DS6 Planning Stage 01
The DS6 Stage01 Planning meeting was held on Thu 24th Feb 2005 in Leicester.
Participants
TonyLinde,
AlasdairAllan,
ThomasBoch,
BobMann,
FrancoisBonnarel,
GiuseppeLongo,
UgoBecciani,
JamesGray,
FabioPasian,
MartinHill,
EduardoGonzalez,
GarrySmith,
KeithNoddle
Agenda
[*] Those who did not attend the
Edinburgh meeting are advised to
have a look at the presentations made there.
Presentations
Presentations are linked from the agenda above.
Notes
Notes here refer to items raised in discussion during/following the presentation with the matching number above:
[1] VOTECH Science Planning (
EduardoGonzalez)
- Demarcation of DS areas. Eduardo's summary of the science drivers arising from the AVO Science Reference Mission, and from the wrap-up of the AVO Science Working group, revealed the difficulty of mapping activities on to the list of Design Studies. In many cases it was not clear whether a particular activity belonged in DS3, DS4 or DS6. It is clear that the staff effort available to DS3 will not be sufficient to cover everything that could possibly be classed as infrastructure, so it is down to each DS to look at these boundary cases and decide whether they are of sufficiently high priority within their own area to be worked on by their own team.
- Specific suggestions for tools. Particular requirements identified by the AVO SWG included tools for the visualization of multiband image datasets, the iterative fitting of SED models to multiband photometric datasets and the running of statistical tests. In the last area, it had been suggested that the R package might be a good one to target. R does include a lot of statistical functionality, and some of it has been implemented for the VO in the VOStat project, but it is not well known amongst astronomers. Many more astronomers are used to using the statistical routines in IDL, which does have licencing implications, although the astrolib library is now available through Python.
[2] DS6 Intro (
BobMann)
- Client-side tools. One of the things that DS3 is planning to develop is a client-side infrastructure, allowing client-side tools to interact with the rest of the VO: this will be of relevance to DS6, where some client-side tools are likely to be used.
- Science Advisory Group. There is no "VOTECH Science Advisory Group", as such: the Euro-VO initiative has a Science Advisory Group, which is a group of external scientists who set priorities, test cases, etc for the Euro-VO; and there is the VOTECH Science Team, which is a group of scientists internal to the project - one from each DS team.
- Deadline for DS6 Study Report. The rescheduling of VOTECH to being a four-year project means that the DS6 Study Report is now not due until Month 21 - not Month 15, as originally scheduled.
- Private VO. A concern was raised about the release of preliminary versions of software: to do so is a good thing, if people can play with them and feed back comments and further requirements, but there is a risk that the release of half-baked software could get the project a bad reputation. One solution to this would be to run a restricted VO, made accessible only to project members and a limited number of tame beta testers. This would provide a good environment within which people could experiment with new software. It could be made realistic by having a registry which harvested from external VO registries - so that external resources would be available - but which was not harvested by other registries, so that test software would not be exposed to external users.
- Long-term plan. One of the goals for the end of Stage01 is to have some level of long-term plan for DS6, which can be put to the Technical Advisory Panel.
- Validation of tools. It would be good to have a section of the wiki in which the assessment of tools can be reported. Some of this work may cross DS6 boundaries, so it may be a good idea to set this up as a cross-DS activity from the outset, with a link from the VOTECH top page, rather than from one of the DS pages.
- Choice of tools to develop. There is clearly a tension in DS6 between the need to elicit generic requirements on the VO infrastructure from data exploration (which motivates the study of a range of tools) and the necessity of developing tools to a point where people can use them in sufficiently realistic scenarios as to yield those requirements (which motivates the development of a limited number of tools, given the finite amount of effort available). The balance obtained between these competing forces must be kept under review as the project develops, but it is clear that the selection of a small set of tools cannot be made before we have a better picture of the science requirements in this domain.
- Data federation. In the discussion of the remit of DS6 the question was raised whether data federation tools well within its scope or that of some other DS. The general concensus was that, conceptually, DS6 should probably start with having data, rather than getting data, so it would seem more natural for data federation tools to be developed elsewhere. However, if the work of DS6 is impeded by their not being developed elsewhere, it might be necessary for DS6 effort to be devoted to this topic: this is something to monitor.
- Getting requirements. Requirements to drive DS6 work can come from a number of sources, but it was felt that the VOTECH Science Team will have a particular role to play here, since they comprise VO-savvy scientists, who should have a good idea of what is technically feasible as well as scientifically desirable. It was suggested that it would be good to get requirements by presenting a series of options to the Science Team and other bodies, since choosing priorities from a list would be more likely to succeed than simply asking for ideas.
[3] CDS interests in DS6 (
FrancoisBonnarel)
- Relation to Aladin. It has not yet been decided how to implement the new functionality being developed as part of the on-going visualization collaboration in Strasbourg. Some may be added to Aladin, while others may be called from Aladin, via CGI or web services: the prirority is on delivering the functionality in a VO-compliant fashion, rather that on doing that via a particular platform.
[4] Stage01 Intro (
BobMann)
- Staff resources in Stage01. Several of the institutions in DS6 - namely Catania, CDS, Leeds and Napoli - have not been able to hire into their new DS6 posts yet, while several of the UK team members will largely be occupied with the AstroGrid roll-out for the next couple of months. So, in addition to ClivePage and most of GarrySmith, the DS6 team effort will mainly be contributed by staff not directly employed on the VOTECH project, as follows: Napoli - GiuseppeLongo (50%), Angelo Ciaramella (50%), Roberto Tagliaferri (30%) and Maurizio Paolillo (30%); Catania/Bologna - UgoBecciani (50%), Marco Comparato (100%), plus ~40% of someone from CINECA; CDS - contributions from various people.
[5] Napoli Stage01 Plans (
GiuseppeLongo)
- Parallel tracks. It is intended that the integration of Astroneural and VisIVO will be undertaken in parallel with work on the VO-compliance of Astroneural.
- Astroneural without the GUI. It is intended that the routines within Astroneural should be available via non-interactive means, but this is not yet operational.
[6] Catania Stage01 Plans (
UgoBecciani)
- VisIVO release. The first stable release of the VisIVO tool will be ready in a few week's time, and will be badged as a DS6 Stage01 deliverable.
[7] CDS Stage01 Plans (
FrancoisBonnarel)
- Extension to SIA/SSA. CDS want to develop extensions to the existing SIA/SSA standards, which allow the user to view collections of heterogeneous data products in a way that makes clear the relationships between them. It is intended that there should be working drafts of these extensions by the end of June, and it would help if Use Cases for the scientific exploitation of such heterogeneous collections could be defined.
- ExtApp. Aladin interacts with external tools, such as VOPlot, through an ad hoc protocol, called ExtApp. This was chosen in preference to using SOAP, because that was thought to be too slow.
- Aladin and VisIVO. CDS and the Catania group are exploring the possibility of using VisIVO as a way of offering 3D visualizations of data accessed through the Aladin workbench, and, conversely, the possibility of using Aladin to offer images of objects selected from a VisIVO visualization. There are some technical challenges to doing this, and these will be addressed when UgoBecciani visits Strasbourg soon.
- Server-side availability from the desktop. TonyLinde pointed out that DS3 plan to develop an framework whereby server-side tools can be made available to desktop applications. It would be good if a single standard for desktop applications could be defined.
[8] Portsmouth Stage01 Plans (
GarrySmith)
- Broker. The essential idea of the broker is that it provides access to computational resources needed for embarrassingly parallel jobs while hiding from the user the details of the underlying job handler (e.g. Globus, Condor, etc).
DS6 Meeting 01 Actions
| Number |
Person |
Action |
Due Date |
| ds6.m01.01 |
BobMann |
Ask NicholasWalton about status of AVO SRM |
March 3 |
| ds6.m01.02 |
AlasdairAllan |
Post wiki note about use of agents in astronomy |
March 31 |
| ds6.m01.03 |
GiuseppeLongo and UgoBecciani |
Prepare wishlist of things to aid integration of VisIVO and Astroneural |
March 31 |
| ds6.m01.04 |
BobMann |
Send to all my version of their plans |
March 3 |
| ds6.m01.05 |
BobMann |
Enquire about placing of VOTECH logo on software, etc |
March 3 |
| ds6.m01.06 |
BobMann |
Draft definitions of "data mining", "visualization" and "data exploration" for DS6 Scope statement |
March 3 |
| ds6.m01.07 |
BobMann |
Follow up suggestion for cross-DS tool assessment/validation wiki page with other DS leads |
March 3 |
DS6 Stage 01 Assignments
Topic revision: r29 - 2005-09-15 - 14:30:52 -
UgoBecciani
Changes
Index
Search
Main
Copyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.