Forth standards meeting at EuroForth 2005

On 2005-10-20, 2005-10-21 and 2005-10-23 there was a Forth standards meeting, right before and partly during EuroForth 2005. The participants were:

Willem Botha, CCS, South Africa
Federico de Ceballos, University of Cantabria, Spain
M. Anton Ertl, TU Wien, Austria
N. J. Nelson, Micross, England (Observer)
Peter Knaggs, University of Bournemouth, England
Stephen Pelc, MPE, England
Bill Stoddart, University of Teeside, England (Observer)

The following topics were discussed:

Accepting proposals into the standard document

Federico de Ceballos chaired this part of the meeting.

The four proposals which had gone to a CfV were discussed:

Various changes were suggested that were necessary for inclusion into the standards document. These changes then resulted in adapted versions of the normative parts of these proposals (performed by Anton Ertl). The resulting proposals were then discussed, and various small clarifications and changes were incorporated, resulting in the following normative parts.

These proposals were all accepted with vote results 5Y:0N:0A.

Future proposals

A number of potential topics for future proposals were discussed, and eventually some participants volunteered to write RfDs for the following topics:

Federico de Ceballos

number prefixes 
0 for NIL 
! and @ for 16-bit and 32-bit signed and unsigned integers, bytes, octets

M. Anton Ertl

separate FP stack 
required 
directory stuff in general 
directory handling for included and required 
key names for EKEY results

Stephen Pelc

{ (locals), fp locals, buffer locals 
S\" .\" 
iors can be THROWn 
SYNONYM 
structures

Bill Stoddart

Using TAB, CR, LF, FF in source code

Integration into the standards document

Should new words be put into the established wordsets, or into new ones?

The eventual goal is to usually integrate the new words into existing wordsets with related functionality; in some cases it may be more appropriate to create a new word set. However, as an intermediate step the new proposals will at first be kept separate, to make it easier for readers of the document to see what is changing.

How are the extension-query names reflected in the standard (if at all)?

The glossary header for new words includes the extension-query string for the extension that proposed it. In addition, there will be a chapter or normative appendix that lists all the extensions, their extension-query strings and the components (word definitions etc.) that it consists of.

Should the tests of a proposal or the reference implementation become normative? No. This could lead to conflicting normative sections; also, making the reference implementation normative would lead to overspecification.

Review of the RfD/CfV process

How well is the RfD/CfV process working at generating high-quality proposals for standardisation and getting information about their popularity? What could be improved? Or should we do something completely different?

Many of the participants were not very familar with how well the process worked in practice, and had no suggestions for improvements.

The (normative part of the) proposals required adaption before integrating them into the document, but there was a widespread feeling among the participants that proposals in the form of unambiguous instructions to the document editor (which is the form that would be voted on by the standards committee) would be harder to understand for the CfV audience.

The resulting idea was to have two forms of the proposal, with the same content: First the not-so-formal form used in the CfV, and later a form for integrating it into the document.

Official standards body

Should we run the standard through a standards body like ANSI, ISO, IEEE, etc.? If so, which one?

Some participants consider the blessing of the future standards document by an official standards body very important, and we agreed to work towards this goal by writing the document in the appropriate style, and by keeping documentation about all our steps. However, the general idea was to first develop the document without involving a standards body, and deal with them at the end.

Various candidate standards bodies were discussed; none was decided on, but it might be that going through ANSI again might be the easiest route.

Chairman and Editor

M. Anton Ertl was appointed as the chair of the committee unopposed.

Peter Knaggs was appointed editor unopposed.

Next meeting

After some discussion about possibly having two meetings per year, we decided to just have one meeting per year for now. The next meeting will again be on the day before the next EuroForth conference (probably in Cambridge, date to be announced).

Open questions

Which standard documents should we start from? Peter Knaggs has a version of dpANS99a in LaTeX form, convertible into a fully hyperlinked PDF file. However, it is yet unclear how this document differs from the ANS/ISO Forth documents and dpANS6.

How do we get from the CfV proposal to the form for integration into the document? One opinion was that the original proponent should do it. On a related topic, there was the opinion that a proposal (de-facto) needs a champion in the committee to get approval by the standards committee. So, if the proponent finds a champion in the committee, they could produce the for-the-document version of the proposal together.

Improvements for future standard meetings

The participants should familiarise themselves with the proposals beforehand.
Paper printouts of the proposals should be available at the meeting.
Proposals should be available in the form needed for integration into the standards document in addition to the CfV form.
There should be a champion for the proposal in the meeting.

Anton Ertl