next up previous
Next: 6.1 Other Systems Up: SETL for Internet Data Processing Previous: 5.4 Summary

6. Conclusions

The strengths of SETL as a data processing language follow largely from its properties which (1) make it a good language for high-level algorithm description and prototyping, (2) encourage adherence to the programming guidelines given in Chapter 5 [On Data Processing], particularly the one which advocates employing plenty of processes, (3) reflect typical data processing environments through a convenient predefined interface to the most commonly used parts of Unix 98, and (4) allow SETL to serve as a highly competent ``glue'' language [156] for interconnecting programs written in other languages.

The case study of Chapter 4 [WEBeye: A Case Study] has shown how the general-purpose SETL I/O facilities, operating system interface, and miscellaneous extensions described in Chapter 2 [Environmentally Friendly I/O] work with the sockets-based network support described in Chapter 3 [Internet Sockets] and the modularity of small, high-level programs to produce a system that is comprehensible, robust, and maintainable.

In this concluding chapter, I will review a few other such systems, and then discuss some needs which have become apparent, particularly in the area of types.



 
next up previous
Next: 6.1 Other Systems Up: SETL for Internet Data Processing Previous: 5.4 Summary
David Bacon
1999-12-10