Early DSRFS-related technical reports

The early technical-report-type publications (at least historically) related to distributed stepwise refinement of finite sequences (DSFRS) are listed in chronological order here. Accordingly, the more application-oriented work precedes the more theory-oriented.

Application-oriented work

An experimental prototype of a proposed tool concept for machine-level programming was described in the following two reports:

• Eero Lassila. ReFlEx - an experimental tool for special-purpose processor code generation. Report B15, Helsinki University of Technology, Digital Systems Laboratory. March 1996. ISBN 951-22-2993-5.
• Eero Lassila. Towards optimizing code generation by domain-sensitive macro expansion. Report A42, Helsinki University of Technology, Digital Systems Laboratory. January 1997. ISBN 951-22-3439-4.

In the tetrasystem terminology, the prototype uses frame ⟨σI, σI, σE, σE⟩. It is implicitly assumed that the refinement rule base is sufficiently constrained so that tetrasystem soundness is achieved. However, the semantics-preservation problem is not explicitly mentioned at all.

The prototype was also presented in a peer-reviewed conference: see the article in Proceedings of the 22nd Euromicro Conference (1996) in my publication list.

Theory-oriented work

The following two early theoretically oriented contributions are now somewhat outdated.

Belt-selectors

Belt-selectors were originally formulated in

• Eero Lassila. On tree belts and belt-selectors. In Nisse Husberg, Tomi Janhunen, and Ilkka Niemelä (eds.), Leksa Notes in Computer Science, pp. 47-58. Report HUT-TCS-A63, Helsinki University of Technology, Laboratory for Theoretical Computer Science. October 2000. ISBN 951-22-5211-2.

However, this old definition differs from the present one by preventing proper descendant nodes of the argument node from being included in the belt selected. (That is, the selected belt always contained the argument node itself.)

Formal devices for DSRFS

Two early DSRFS-oriented formal devices were then presented in

• Eero Lassila. A tree expansion formalism for generative string rewriting. Report HUT-TCS-B20, Helsinki University of Technology, Laboratory for Theoretical Computer Science. June 2001. ISBN 951-22-5554-5.

In particular, the concept of a trisystem (whose definition is moreover cumbersome) has now proven to be obsolete. Furthermore, some terminology in the report is outdated: for instance, "normalness" has replaced the old "idealness". (Some errors have also been found in the report. Part 2 of the conjectured theorem H.31 is false. In addition, the first sentence of the second paragraph of section A.1.2 is best dropped out.)