domo@tsa.co.uk (Dominic Dunlop) (02/10/91)
The following is a summary (without pictures or colour) of the
presentation materials (well, hand-written foils) that I used at the
USENIX Standards birds-of-a-feather session on 24th January. My
interpolations and clarifications appear in square brackets.
Foil 1.
USENIX
Winter 1991
Standards b.o.f.
Dominic Dunlop
ISO POSIX Observer, USENIX/EurOpen
IEEE-CS TCOS Institutional Representative, EurOpen
POSIX Internationalization Rapporteur to JTC1/SC22/WG15 (POSIX), U.K.
[I was justly accused of going over the top with this list of
my responsibilities]
Jeff Haemer
USENIX Snitch Editor
Foil 2.
This b.o.f. is ``slightly structured''
1. Introductions
2. Who's who?
-- Standards organizations
-- Special interest groups
3. What's what?
-- Standards activities related to ``open systems''
[- Including new activities relating to an IEEE standard for the
SPARC microcomputer architecture, and an industry grouping
developing a 16-bit wide character set for worldwide use.]
4. Where's USENIX?
-- Your dues dollars working for you(?)
5. Why, oh why?
-- General discussion
Foil 3.
Organizations -- top down.
[This was a picture. The issue is discussed in a forthcoming report
on the New Orleans IEEE-CS TCOS meeting of 7th-11th January, which
will shortly be posted to comp.std.unix.]
Foil 4.
Organizations -- bottom up.
[Another picture, featuring meat grinders (mincers in British English)
of varying sizes standing in for the various groups which get to process
POSIX standards. Again, this issue is discussed in the forthcoming
report.]
Foil 5.
Utter speculation
``ISO'' [strictly, Joint Technical Committee 1 of ISO and the IEC]
unimpressed with ANSI's recent performance as a [standards] development
agency [for information technology].
-- [Bad experiences with] Fortran, C, (SQL) [ANSI not strictly a
development agency for ISO SQL standard, but was heavily
involved]
-- POSIX OK, but could get caught in cross-fire
-- ANSI may, or, more likely, may not be development agency for C++
ISO standard
-- ISO may or may not decree that C++ is to be true superset of ISO C
(revision to IS 9899 [handled by existing SC22/WG14]) [We learned
at the bof that WG14 has said that it does not want to pick up C++
work]
IEEE _could_ get liaison status at ISO (bypass ANSI)
(See also Datamation, Sept. 15 1990 (vol.36, no. 18) [``The Standards
Process Breaks Down'' -- reproduced in part as comp.std.unix Volume 21,
Number 177]
Foils 6-8.
IEEE POSIX
[In the following, ?? means that information on the ballot target was
not available to me when I drew up the foils. In these cases, a ballot
target may or may not in fact exist.]
Project Title [abbreviated] Ballot target
1003.0 POSIX Guide mock 2Q91
1003.1 System interface complete
1003.1A Extensions 1991?
1003.1LIS Language independent version 1991?
1003.2 Shell & utilities in ballot
1003.2A User portability extension in ballot
1003.3 Generic test methods complete
1003.3.1 Test methods for 1003.1 1Q91
1003.3.2 Test methods for 1003.2 ??
1003.4 Real time in ballot
1003.4A Threads ??
1003.4B Real time language independent version ??
1003.4C Extensions to real time ??
1003.5 Ada bindings ??
1003.6 Security 2Q91
1003.7 System administration mock 4Q91
real 4Q92
1003.8 Transparent file access 2Q91
1003.9 FORTRAN bindings in ballot
1003.10 Supercomputing application environment profile none
1003.11 Transaction processing AEP mock 4Q91
1003.12 Protocol independent interface none
1003.13 Real time AEP none
1003.14 Multiprocessing AEP 4Q91
1003.15 Batch services none
1003.16 C bindings 4Q91?
1003.17 Name/directory services none
1003.18 POSIX AEP none
1201.1 Windowing toolkit ??
1201.2 Drivability recommended practice ??
1224 X.400 application programming interface 2Q91
1237 Remote procedure call KILLED
(-> ANSI X3T4)
1238 ) Common OSI & FTAM interface ??
1238.1)
Foil 9.
ISO Activity
POSIX.1 = 1003.1:1990 = ISO 9945-1:1990
Approved and available
Buy from IEEE:
(800) 678-IEEE
+1 908 981 1393
$75.00 -- normal price
$52.50 -- IEEE members
NOT available by ftp or in machine-readable form.
-- IEEE investigating electronic draft distribution to working
groups
-- (just maybe) CD-ROM for published standards
-- concerns about loss of income, [and of] passing off [amended
versions of the standard as the real thing.
Foil 10.
ISO (in)activity
Language independence: summary
a. C-based standards first
b. Language-independent with C bindings second
c. ``Thin'' bindings to other languages no sooner than b.
Summary of the summary:
No Ada or FORTRAN bindings to POSIX in ISO for quiet a while.
Foil 11.
An interesting development
IEEE Project 1794 -- Open microprocessor architecture
(decoded: a project to make SPARC a standard -- subject, of course, to
the normal considerations of consensus.)
Contact: Clyde Camp
Phone: +1 214 995 0407
Email: camp@csc.ti.com
[There was a fair amount of mayhem in the bof when those present learned
of this project, which was approved in September 1990. If you are
interested, please contact Clyde.]
Foil 12.
A <your-adjective-here> development
Unicode
-- 16 bit wide fixed width (well, I'd argue about that [because of
its floating diacritical (accent) characters]) character set
-- intended to handle all (most) of the world all (most) of the time
-- developed by Apple, Claris, IBM, Metaphor, Microsoft, NeXT, Sun,
Xerox, Research Libraries Group and others [all U.S.-based]
-- incompatible alternative to the ISO way [draft international
standard 10646]
-- mail list: unicode@sun.com [try unicode-request@sun.com to
subscribe]
-- review copies of ``final draft'' from Asmus Freytag --
microsoft!asmusf@uunet.uu.net [I misspelled this on the foil!]
[This news too was greeted with displeasure. The comment ``Sounds like
an EBCDIC for the 90's'' one to savour.]
Foil 13.
USENIX activity
Snitch reports
-- Timely reports from volunteer working group members
-- Overview from editor
-- Published in ;login: [EurOpen Newsletter] and comp.std.unix
ISO monitor
-- Like snitches, but paid
-- Informal liaison to WG15 (ISO POSIX)
IEEE TCOS Institution representative
-- Politicking
-- Balloting
-- White papers
Foil 14.
Free for all (at last)
[This was intended to herald general discussion, but, as we'd been
discussing things generally for the preceding two hours, we adjourned
to the places of refreshment of our choice instead.]
--
Dominic Dunlop