Talk from the bench, Guy Laffitte, Henri Habrias

  At the Vertou Stadium 

 

1.1 First Morning 

The wind was favorable. The BN (Biscuiterie Nantaise) factory offered us delicious smells of biscuits. And Guy answered our questions while my grandson was at French soccer practice.  

H.H.: You met J.R. Abrial when you did a training course at CEPIA on the Domaine de Voluceau in Rocquencourt. What year was that? 

 

 G.L.: In October 1985.

 

H.H.: I was also there in October and in the spring. We could have met there. But we didn't know each other yet. What did you like about it then? 

 

 G.L.: Strong point: instead of vainly trying to prove a problem on any program, we start by creating a specification that can be proven, then we focus only on programs implementing the specification.  

 

H.H.: J.R. Abrial thanks you in the B-Book for your contribution. Tell me about it.  

 

G.L.: I used to work at INSEE on Boulevard Adolphe-Pinard in the 14th arrondissement. 

 

H.H.: We have the population file in Nantes, and the NNI served as an exercise for learning B! When I was in the IT department at the town hall, I remember learning that at birth, when sex is undetermined (we wait for the testicles to fall off. Have they taken into account the property of inevitability?), the sex code is set equal to 1. In my time, the only change was the change to 1.)

 

G.L.: And it wasn't far from where he lived, at the address that appeared in his articles, at 26 rue des Plantes, an Art Deco-style building, composed of duplex artists' studios with the studio on the lower level, bedroom and kitchen, living room on the mezzanine, with double-height windows. Max Ernst lived there. Marx Ernst was the one who illustrated my "Logique sans peine" (1966 edition) by Lewis Carroll, published by Hermann, the publisher of the first Bourbaki. You bring back memories. If memory serves, there's a plaque in memory of Jean Moulin, who had rented a studio there. 

 

G.L.: Yes. I could walk from INSEE to rue des Plantes, so we were able to talk.

 

 H.H.: What were your contributions? 

 

G.L.: Theoretical contributions. * Obligation to prove initialization. * Loop Introduction pp. 377 and 378 of the B-Book. Initial (dual) version of  Th. 9.2.1 p. 379 with unions instead of intersections: drastic simplification of hypotheses. 

B-Tool. * Improvement of the sequence machine. * Plagiarism assessment between the initial version (owned by BP) and the new version (STERIA at the time). * Analysis of a new version of the parser, an LR-inspired pushdown automaton replacing recursive descent. Subtleties: generation of saturated inverse polonaises and insertion of implicit operators (substitution and evaluation). 

 

H.H.: Can you tell us about it during the day? And your application of B to the population census. You gave a presentation to the students. I remember the study of administrative geography.  

 

G. L.: Okay. * Tools for the 1990 Population Census 1 * Systematic use of the B-Tool for parsing tools and code generation (Pascal). * TRANS xx. Data storage tool using nonvolatile virtual memory and sequence machines based on an initial idea from JRA. * Basic, system, and application concepts. * Mirror basic concepts. * Development on Unix, portable to IBM OS and DPS7. 

 

Bernard, P., Laffitte, G. (1995). The French population census for 1990. In: Bowen, J.P., Hinchey, M.G. (eds) ZUM ’95: The Z Formal Specification Notation. ZUM 1995. Lecture Notes in Computer Science, vol 967. Springer, Berlin, Heidelberg