml spec which is not present in the relational spec. Brian Collins brought this up at the f2f - there is an issue as to where we draw the line ... Susan: we intend to support DDL in the database ... Dave: I agree we need to support DDL but we need to find out where to draw the line ... we are supporting management .. there were diffs ... we have to make a decision as to where the scope ends. There are something in DDL that are needed by data access ... Susan: we need to make a statement in XML as to where data management starts .... Norman: Dave, as you have spotted, this can you write down a proposal and see if we can implement that proposal so we'll have that for the next GGF... Susan: are you available after this call? ... you, Mario and I can discuss this? Are we going to ask for a longer time to produce the specs? ... Norman: may need to modify the charter but we decided we would do this after GGF10 to see what the lay or the landwas ... we should then be in a position to revise what has to be done. o DAIS requested session for GGF10 on WS Resource proposal Norman: we have a plan ... Simon: plan is that we do work to investigate a common scenario and apply WSRF and WS-GAF and come up with a slide set that take us through the issues of doing this with the different models. Can scope this out now. Norman: Simon/Savas and Sastry would lead this and could have slides to lead the discussion at GGF10. You guys are going to have a telcon. Sava: we were going to use the telcon afterwards but Mario, Susan and Dave have usurped this ... Simon: we are going to have a telcon today... Sastry: start with simon's slides and the look at the sldies ... and we were going to add our set.... Savas: but Simon is going to revise the sldies now ... Simon: these are the things taht we need to decide, the technologies, the use cases to be considered and what the success criteria are to be... we could do that now ... Savas: we can do that amongst ourselves ... unless the others are interested? Dave: the focus of this exercise is to determine what is best for DAIS ... we need some use cases ... but also what functions and operation are required for DAIS and how these map on to WS-* and what needs to change in these or what needs to change in DAIS... Savas: but we can't concentrate on the whole of DAIS ... we need to look at sub cases ... Simon: sounds like a good idea... Norman: I thought that Sastry had seen your slides ... Simon: I did send them ... Sastry: checking for slides now ... Norman: Simon demonstrated how some of DAIS mapped to WSRF ... we could now take a couple of scenarios and see how they fit within these frameworks ... the things that matter to us are the scenarios ... have some existing scenarios that could be worked through ... and where there are things that can be mapped to WSRF, these could be mapped to WS-Context and then we could tackle the different concepts ... thinking of Greg's scenarios but there are some people that are working on an informational document that are trying to make a coherent scenarios for an informaitonal document .... Greg Riccardi and Mahadevan Subramanian ... not just the delivery scenarios ... they may be in a position to share a draft. As you are not trying to produce a document you'll be able to work on this after the 20th ... could take the scenarios from there or get an early draft ... Simon: it's not just the scenarios but also the proof points ... Sastry: looked at the sldies ... reading of them is how to use WSRF ... look at the value which each model brings to the party ... what is the use case scenario and what we want to achieve and try to cast this in existing technologies and what you cannot do with the existing technologies... Simon: we need to identify the added values ... Norman: in essence an agreement needs to be arrived as to how this going to be achieved ... my inclination is that you have a component in the doc - here are the concepts and how they are presented in the different frameworks ... and how these are implemented in different representative scenarios ... put the mappings first and then the scenarios as the scenarios could grow much more quickly than a rather relatively stable mapping ... Simon: before the mappings we need to find out the points we need to bring out .... then we build the points that we want to demonstrate as quickly as we can and then the rest of it can be done independently and the bits can be put together ... Savas: will have to work via email as I'll be away for about a month ... will not be at the office but will have acces to mail ... Simon: if we circulate a list of points that we can agree to in the first instance and the technologies that we wish to consider in the mappings ... what do we mean by web services, wsrf, etc .... The list of point can be evolved in parallel ... can do that and move on from there ... AOB --- Susan: curious if anyone is going to WS-Notifcation ... Simon: is Martin going? Mario: I will ask Neil tomorrow. Sastry: there is a legal requirement in order to go ... Susan: I would like to go .... there is a BOF in informatio dissemination at GGF10. Norman: we meet again in a week ... quite close to the document deadline. [Susan dropped off at this point] Dave Pearson will arrange a meeting so that the boundary between data access and data management can be discuss tomorrow: with Susan, Amy, Mario and Dave and Brian. Norman: deadline for documents is now 24/2 which gives us a fraction longer [As Susan drops off Dave offers the telcon to Simon, Savas and Sastry ... except that Sastry drops off :-)]