pamel;afox DOUBLETREE MARLON OKC *********************** Itinerary 3 Price per passenger: $345.50 (USD) Select Itinerary 800 MSU Carrier Flight # Cabin & Class Departs Arrives Stops View Seats Upgrade Pts From Time/Date To Time/Date Not Applicable N/A Delta 6012* Coach (K) IND 1:25pm 09 Jun 2002 CVG 3:07pm 09 Jun 2002 0 Available 2 Delta 981 Coach (K) CVG 4:15pm 09 Jun 2002 AUS 5:42pm 09 Jun 2002 0 Available 2 Delta 1524 Coach (K) AUS 2:45pm 13 Jun 2002 CVG 6:17pm 13 Jun 2002 0 Not Applicable N/A Delta 5346* Coach (K) CVG 7:15pm 13 Jun 2002 IND 7:01pm 13 Jun 2002 0 Note: The following fare rules apply to this itinerary - Rule 1 Itinerary 1 Price per passenger: $345.50 (USD) Select Itinerary 800 MSU Carrier Flight # Cabin & Class Departs Arrives Stops View Seats Upgrade Pts From Time/Date To Time/Date Not Applicable N/A Delta 6012* Coach (K) IND 1:25pm 09 Jun 2002 CVG 3:07pm 09 Jun 2002 0 Available 2 Delta 981 Coach (K) CVG 4:15pm 09 Jun 2002 AUS 5:42pm 09 Jun 2002 0 Not Applicable N/A Delta 4771* Coach (K) AUS 5:30pm 13 Jun 2002 DFW 6:32pm 13 Jun 2002 0 Not Applicable N/A Delta 5155* Coach (K) DFW 7:40pm 13 Jun 2002 IND 9:52pm 13 Jun 2002 Earthlink Netscape6 need Internet Access Find orange cable Find dead sprint cable call england gatwick hotel NCSA EOT Euro 35 25 16 // 16 20 // 21 20 // 20 46 taxi This review cycle did not include a detailed review of the work of D2 and related computational issues, as this was done in 2001. We did however see many examples of the division's work on modeling and simulation. On the other hand, much of the work done elsewhere in the division does not require computational sciences that are path-breaking and world class. For that set of projects, the computing work in D division is largely workmanlike- certainly "good enough" to get the job done and not exhibiting serious intellectual deficiencies but also not great computational science. This level of performance is entirely defensible, in terms of programmatic requirements. But the division and laboratory management more generally should consider more carefully whether it is also desirable-whether there are not further innovations to be won through a broad-based improvement in the quality of computational science in the division. For quite a long time Los Alamos was in the leading in scientific computing. In recent years, Argonne, Livermore, and Oak Ridge have greatly improved their scientific computing while Los Alamos has struggled to stay even. Thus the fact that computational science within D division is somewhat mixed reflects a broader malaise at the laboratory, one requiring leadership from the top. I feel very uncomfortable about section on Modeling and Simulation Science I don't think this field has an agreed meaning and I am not aware for instance of any recognition in computer science that such an area is well defined Outside D division one would talk about Computational Science and more recently Science and Engineering Grids e-Science and e-Engineering BUT D Division is not state of art in modern large scale computing technologies in these areas I believe Chris Barrett's division has a name that reflects some peculiar local politics -- as you know I think his work is excellent but it is focussed on one particular approach to simulation. So I think this section is handicapped by the weaknesses in D division in computing (both vision and practice) and so it is hard to organize properly I would separate off discussions starting An excellent example of the highly sophisticated and complex modeling work .. (page 16) to Moreover, the problem focus of LANL-they model because they have a problem, rather than modeling for its own sake-creates a built in disciplinary mechanism to continually develop the science and art of simulation. (top page 18) and call it Modeling Complex Systems or something like that This work is in my opinion generally weak from a computer science port of view but strong from an operations analysis/statistics point of view By separating I think discrepancies in evaluation in current long sedction become less pronounced I would remove or modify a) Such work, motivated by a real problem, transcends its application and has general value in other avenues such as signal processing, weather modeling, etc. Many engineers and physicists tend to work on problems involving multivariate Gaussian models. Unfortunately, such work tends to be ad hoc (i.e. not mathematically formal) and often fails to exploit the modern computational tools that are now available and known to those who specialize in stochastic process theory. The work done in the division proceeds on a formal footing and in doing so is scientifically meritorious. ** D division may be good in some ways but weather modeling field and many other fields are way ahead computationally b) Elsewhere in this assessment we have flagged for praise the work in modeling large-scale program management. ** I couldn't find flagged praise! The work on program mangement was humdrum from an information technology point of view in my opinion c) A distinctive feature of D division's approach is that the research staff has absorbed and developed an enormous amount of this "art." and further notessentences. ** Clearly D division has learnt "art" in its application area of expertise BUT all "O/E" application areas around the world must understand their application deeply -- I believe for example that the high enery physicists and astronomers for example in leading universities understand best practice in data analysis . I find controversial assertion that It is hard to imagine how this could come about at a university, Finally we come on page 18 to a section one could call Computing and Computational Science ------------------------------------ Starting Especially in D2, the theoretical work is path breaking. .... ---------------------------------------------------------------------- We continue to work with Tribal College and other minority communities on development of a collaboratory. We have identified an early prototype to demonstrate value of use of Web services in learning objects. Substantial Progress was made in developing collaboration technologies. Highlights of this work are: a) Integration of our Collaboration Environment Garnet with Peer to Peer technologies. b) Initial demonstrations of Audio/Video conferencing as a Web Service with a common XML specification used for integrate H323 (e.g. Polycom) and SIP (e.g. HearMe) and MBONE (e.g. Access Grid) c) Design of a new approach to collaboration based on shared Web Services using portlet architecture d) Collaborative SVG viewer between PDA and PC's We gave several talks that inluded these topics including Global Grid Forum, Alliance Portal Expedition meeting and a visiting group from Taiwan planning a new national initiative in distance learning. We discussed collaborative portlets with Sybase and are discussing a new initiative with San Francisco School District who wish to plan the infrastructure needed to support emerging learning environments. Papers on these topics will be presented at PDPTA and a SVGopen conferences this summer while Fox gave an ICCS 2002 keynote talk (April24 2002) on applications of Web Services to computing and collaboration. Geoffrey Fox, Ozgur Balsoy, Shrideep Pallickara, Ahmet Uyar, Dennis Gannon, and Aleksander Slominski, "Community Grids" invited talk at The 2002 International Conference on Computational Science, April 21 -- 24, 2002 Amsterdam, The Netherlands Typical talks are: http://grids.ucs.indiana.edu/ptliupages/ggf4feb02/ggf4tutorialfeb02.ppt http://grids.ucs.indiana.edu/ptliupages/distedtalkfeb02.ppt http://grids.ucs.indiana.edu/ptliupages/ncsaexpeditionmar02.ppt http://grids.ucs.indiana.edu/ptliupages/collabwsncsaapril02.ppt ---------------------------------------------------------------------- Data Deluge Data private 1 year Comb - e - Chem MyGrid BioInformatics Meta-data & Search aircraft engines generate 1 gigabyte of sensor data per flight IBM Oracle Tamanoir talk Active Grid INRIA interesting DyRAM Active Multicast Stuer Multicast Java Objects Vazquez Interesting Compression Sunderam PVM as a Web Service