Subject: Access Grid, Training, TangoInteractive etc. Resent-Date: Thu, 29 Apr 1999 13:10:45 -0400 Resent-From: Geoffrey Fox Resent-To: p_gcf@boss.npac.syr.edu Date: Mon, 26 Apr 1999 08:25:52 -0400 From: Geoffrey Fox To: stevens@mcs.anl.gov, reed@oboe.cs.uiuc.edu CC: roscoe@bu.edu, Scott Lathrop So I was in the "portal" meeting in Feb 1-2 gettogether and so did not participate in Access Grid discussions then. I also saw emphasis was to the high end in visualization which is not my expertise. However if access grid is to be used for education and training, then we can probably learn from our experiences with TangoInteractive which has now been used for several reasonably succesful such activities over the last 2 years. My work is based on the assumption that web pages are the most important material (for training) to be shared and here we can use TangoInteractive's ECMAScript (JavaScript) interface. This exploits Tango's integration with browser, In this area, my main Alliance thrust has been a) Collaborative versions of form based applications like the Biology Workbench b) Shared dynamic HTML for telepointers etc. c) and most interesting (to me) the issues of universal access to web material using JavaScript to modify style sheets to integrate an audience across disabilities. This is with EOT partner the Trace center and we are working on supporting sharing of new W3C Document Object Model In developing Tango, most pressure I get is not at high end but to do things like support MacIntoshs, Internet Explorer etc. This is a a nontrivial engineering problem and consumes quite a bit of effort as does avoiding the myriad bugs in today's web browsers! I see this as quite helpful and somewhat orthogonal with the main AG thrust. So I wonder if these issues should appear somewhere in AG roadmap and if I can supply some material along these lines? BTW on a different tack of collaborative computing, I think we need to indicate that some issues are research. Namely it is not obvious (to me) what support users want. Even in a simple thing like Shared Biology Workbench, the mode of sharing needs to be evolved by several experiments. If you ask about shared debuggers etc., the collaborative model seems to me even less clear This uncertainty does of course make area pretty interesting but I think any project plan should recognize that these areas such as portals and collaboration have substantial esearch and so cannot be laid out programmatically without some "if statements" Geoffrey Fox gcf@npac.syr.edu, http://www.npac.syr.edu Director of NPAC and Professor of Physics and Computer Science Phone 3154432163 (Npac central 3154431723) Fax 3154434741