Jeff Thomas

Scalable software libraries

Many software libraries (e.g., the Booch C++ Components, libg++, NIHCL, COOL) provide components (classes) that implement data structures. Each component is written by hand and represents a unique combination of features (e.g. concurrency, data structure, memory allocation algorithms) that distinguishes it from other components.We argue that this way of building data structure component libraries is inherently unscalable. Libraries should not enumerate complex components with numerous features; rather, libraries should take a minimalist approach: they should provide only primitive building blocks and be accompanied by generators that can combine these blocks to yield complex custom data structures.In this paper, we describe a prototype data structure generator and the building blocks that populate its library. We also present preliminary experimental results which suggest that this approach does not compromise programmer productivity nor the run-time performance of generated data structures.

The GenVoca model of software-system generators

An emerging breed of generators synthesize complex software systems from libraries of reusable components. These generators, called GenVoca generators, produce high-performance software and offer substantial increases in productivity. GenVoca is a blueprint for creating domain-specific software component technologies. Predator and Adage are based directly on GenVoca. Experimental evidence from the Predator project supports claims of performance and productivity with this model.<<ETX>>

P2: A Lightweight DBMS Generator

A lightweight database system (LWDB) is ahigh-performance, application-specific DBMS. It differs from ageneral-purpose (heavyweight) DBMS in that it omits oneor more features and specializes the implementation ofits features to maximize performance. Although heavyweight monolithic andextensible DBMSs might be able to emulate LWDB capabilities, they cannotmatch LWDB performance.In this paper, we describe P2, a generator of lightweight DBMSs, andexplain how it was used to reengineer a hand-coded, highly-tuned LWDB usedin a production system compiler (LEAPS). We present results that showP2-generated LWDBs reduced the development time and code size of LEAPS by afactor of three and that the generated LWDBs executed substantially fasterthan versions built by hand or that use an extensible heavyweight DBMS.

Reengineering a complex application using a scalable data structure compiler

P2 is a scalable compiler for collection data structures. High-level abstractions insulate P2 users from data structure implementation details. By specifying a target data structure as a composition of components from a reuse library, the P2 compiler replaces abstract operations with their concrete implementations.LEAPS is a production system compiler that produces the fastest sequential executables of OPS5 rule sets. LEAPS is a hand-written, highly-tuned, performance-driven application that relies on complex data structures. Reengineering LEAPS using P2 was an acid test to evaluate P2s scalability, productivity benefits, and generated code performance.In this paper, we present some of our experimental results and experience in this reengineering exercise. We show that P2 scaled to this complex application, substantially increased productivity, and provided unexpected performance gains.