AVO Telecon: Mar. 29, 2004 participants: MA Mark Allen TB Thomas Boch LC Laurent Cambresy SD Sebastien Derriere FG Francoise Genova ML Marco Leoni AM Alberto Micol AR Anita Richards chair: PP Paolo Padovani minutes: MD Markus Dolensky, revised by PP excused: PO Patrizio Ortiz NW Nicholas Walton next telecon: Apr. 5, 2 p.m. GMT (WA1 - science only) Apr. 19, 2 p.m. GMT (the usual crowd) New Action Items: PP narrow down date for SWG meeting at CDS to either 8+9 July 04 or earlier; there is still the option to involve an ESA site depending on political circumstances PP prepare plan to present to SWG which is based on his elaboration of the SWG feedback and FG's WA2 plan (appended) MD prepare another iteration of topic Avo.PublishDataToVO Previous Action Items: MA inquire whether CDS could host next SWG meeting in July Answer: Yes. New action on PP. Closed. PH continue evaluation of improved ACE FG raised awareness of potential project at meeting in Paris Ongoing. MA provide a data tree example of UDF; also write some paragraphs of text for topic Avo.PublishDataToVO Ongoing. Question to Francois Bonnarel: Can users load data cubes into the current version of the AVO prototype? If so, what is the recipe to do so? Answer is no. Closed. FG summarize infrastructure goals (complementary to SWG recommendation) and send to Paolo CDS's comments/timelines on recommendations Summary appended. Closed. PP get more detailed feedback from P. Rosati about sample cases where improvements are desired (e.g. contour tool, smoothing, RGB dialog) Ongoing. LC 'Error propagation' contacts VO-India VO-India agrees to support error expressions; need to work out schedule and details. Ongoing. **** > Dear colleagues, > > I hav the action to produce a document explaining WA2 view > of the AVO demos. Here is a draft. Comments welcome > > Regards, > Francoise > > _________ > > WA2 view on the January 2004 AVO demo > _____________________________________ > > Draft 0.2, 28 March 2004 > > VO demos have three aspects: one is the development of prototypes > allowing science usage, to understand better the science specification of > the IVO, and to begin the training of scientists in the usage of these > new tools, the second one is the usage of underlying infrastructure, > in particular the interoperability standards, to improve their > specification and test their implementation, and finally demos also aim > at advertizing the new tools towards the science community, Committee > members, and funding Agencies. This document deals with these various > aspects, at diffeent levels. > > 1. Documentation > > - prototype and also Aladin and VizieR functionalities are more and more > complex. In addition to the Demo viewgraphs, tutorials on specific > science usages are provided. New subjects are welcome. > http://aladin.u-strasbg.fr/tutorials/ > > - A new 'trace' mode is being developed. It keeps track of an Aladin > session and allows one to replay it automatically. This will allow easy > maintenance of tutorials when the software is updated, > by simply replaying the script. > > 2. Service development > > Updates and additions proposed by the SWG fall into several categories. > We have considered separately in the following pixel > manipulation/visualization, on one hand, and additional services, > on the other hand. > > Obviously all VO-required functionalities will not be included in one > single tool, but astronomers will use a set of interoperable tools and > services. > > 2.1. Pixel manipulation/visualization: > > The following points, brought up by the SWG, will be developed in the > coming months: > > - access to 'real' pixels in Aladin, with additional functionalities > such as intensity profiles, and improvement of image-oriented tools > (contours, RGB, ...). Different transfer functions will be made > available (log, sqrt, pow2, ...). This development also means that > Aladin will be interoperable at the image level, e.g. with image > analysis tools > Foreseen schedule: May 2004 > > - Multi-view data visualization, with interaction between the views: > This requires deep changes in the Aladin software and is presently > being assessed. > A 'light' development allowing the display of several windows by sending > a script to Aladin: feasibility and schedule remain to be assessed. > > 2.2. Connection with services: > > - improved cross-match tool: this will likely be an incremental > development, with first release by Summer 2004. The tool will be > distributed as a plug-in in VizieR and Aladin. In addition, a "light" > cross-match version, which has proven to be very effective to obtain > science results in the 2004 demo by combination of images, catalogues > and cross-match, will be integrated into Aladin; this will improve > performance for cross-matching of local catalogues (foreseen schedule: > May 2004). > > - plug-ins of new remote process: the possibility to plug new remote > services to the prototype easily, by describing input/output in > the GLU, will be rapidly developed (May 2004). Are there candidates for > testing this implementation with services developed by the other AVO > teams? > > - To be explored before the end of 2004: > . GLU interaction with IVOA registries > . prototype interaction with the AstroGrid workflow (inetraction > between teams has to begin ASAP if this is to be included in 2005 demo) > > - VOPlot: updates of VOPlot are discussed with the Indian VO. > e.g. error propagation > > > Interoperability standards > > 2003 and 2004 demos made an extensive usage of VOTable and UCDs, and > in 2005 emphasis will likely be on other interoperability standards, in > particular registries and data models. However proper completion > of the new versions of VOTable and UCDs is critical for the VO, and the > amount of work must not be underestimated. In addition, lessons learned > from 2003 and 2004 demos in terms of data model, data access layer, > registry, web services, have to be taken into account. > > - VOTable is a critical tool for VO implementation, by allowing the > exchange of metadata and data between services developed > by different groups. Version 1.1, allowing in particular the > grouping of data, has to be completed by the May 2004 > Interoperability meeting. > > - UCDs are used in data retrieval, selection and cross-match, and a > rapid convergence with VOX (used in SIAP) is required. WA2 has put a > huge amount of work in the last year in the systematic checking and > pruning of UCDs, and in preparation of UCD2, then UCD1+ (collab. R. > Williams and the UCD Steering Committee). UCD1+ has to be > completed by the May 2004 Interoperability meeting. > > - data model and data access layer: AVO 2003 and 2004 demos have shown > a powerful example of usage of hierarchical data tree, with a set of > metadata available at each tree level. 2004 demos has shown the > extension of this type of data description and data access to data cubes. > . lessons learnt from the successful development of > the IDHA data model (of which the hierarchical data tree > is a view), have to be taken into account in > the 'Observation' data model, which is presently discussed in the Data > Model WG. The Data Model WG is about to publish the first draft of > Observation by May 2004. It is for the moment fully focussed on the main > concepts and entities to be modelled for the exchange of observational > data. A textual and UML description of the Observation model will be > provided for the IVOA May 2004 meeting. The data model serialisation is > still a question to be discussed. > . the possibility to include structured access to the data > has to be included rapidly in the DAL (SIAP evolution). The fact that > the inclusion of structured access by the DAL WG might be > delayed after the end of 2004 is a serious concern. Proposals will soon > be distributed for discussion at the IVOA May meeting. > > - web services: one conclusion of the demo is that although it is > relatively easy to wrap existing programs into web services, the user > interface and documentation aspects have to be seriously taken into > account. > > - registries: at present the GLU registry is a fundamental building > block of the demo. Interaction between the GLU and IVOA registries has > to be explored, as explained above.