(Feb 1, 2004: several links in the following text have gone stale. We are leaving the original URI in the document for your reference, but have not fixed the link.

Periodic Status

What was accomplished and learned during various time periods of this LDRD.

April 1999

We met with ANAG and began to discuss what we could do for them and what they could do for us. Basic functionality for a set of visualization tools was laid out and given priorities. We got a copy of the the Chombo distribution, ANAG software, and some sample 2D and 3D data. We were able to compile Chombo on an SGI workstation using the native compilers and compile programs we wrote that read the 3D data. These programs produced geometry and a flat data field that AVS could read in. We did some basic visualization in AVS to demonstrate this.

May 1999

We got a copy of the current distribution of Vtk, version 2.3. We compiled, tested, and installed Tcl, Tk, and Vtk on our SGI workstations using the native compiler. We inquired about additions to Vtk which would let it work with the CAVE library.

May 31, 1999 - June 4, 1999

We began weekly meetings with ANAG on Friday mornings alternating between their offices and the Vis. Lab. We discussed their data formats in some detail (both the embedded boundary and AMR data). We continued to discuss visualization needs and possibilities.

June 7, 1999 - June 11, 1999

We were able to compile and run programs which used both the Chombo and Vtk libraries. These programs mirror the work that was done earlier with AVS and produced Vtk compatible output of the geometry and AMR data flattened onto a single grid. This output could then be input to an Vtk visualization pipeline and all the basic visualization we had done in AVS could be done in Vtk. Thus, we could now use a tool which was freely available, come with source code, and could be extended.

June 14, 1999 - June 18, 1999

We began thinking about how Vtk could be extended to visualize the ANAG data sets. Some rough ideas were put forward at our weekly meeting and some more concrete visualization requirements were discussed. We decide to start a design document that would document our approach to using Vtk for visualizing the ANAG data. We also discussed ideas for alternative interfaces to ANAG data, especially for visualization. This let to another document which detailed this.

June 21, 1999 - June 25, 1999

The documents mentioned above were begun along with the design work necessary to write them. We discussed the Vtk design document at the weekly meeting and there were no major objections raised. We tried one of the extensions of Vtk to the CAVE library done by Matthew Hall at NSCA. This extension had fairly limited functionality (which the author indicated) but what was there seemed to work well.

June 28, 1999 - July 9, 1999

We exchanged some information about with ANAG about using HDF 5 in conjunction with FlexIO. We built Chombo and Vtk for our Solaris Sun workstations using g++ since the native compiler won't work for Chombo. We began to build Vtk for the ANAG Solaris PC's and ran into various problems with compilers, shared libraries, other packages (e.g. Mesa), and disk space. We began these WWW pages documenting our progress.

July 12, 1999 - July 16, 1999

We finished successfully building Vtk on the ANAG Solaris PC's with shared libraries. We were able to build and run the applications mentioned earlier which used Chombo and Vtk. We gave a tutorial on Vtk to ANAG. We tried another extension to Vtk which uses the CAVE library from Michael Papka at ANL. It is a more complete extension than the one we tried earlier. It also work fairly well. We're still evaluating these extensions.

July 19, 1999 - July 23, 1999

We began some work to do some visualization of ANAG data in AVS which is analogous to what we are doing in Vtk. We began gathering information about extending Vtk the way we need to for the ANAG data. This included using the Vtk users group e-mail list through which we found out about some similar work being done at NCSA by Matthew Hall. We are looking into this.

July 26, 1999 - July 30, 1999

We worked on getting AMR data into AVS via a preliminary version of a network based interface. We continued communicating with the group at NCSA doing visualization of AMR data. We found out that what they have done maps very closely to what we'd like to do and we will be exploring what can be done to bring their work here, use and extend it, and give this back to them.

August 2, 1999 - August 6, 1999

We continued to explore visualizing AMR data in AVS. This was documented and a Visualization Communication Protocol (VCP) was defined. Some results of a shared, distributed volume renderer were generated from AMR data from CCSE. We made arrangements to meet with meet with John Shalf from NCSA after SIGGRAPH to discuss their work with AMR, AVS, and Vtk. We began to learn how to extend Vtk to meet our needs (independent of the NCSA effort).

August 9, 1999 - August 13, 1999

One of us went to SIGGRAPH for the week and the other continued work on extending Vtk handle AMR-like data. This included identifying the issues in preparation for meeting with John Shalf.

August 16, 1999 - August 20, 1999

We met with John Shalf and ANAG. This meeting was very productive and it was decided that ANAG would benefit the most by using the http://infinite-entropy.ncsa.uiuc.edu/AMRpics/ - work John has been involved in at NCSA ( http://infinite-entropy.ncsa.uiuc.edu/ - this link is also useful). Thus, instead of providing ANAG with a bare bones extension to Vtk, we will be getting the extensions from John and helping to install them for ANAG. ANAG will be setting up I/O necessary, http://infinite-entropy.ncsa.uiuc.edu/FlexIO/ - FlexIO, to use these Vtk extensions.

August 23, 1999 - August 27, 1999

We communicated more with John Shalf and got all the pieces necessary to try out the work they have done. The pieces included http://hdf.ncsa.uiuc.edu/obtain.html - HDF 4.1r3, http://bach.ncsa.uiuc.edu/FlexIO/ - FlexIO, and ftp://bach.ncsa.uiuc.edu/outgoing/vtkMultiGrid.tar.gz - Vtk multigrid extensions (as a compressed "tar" file). We also got some previous work John's group had done, ftp://bach.ncsa.uiuc.edu/outgoing/AMRSlicer.tar.gz - AMRSlicer, and an http://bach.ncsa.uiuc.edu/FlexIO/VizReaders.html - AVS module which reads their AMR format via FlexIO. We began installing these this week and will continue next week.

August 30, 1999 - September 3, 1999

We successfully installed all the packages and tools listed above on our local machines and on the ANAG Solaris PC's. Many minor changes were necessary but overall everything was fine. We began looking at the tools and exploring their capabilities.

September 6, 1999 - September 10, 1999

Having explored the capabilities of the NCSA tools (extensions to both AVS and Vtk) we began documenting what we had found. So far all the data used in this process was provided with the tools. We are in the process of working with ANAG to get some of their AMR data in a FlexIO/HDF format. We communicated with John Shalf and gave him a list of problems we had installing the software, any fixes we had, and suggestions for changing to the software itself (especially the Vtk extensions).

September 13, 1999 - September 17, 1999

We continued to learn about the Vtk extensions which are being actively developed at NCSA by using them. We are waiting for ANAG data in a format that the Vtk extensions can read so we can do visualization with their data. We are talking to John Shalf about how this code will be released/made available so that ANAG can point to it in their release and include visualization code which uses these extensions. We outlined what we would deliver at the end of the LDRD. Specifically, we will provide ANAG with the NCSA Vtk extensions compile and installed on their machines, an example of using the extensions to visualize their data, and any other documentation necessary for them to begin using the extensions.

September 20, 1999 - September 24, 1999

We got ANAG data in a form the Vtk extensions can read but seems to be some problem in the format that is causing problems when read with these extensions. The data appears to be correct when viewed with a lower level, AVS interface but incorrect when read into the Vtk extensions (i.e. there seem to be zero "active" grids on each level of refinement even when all are selected/shown). We are tracking down this problem so we can visualize this data with the extensions.

September 27, 1999 - October 1, 1999

We managed to get ANAG data read into the NCSA Vtk extensions and did some visualization of that data. We handed all this over to ANAG and wrote began writing the final report.