Present: Anton Ertl Andrew Haley (from 21st 4:45pm) Peter Knaggs Bernd Paysan Stephen Pelc (plus proxy vote from Willem Botha) Bill Stoddart (from 22nd afternoon) Carsten Strotmann (observer) Leon Wagner Gerald Wodni (absent on morning 22nd until 11:15) Minutes: Accepted minutes of last meeting: 7Y/0/0 Snapshot: Stephen Pelc proposes taking what we have and using it as a snapshot: We commit to use the 10.2 document with the changes from this meeting as the 2011-rc1 snapshot document, with no normative changes except for those resulting from feed back from the public review process. 7Y/0/0 Discussion about whether to go through ANSI: stick to ANSI procedures, but don't go through the organization for now (may change after commercial reasons require it). Review of the document: Discussion about whether the reference implementations should be included in the document. Legacy: Finetuning of the wording of the rationale for FORGET. Reference Forth-94 as "Forth 94" and define somewhere what document this refers to. Correct section numbering for Extension queries and Obsolescent Environmental Queries in Annex A. Fine-Tuning of the wording of the rationale Obsolescent Environmental Queries. X:structures2: correct stack-effects to use the correct address types; some discussion about whether it should not be, e.g., u|df_addr (or x) in order to accomodate the usage for computing offsets, but eventually decided against it (for aesthetic and clarity reasons). Rationale for FIELD: add stack comments and remove the other comments to implementations. 4. Consideration of proposals X:enhanced-locals: some wordsmithing of the Rationale accept with revised Rationale: 7Y/0/0 X:buffer: discussion about alignment issues (deferred) wordsmithing of Rationale vote: 7Y/0/0 X:text-substitution: added restrictions to REPLACES (reflecting implementation practice) Rationale will come from Intro of proposal Vote 8Y/0/0 5. Considerations of draft proposals. X:is Rewordings. 8Y/0/0 FOUND: Addresses a problem that exists, but the proposed solution was not favourably received by the committee. Some response was written by the TC. Actually having this proposal in this section is a mistake (it was not submitted as CfV). Multi-tasking Block Buffers: Minor wordsmithing, then the suggested (non-normative) change was accepted: 9Y/0/0. SOURCE-ID discussion: Implementations that slurp/mmap on including are standard; programs cannot rely on file positions for files that are being text interpreted, nor that changing the file position has any particular effect. 6. Considerations of Snapshot Candidate Release a) Numbering of new words: 8Y/0/0 b) Removal of extension queries: Adjusted for number prefixes (optional) Vote: 9Y/0/0 c) Remove Annex F (reference implementations); they are available through the Forth200x web site in a separate document. 9Y/0/0 d) Remove Annex G (test suite); it is available through the Forth200x web site in a separate document. 9Y/0/0 e) Remove the change log (to be replaced with a summary in another Annex). 9Y/0/0 f) Annex C: Remove it. It is somewhat dated and everyone who reads the standard knows Forth. 9Y/0/0 g) Annex D: Remove most of it and replace it with a summary of the changes from Forth-94. Action: PK; 9Y/0/0 h) Annex E. Revise it. Action: AE Remove description of proposal process. 7) Progress Reports: a) Text Substitution: Done b) I18n: Dependent on Text substitution, now progress is possible c) Cross compiler: A little bit to report d) IEEE FP: Nothing to report. Requires a lot of work. e) File Names and Directories: SP reports that this can be implemented in his embedded native FAT file system implementation, so we may approach consensus. f) Directories: Maybe progress from having text substitution. SP reports that VFX now does source-relative directories through a text substitution macro %idir%. g) Memory access: some discussion 8. Matters arising Documentation requirements: AE will produce a version that removes most of it; ambiguous conditions are collected in an Annex. Compliance and Labeling: PK will move extensions from other chapters into a table here. Incomplete xchars in I/O: consensus that this should work. BP is to draft a response. 9. Next meeting If the snapshot-rc1 is released in 2011, we will have a spring meeting in NL. In any case we will have a meeting at the next EuroForth.