|
|
|
|
|
|
|
|
Prolog Standardization: A Progress Report Roger Scowen
17th June 1994
Appeared in Volume 7/3, August 1994
Keywords: standard.
Summary
The standardization of Prolog -- Part 1 (General Core) is approaching completion, and work continues on Part 2 (Modules). Good progress has been made in 1993-94 but more support is needed if the work is to continue in 1994-95 and until both parts of the standard are published.
Background
The International Standard for Prolog is being defined by an ISO working group whose full title is 'ISO/IEC JTC1 SC22 WG17'. The standard is being developed in two parts: Part 1 (General core), and Part 2 (Modules). Part 1 is now largely complete, and a ballot to approve or disapprove the technical content of a Draft International Standard (DIS) has just started. It will last for four months and the national standard bodies (e.g. ANSI, BSI, DIN) will doubtless consult their Prolog experts. Assuming the ballot is successful, and WG17 do not accept comments that would change the substance of the draft, the standard will be published by ISO/IEC early in 1995. Part 2 is less advanced and CD (Committee Draft) registration is not expected before 1995.
Up to March 1994, this work was funded first by the UK Department of Trade and Industry, and then NPL (National Physical Laboratory) with sponsorship from Association of Logic Programming and Quintus Corporation. NPL's funding has now almost completely ceased and Roger Scowen has left NPL in order to complete the standardization of Prolog and pursue other research. He is currently an unpaid guest worker at NPL with an office, workstation, and email. The costs of computing, printing, and distribution also continue, for the time being, to be borne by National Physical Laboratory. Ideally, another base for the work should be found.
Progress in 1993-94
Highlights of the last twelve months include:
However progress on Part 2 (Modules) has been slow. Klaus Daessler, the project editor, has revised the working draft (WG17 N122, N126), but the objectives, theory and existing practice are still subject to much debate and dissension in WG17.
Roger Scowen is continuing as a (currently unpaid) convener for WG17 and project editor for Part 1. News of the standardization is broadcast by email, but WG17 papers are sent free only to those nominated as members of WG17 by the national standards bodies. Other Prolog experts and companies must either arrange for copies to be forwarded by someone else or become a subscriber or sponsor. Most sponsors will participate because they themselves recognize the benefits of a Prolog standard defined, agreed and published as rapidly as possible.
It is necessary to emphasize that sponsorship will not provide the right to directly influence and define the standard. Content and organizational procedure will continue to be defined according to ISO/IEC JTC1 directives.
Further standardization?
The Prolog community should now consider whether further features of Prolog should be standardized. And if so, whether by ISO or some other group. Candidates include some features which were left out of Part 1 because their inclusion would have delayed its completion (e.g. grammar rules). Other features would fill a semantic gap (e.g. a 'read_token' built-in predicate), or extend the logical and computational framework (e.g. constraints, breadth-first searching, a foreign language interface).
Roger Scowen WG17 (Prolog) Convener & Project editor 9 Birchwood Grove, Hampton Middlesex TW12 3DU, UK Email: rss@seg.npl.co.uk
|
|
|
|
|
|
|
|