[cfp-interest 3379] Re: CFP accomplishment and future

[Mike Cowlishaw]

Interesting question. 
 
As I recently suggested on the 754 list, it might be a good idea to start
pruning some of the ancient 'dead wood' in 754 that is now no longer useful.
 
Perhaps this would be a useful activity for CFP, too.  That is, review the
additions made by CFP over those 16 years (and, of course, the relevant
material that existed before then) .. and see if any can be usefully
consigned to non-normative annexes or removed completely.
 
Mike
 


Publication of the TS-4 and TS-5 revisions completes CFP's purpose as stated
on the CFP webpage: "Develop a C language binding for ISO/IEC 60559
Floating-Point Arithmetic." C23 together with the two TS revisions specify C
language binding for all required and reasonably applicable recommended
features of the latest version of ISO/IEC 60559. 

CFP started in Feb 2009--a 16-year effort! That is a very long time for a
tech project, but it could not have been shorter. We started with C99, based
on IEEE 754-1985. C23 was the next major C revision, which was the first
opportunity to make the upgrade needed to support the 2009 and 2019
revisions of the floating-point standard. 

What is CFP's purpose now? Based on the direction we're going, the purpose
appears to be to:

1. Advise WG14 on maintenance of C support for ISO/IEC 60559, i.e. propose
bug fixes and clarifications, and comment on related new C proposals.
2. Advise WG14 on issues related to floating point in general.
3. Provide liaison between WG14 and the IEEE 754 Working Group.

Your thoughts?


- Jim Thomas


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20250211/2482197e/attachment.htm>