This report describes the progress made over the past six months and the project’s plans until January 1, 2001.
What is the difference between the Scientist Expert Assistant (SEA) project at Goddard and the Astronomer’s Proposal Tool (APT) at STScI? SEA was a research project developed at Goddard with STScI science input to prototype new Proposal Preparation Environments. The result of this two-year effort is the SEA prototype. APT is STScI’s project to build upon, adapt, and re-engineer SEA for HST operational use. Think of SEA as a proof of concept that resulted in a useful research prototype which can be built upon. APT is taking the SEA prototype to its next step, HST operational use.
A major goal of APT is to incorporate RPS2’s functionality into APT such that RPS2 can be retired. We expect that the Phase 2 Cycle 11 (i.e., December 31, 2001) release of APT will provide users with an alternative to RPS2 for proposal preparation. We expect that this release will not support all the existing RPS2 functionality. Some users will still have to use RPS2 and others will be able to use APT. We expect that the Phase 2, Cycle 12 (i.e., December 31, 2002) release of APT will allow us to retire RPS2.
APT is envisioned to be more than just a Phase 2 proposal preparation tool. We are currently working on deploying Exposure Time Calculators for ACS, STIS, & NICMOS. The Phase 2 Cycle 10 (i.e., December 31, 2000) release of APT will incorporate an ACS Exposure Time Calculator. The Phase 1 Cycle 11 (i.e., June 1, 2001) will incorporate the STIS and NICMOS Exposure Time Calculators. We are also investigating the technical feasibility of integrating the Visual Target Tuner and STScI’s archival research tool (Starview2).
APT’s philosophy is to provide the user community with useful prototypes (tools) as soon as possible. To this end, we are releasing the Visual Target Tuner on June 1, 2000 and December 31, 2000 for Phase 1 and Phase 2 use. We will continue to release parts of APT as early as possible, gradually working to our goal of replacing RPS2.
Why was the January 1, 2000 release of the Visual Target Tuner (VTT) and ACS Exposure Time Calculator (ETC) delayed? STScI did not have enough resources available to test the VTT’s and ETC’s accuracy or to provide quality user support after it was released. The following were contributing factors that did not allow STScI to free up the appropriate resources needed to release the VTT and ETC;
- STScI Re-Organization
- HST servicing mission
- RPS2 release had higher priority in testing/documentation
The SEA prototype was the culmination of a Goddard (software development) and STScI (science input) collaboration. With STScI now beginning to adapt and extend SEA prototype for HST use, we needed to work out a new collaboration agreement. Here are the highlights of the agreement.
- STScI and Goddard will share all software changes
- STScI and Goddard are working as one software team to deliver the VTT on June 1, 2000
- Goddard will provide technical advice on modifying/enhancing the SEA software
- STScI and Goddard will hold monthly development meetings
- Goddard will begin extending the SEA software to simulate what an image might look like to a user before it ever flies on the spacecraft
- STScI will concentrate on extending the SEA software to replace RPS2 and add additional user tools.
STScI and Goddard are already working under this agreement. Goddard is currently funded until October 1, 2000.
An STScI user group has been formed to provide the APT project with the following input;
- Provide project vision and priorities
- Define high level user functionalities
- APT software evaluation, look and feel, functionality
- Testing and Documentation Support
The science members of the user group are;
- Steve Lubow (leader)
- Chris O’Dea
- Ray Lucas
- Ron Downes
- Megan Donahue
- Max Mutchler
We have been meeting and coordinating our efforts with the Goddard (SEA) team’s scientist, Anuradha Koratkar.
The VTT is scheduled for release on June 1, 2000 as an operational prototype for Phase 1 cycle 11 use. The June release will help the Phase 1 submitters answer the following types of questions;
- How many exposures will I need to mosaic my extended target?
- Which of my potential targets "fits best" in the aperture.
- Would a coordinated parallel in another aperture provide
additional data related to the science in my proposal?
- Do any of my potential targets have nearby bright objects that
Could spoil the observation? Is there an orient which would avoid the
bright object?
The next external release of the VTT is planned for Phase 2 cycle 11 (i.e., January 2001 timeframe). The project will work with the APT user community to help determine and prioritize the features that should be put into this release.
The ACS ETC is scheduled for release on January 1, 2001. Work has already begun to re-engineer parts of the ETC software to make it easier to maintain and extend, thus reducing the cost to incorporate new instruments.
One of the biggest complaints about RPS2 over the past 6 years is that it runs too slow. RPS2 consists of numerous software systems, each being responsible for a specific function, such as syntax checking, laying out exposures in their orbits, and visit scheduability. Simply integrating these systems as they are into APT, will transfer the RPS2 speed problem into APT. From our RPS2 experience, we know that the slowest RPS2 component is the software component that lays out exposures in their orbits.
To address this problem, the “Quick-Trans” project was started, with the following charter;
- Study alternative approaches to solve this problem
- Provide a recommendation on how to proceed.
Recommendations on how to proceed are due on May 9, 2000
One of STScI’s visions for APT is that APT is not just a Phase 2 submission tool (RPS2), but also a tool that integrates many of the STScI tools that our user community uses. We have begun a project to assess the technical feasibility and level of effort involved with integrating STScI’s archival research tool (Starview 2) and the Visual Target Tuner Tool. This project is a collaborative effort between the Goddard SEA team and the STScI APT team. The technical evaluation and a decision on how to proceed will be made by mid April 1999.
The Goddard SEA team just finished an evaluation of the SEA prototype by asking ~20 HST PIs to use the SEA prototype. The evaluation showed that the SEA Graphical Interface (GUI) was not intuitive to a large number of the evaluators. After the June release of the VTT we will begin working with the APT user group on improving the intuitiveness of the GUI.
The software development team has evaluated and selected an object-oriented design tool and a Java development tool to support the project’s Java development. We plan to support APT on both SUN and PC computers, and have ordered PC hardware for our software development team.