[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Edinburgh meeting



Dear Participant,

Let me first remind you that

     The FINAL DRAFT of the Tentative Design Summary for
     CASL, the COFI Algebraic Specification Language, is now available.

     WWW URL: http://www.brics.dk/Projects/CoFI

     Please read it!
     Deadline for comments: Monday 9 December

Now it is time for the TOOLS task group to launch.
Please read carefully the minutes of the TOOLS task group meeting
in Edinburgh and react if you are interested in participating to 
the project. Do not forget that the success of CoFI relies on 
YOUR participation.

Best regards,

Helene

==================================================================

TOOLS task group Meeting

Edinburgh, November 11, 1996

Participants:

Egido Astesiano (EA)
Michel Bidoit   (MB)
Helene Kirchner (HK, chair)
Bernd Krieg-Brueckner (BKB)
Peter Mosses    (PM)
Michal Walicki  (MW)

These minutes are written from the notes taken by HK and MB.

AGENDA:

Aims and scope 

Impact of Tools on the language 
- concrete syntax
- annotations

Minimal set of tools
- Latex style 
- Emacs mode
- Interchange format
- Parser
- Relation to existing tools for sublanguages

Desirable set of tools
- Catalogue of existing tools

Ressources (who does what)


DECISIONS:

I. Aims and scope: 

Polish and update the current proposal approved
during the previous tools task group meeting in Munich.
The final version should be send to PM 
by Sunday 8 December. A first version is goind to be sent 
to the mailing list. 
Please send comments and reactions to HK before Friday 6 December.


II. Impact of Tools on the language: 

1. Concrete syntax

The following steps for elaborating the concrete syntax are proposed:

a) gather the requirements for the concrete syntax from people having 
experienced the implementation of parsers for different 
kinds of syntaxes.
b) propose a concrete syntax for CASL.
Complete proposals (with examples) are solicited  from
anyone who cares about such matters.
c) bring the proposal to discussion.

BKB proposes a questionaire for the step a) to be send to the Tools 
mailing list as soon as possible.

According to previous minutes, people interested in working
on concrete syntax were 
BKB, MB, Christine Choppy, Gianna Reggio and Olaf Owe.
(Reactions and participation to points a) and b) are expected
from them in particular!)

The design of a concrete syntax for CASL is a task of the
Language Design Group and the initiative is left to it for b)
and c) on the basis of any requirements/constraints proposed 
by the Tools group, following a).



2. Annotations and attributes

The concept of attributes used here subsumes all annotations 
and comments.

The notion of abstract syntax trees with attributes for CASL 
has been studied in a  note  written by PM,
based on the approach taken in the Z standard, which all Tools
participants are encouraged to read. It is available 
at URL: http://www.comlab.ox.ac.uk/oucl/groups/zstandards/
  
The note should be revised according to version 0.95 of CASL and will
be sent to cofi-tools for comments by early December.


3. Interchange format

It looks like the most urgent thing to work on.
An interchange format defines a portable representation of CASL,
allowing CASL documents to be transmitted between different products 
or machines.
The interchange format should support attributes to encode layout 
information  and other information which is not considered to be 
part of the structure of a CASL specification.

Different experiences should be gathered:
- Z interchange format based on SGML (PM and BKB)
- GIPE and the interchange format used in ASF+SDF (PM)
- SALSA interchange format (MB, Christine Choppy, Frederic Voisin).
(Please read their paper on this topic in the Proceedings of the 
Oslo WADT, lncs 1130, pp102-124).

A study note on each should be provided for December 30.


4. Parser

BKB proposes to use Isabelle to prototype static semantics for CASL
and get a parser from it. Several concrete syntaxes could be 
experimented until December 30. 

An alternative is to use the ASF+SDF meta-environment
to generate parsers for sublanguages of CASL.
Jan Bergstra is to be asked for help with using ASF+SDF 
for translating existing languages into interchange format.


5. Emacs and LaTeX modes

PM mentioned the following paper at Tapsoft 95:
"ASF+SDF - Documentation Tools for Free"
by M G J van der Brand and E Visser, LNCS 915, pp 803-4.
(Description of a Tool Demo.)

A proposition is to have a LaTeX style for use independently of any
pretty-printing system, like the zed or oz styles for Z.

We need a volunteer to look at these points by the end of december.


III. Desirable set of tools

The question of elaborating a catalogue of existing tools is postponed 
until january.


IV. Next meeting

The next meeting on tools will be held in Paris, back to back 
with POPL and will be hosted by ENS Cachan.
Further information and a precise date will be sent soon.