Introduction
People
- Lancaster/Daresbury/ReDRess
- Rob Crouchley
- Rob Allan
- Ties van Ark
- Dan Grose
- Manchester/NCeSS Hub
- Leeds/MoSeS
Agenda
- Provisional Agenda
- Additional Agenda Items:
- Possibly, lets see how it goes.
- We should have also had the following on the agenda:
- Use of WRG and NW-GRID:
- Discuss joint use of 2 regional Grids for the e-Infrastructure project.
- This follows from useful discussions at AHM.
- Actual Agenda
- Whether/how MoSeS makes use of GROWL-related technologies (Sabre?):
- We at one time discussed running Moses via GROWL and then running Sabre via GROWL to demonstrate two apps.
- We might want to control this using a workflow interface.
- This would be a nice demo!
- How MoSeS can make use of workflow technologies:
- (This is why Rob Allan is part of this meeting.)
- May be related to the first item of discussion - see above
- Further progress on how GROWL and MoSeS and Grid-enabled BHPS all link up:
- Access to data is critical:
- This has been discussed a lot in the CQeSS project.
- The GEMS work of Keith Cole and Pascal is key to this?
- Discussion about which datasets we need:
- This feeds into Work Package 1.1
- Portal work:
- Initial discussion of requirements.
- We will probably follow this up with a visit to Leeds by myself and Xiaobo Yang when Andy Turner and Paul Townend are available.
- How to keep in contact and have regular developers meetings - AG, Skype, IM?
Documentation
- http://www.ncess.ac.uk/portal
- Junaid Arshad's MSc Thesis on Security for e-Social Science.
- Links to other Portal Resources...
- Gist of some pre-meeting emails:
- Pascal's experience with workflows tools has been somewhat limited to beta-testing the P-Grade workflow portal for the University of Westminster
- (nice tool which he appreciates a lot).
- OGSA-DAI is not compatible with BPEL or Taverna and only offers minimal workflows through its activity steering.
- However, OMII UK is keen to further push the the integration of OGSA-DAI with its other technologies (service discovery, worflow tools).
- Pascal and Andy have no experience of GROWL.
- Rob Allan is involved in a project with OMII-UK and will probably be looking at the Cardiff workflow tools in that context in addition to Taverna, Kepler and Pegasus:
- Probably only enough resource to do an initial set of tests and a report at this stage.
- OSS Watch Wiki: Integrating Sakai and Shibboleth
Notes
- Going through the agenda:
- Whether/how MoSeS makes use of GROWL-related technologies (Sabre?):
- Dan Grose asked for a summary of MoSeS activity and progress:
- I explained that I'm making an effort to put everything online on the world wide web to expose what we are doing:
- Advertising what we are doing is gaining interest from our target end user communities.
- We are not as yet getting technology developers finding us wth solutions to our problems:
- But we are at least trying to make these links in this and other ways!
- I gave a verbal outline of the demographic modelling work and described the application areas in health service planning, transportation modelling and housing research.
- Dan had many questions, which opened up some of the details of what is being done, particularly in respect of the Population Initialisation:
- This was probably useful background/revision/update for the others too.
- I explained that the NCeSS e-Infrastructure for the Social Sciences proposal was written primarily by Mark Birkin who had a slight misconception of what GROWL was.
- Mike Daw explained that there was some flexibility and that we do not have to develop something just because it is written in the proposal:
- What he would prefer is for us to develop something useful, some e-Infrastructure of benefit to the social science community.
- He and I have had discussions about this and I explained that I had believed that it was best for us to continue developing the GIS type parts of the work starting with GeoLinking, overlaying contextual data and styling issues.
- We at one time discussed running MoSeS via GROWL and then running Sabre via GROWL to demonstrate two apps.
- I think the applications were R and SPSS
- Is an interface to R what the Health users want?
- AFAIK it hasn't been asked, but it is unlikely!
- Talk of where the various interfaces were between the different organisations and the work they do.
- For Health and Social Care planners MoSeS aims to support them by automating the generation of reports they want in order to help in policy meetings.
- The reports are to include figures (maps, tables, charts) and summary statistics for a range of variables for usually at least two points in time (usually present and some time in the current planning horizon):
- The forecasts are provided under specific scenarios.
- It is expected that the differences between the scenarios and present will be of considerable interest.
- To automate the process we need to do it at least once.
- There is an on-going to and fro process where MoSeS produces reports and the PCT users ask for modifications.
- Once the process is automated and there are some useful report templates, the project may be over or it may be that it has only just begun and that this is recognised to be generally useful and resources are made available to develop it...
- JSR168 with WSRP is probably not enough.
- Security discussion
- Junaid Arshad's MSc Thesis on Security for e-Social Science.
- Oxford Sakai have done work on Shibboleth integration:
- Does MoSeS need it's own Sakai instance or can it use the NCeSS portal?
- Does GROWL have a job to play?
- Controlling the workflow interface?
- Integrating GROWL and OGSA-DAI.
- Using GROWL to access remote data
- Taverna into BPEL
- Rob Allan knows about some work on this.
- How MoSeS can make use of workflow technologies:
- I mentioned that I had gone through a SAW-GEO tutorial using ActiveBPEL recently, but that was my only direct experience of Workflow technology.
- MoSeS Archive
- This links with metadata work:
- What did you create and why and what can it be used for etc?
- Is this result useful to anyone else?
- How much resource did it take to compute?
- Further progress on how GROWL and MoSeS and Grid-enabled BHPS all link up:
- Access to data is critical:
- This has been discussed a lot in the CQeSS project.
- The GEMS work of Keith Cole and Pascal is key to this?
- 3a) Discussion about which datasets we need:
- This feeds into Work Package 1.1
- BHPS and Census:
- I am keen for an improved grid enabling of 2001 Census data.
- Rob Crouchley outlined a Labour Market Economist Use Case for the BHPS and General Household Survey integration
- Merge all files to generate work histories for individuals so as to look at labour market provision in different parts of the country.
- Portal work:
- Initial discussion of requirements.
- We will probably follow this up with a visit to Leeds by myself and Xiaobo Yang when Andy Turner and Paul Townend are available.
- How to keep in contact and have regular developers meetings - AG, Skype, IM?
- AoB
- Miscellaneous
- How will the new NCeSS nodes fit with all this?
- Do MoSeS clients need to have UK e-Science certificates to run simulation models?
- The hope is that this can be avoided.
- I asked and Rob Allan agreed to support a MASS WorkSite on the NCeSS portal :-)
- There seems to be a notion that the social scientist that we are to support are the technically challenged:
- I would argue that a better user group to target would be those which do develop programs and struggle to run simulations on the non-grid resources available to them.
- I believe this is a significant group that would benefit by becoming engaged.
- Both Rob Crouchley and myself can be regarded as social scientist and we are reasonably technically competent.
- Although it is selfish, I would like much of the portal development to be geared to supporting us and drawing in our closer colleagues rather than trying to collar the masses who might not be even interested.
- At least people building (complex) simulation models of social systems can gain some tangible benefit by learning about and engaging with Grid technology.
Action List
- Andy Turner
- Arrange meeting in Leeds with Xiaobo Yang, Rob Allan, Wei Jie, Junaid Arshad, Paul Townend and possibly others.
References
![[Validate XHTML 1.0]](http://www.geog.leeds.ac.uk/people/a.turner/images/valid-xhtml10.png "W3C XHTML 1.0 Validation")
![[Validate CSS]](http://www.geog.leeds.ac.uk/people/a.turner/images/vcss.gif "W3C CSS Validation")
![[An image of Andy Turner]](http://www.geog.leeds.ac.uk/people/a.turner/a.turner.png)