[Cfp-interest] WG14 IEEE 754-C binding meeting minutes 2017/05/23
Rajan Bhakta
rbhakta at us.ibm.com
Tue May 23 10:49:28 PDT 2017
Attendees: Rajan, Jim, Mike, Ian, David C., David H.
New agenda items:
Teleconference number backup.
Last meeting action items:
Rajan: Bring up "Understanding" points 4 and 5 when discussing DR501
(from Jim's email on 2017/03/17). - Done.
David H: Ask about the IEEE dependencies (rounding mode, infinities)
for the augmented functions and how they could be used for non-IEEE
formats. - Done.
Jim: Ask David Keaton what we should be doing to close off the group
or extend it (IEEE-754:2018 binding for example). - Done.
New action items:
Jim: Reword
http://wiki.edg.com/pub/CFP/WebHome/augop_spec-20170422.pdf to have the
additions done in Annex F only.
Jim: Work http://wiki.edg.com/pub/CFP/WebHome/min-max_names.pdf into a
specification for the main body of the standard. Also add a footnote
saying for implementations without NaN's the functions have the same
results.
All: Look into any other teleconferencing facility (Ex. Cisco Webex,
Skype, etc.) to have a backup in case the main one fails.
Next Meetings:
Tuesday June 20th, 2017, 12:00 EDT, 9:00 PST
Tuesday July 11th, 2017, 12:00 EDT, 9:00 PST
Same teleconference number.
Discussion:
IEEE 754 revision:
Progress on min/max and augmented functions.
Augmented for non-IEEE: Depends on what the other formats handle it.
Non-issue from the 754 committee.
Result: Put in Annex F when we work on this.
A meeting will coincide with the ARITH conference (Sunday, July
23rd).
NaN encoding with reserved bits is being discussed.
Looking at a distinction between error NaNs and missing value
NaNs.
Not compatible change.
Avoids the need of separate functions for each use of NaN.
Thinking of finding a way to make the two kinds compatible.
C++ liaison:
No update.
WG14 meeting:
Note sent out by Rajan in April.
C2x proposals:
See note sent by Rajan in April 6th.
DRs (
http://wiki.edg.com/pub/CFP/WebHome/Defect_reports_-_TS_18661-20170425_Sheet1.pdf?twiki_redirect_cache=0d161c3411487fb5082b05aa84198691
):
DR9: See if Blaine had the proposed changes from our paper folded
into the DR.
DR11: For binary it is the number of digits after the decimal point,
for decimal it is the total number of digits.
The reason is %a is based off of the %e formatting for binary, for
decimal it was modelled after %g.
Was already this way in the TR.
This DR was due to the TR being underspecified.
Result: Keep it as is and push the understanding that the TR had
it this way.
C DR501:
Jim: Change the definition of DECIMAL_DIG to the "widest supported
floating POINT type". A floating point type fits the floating point model
and does not have to be just a float, double or long double that "floating
type" would mean.
A similar change could be made for float_t and double_t.
Result: Make this a proposal to let the current DR close.
We can integrate all the changes to the TS's (from these DR's) to
get base documents for C2X input.
Binding for IEEE 754-2018:
augmented add and multiple operations (
http://wiki.edg.com/pub/CFP/WebHome/augop_spec-20170422.pdf):
*Jim: Reword to be redone to have everything in Annex F only.
min/max operations (
http://wiki.edg.com/pub/CFP/WebHome/min-max_names.pdf):
First set uses the 754 names with some tweaks (no camel-case,
num/mag abbreviations, using _'s).
Feedback was to use the first set.
*Jim: Work into specification for the main body of the standard.
Add a footnote saying for implementations without NaN's the functions have
the same results.
Other:
Joseph Myers 14704 email (missing underscores for HUGE_VAL*):
Will be handled editorially.
Teleconference backup:
Cisco webex? Free?
*Look into any other teleconferencing facility (Ex. Cisco Webex,
Skype, etc.)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20170523/7eecb13c/attachment.html
More information about the Cfp-interest
mailing list