Towards Increased Productivity of Algorithm Implementation in
Scaled Up Applications
Robert Paige
ONR Grant N00014-97-1-0128
| Project Objectives:
| Increasing Productivity of Large Scale,
| Labor Intensive, High Performance
| Systems
Scientific Approach: | Recent Results:
1. Design statically typed | 1. An initial design of Low SETL, a
variants of SETL as specification | low level, statically typed variant of
languages. The type system allows | SETL2 was used to speedup the linear
errors to be detected at compile | expected time fragment of Willard's RCS
time, helps to model complex data | query language to linear worst case time.
structures, and supports | 2. Low SETL was extended with disjoint
computational transparency. | disjoint unions, user types, recursive
2. Transform these specification | types and subtypes. The formal
languages into C using partial | semantics of a batch read method
evaluation, data structure | for the new Low SETL was developed.
selection, finite differencing, and | An algorithm implemented in the old
chaotic iteration iteration. | Low SETL was developed to convert
| input in string form into efficient
| data structures for a variable list with
| any signature from our type system in
| linear time in the length of the string.
| 3. An initial solution to the hidden
| copy problem for SETL2 was discovered,
| and was used to implement a new SETL2
| backend. This result solves a difficult,
| important problem that has plagued SETL
| compilers for 20 years.