[Cfp-interest 2029] WG14 meeting results for CFP papers

[Rajan Bhakta]

Hi,

Good news: All of our papers scheduled for this week, as well as some that
were not scheduled were accepted by WG14!
Note that there were other issues that touch on us including the Austin
Group's question Fred posted, so this is just a paper summary.

Details are below.

5.8 Thomas, C2X proposal - signbit cleanup [N 2650]
  Straw poll: Does WG14 want to adopt N2650 into C23?
    18/0/2. Consensus to adopt into C23.

5.9 Thomas, C2X proposal - fabs and copysign cleanup [N 2651] - We only
asked for the last two changes to go it as the first couple need work.
  Myers: For the Annex F changes that are not being proposed right now, for
implementation defined assignment convert format may lose information about
the sign representation. This applies to SNaNs for hypot and other
functions.
  Svoboda: For functions that differ on types of input we have different
functions. Should we say something about it still being a floating point
types.
    Rajan: I don't see a need for a change. If WG14 wants one, let us know.
  Tommy: Does this open it up to strings?
    Rajan: No.

  Straw poll: Does WG14 want the change to 7.12.7.3 into C23?
    20/0/3. Consensus to put into C23.

  Straw poll: Does WG14 want the change to 7.12.11.1 into C23?
    17/0/6. Consensus to put into C23.

5.10 Thomas, TS 18661-5 revision [N 2652]
  If plenary is not needed, we want approval to reissue the TS.

  Myers: What happens with part 4b?
    Rajan: We will have new proposals if we don't already.

  Straw poll: Does WG14 want to create a new revision of 18661-5 based on
C23?
    19/0/2. Consensus to revise part 5.
  Action Item: Keaton: Get a new revision started for CFP part 5 via asking
for permission from SC22.

5.11 Thomas, C23 proposal - zeros compare equal [N 2670]
  Svoboda: The committee said 2's complement for integers. This should only
apply to floating point.
    Rajan: Yes, by elimination it does apply only to FP.
  Seacord: Any way to tell if you have a negative zero?
    Rajan: Some library functions (Ex. memcmp), bit compare, unions.
  Ballman: Maybe nice to say explicitly that the 6.5.8 change only applies
to floating point.
    Rajan: Other types may exist. Ex. Fixed point.

  Straw poll: Does WG14 accept N2670 into C23?
    19/1/3. Consensus to put into C23.

7.1 Thomas, C23 proposal - negative values [N 2671]
  Svoboda: -Inf is negative. What about complex numbers?
    Rajan: Complex is something to consider, but I believe it works. i.e.
Not allowed to do < 0, but parts can be compared.
  Seacord: Like the first paragraph better.
  Myers: Have you looked at the rest of the standard. Ex. printf + flag.
This may need to change for wide as well.
    Fred: Yes, I have a paper in the queue for handling that. N2643.
  Ballman: Can this break implementations? Ex. fsqrt?
    Chat (Myers/Bachmann): It is handled in the spec for IEEE.
  Svoboda: Shouldn't we vote on this and N2643 at the same time?
    Keaton: We voted N2643 in already last meeting.
  Seacord: Is the opposite true? i.e. Not negative means -0 is positive?
    Rajan: No. NaN and Zero are not positive or negative.

  Straw poll: Put N2671 into C23?
    16/0/2. Consensus to put into C23.

7.2 Thomas, C23 proposal - 5.2.4.2.2 cleanup [N 2672]
  Straw poll: Put N2672 into C23?
    21/0/1. Consensus to put into C23.

Regards,

Rajan Bhakta
z/OS XL C/C++ Compiler Technical Architect
ISO C Standards Representative (Canada, USA), PL22.11 Chair
C/C++ Compiler Development
rbhakta at us.ibm.com

IBM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20210617/799a46c6/attachment.htm>