Minutes from the September 19 2002 Teleconference Austin-139 Page 1 of 1
Submitted by Andrew Josey, The Open Group.              September 20, 2002


Attendees:

Andrew Josey, Chair
Mark Brown, TOG OR
Don Cragun, PASC OR
Ulrich Drepper
Joanna Farley
Nick Stoughton, WG15 OR


The purpose of the call was to review the aardvark reports
for TC1 draft 4, and also other open defect reports that had not
made it for TC1.

XBDtc1d4 ERN 1 Accept
XBDtc1d4 ERN 2 Accept as marked


 Add _SC_RAW_SOCKETS before _SC_RE_DUP_MAX

 CROSS VOLUME CHANGE

 In sysconf()
 add to the table in alphabetical order

 _POSIX2_PBS_CHECKPOINT  _SC_2_PBS_CHECKPOINT

 _POSIX_IPV6     _SC_IPV6

 _POSIX_RAW_SOCKETS      _SC_RAW_SOCKETS

 _POSIX_SYMLOOP_MAX      _SC_SYMLOOP_MAX

 _XOPEN_STREAMS  _SC_XOPEN_STREAMS


Final text aardvark
---------------------


XSHft6 ERN 1 Reject 

ISO C and XSH  state that
if exit() is called more than once, the behavior is undefined,
so no change is needed.


XSHft6 ERN 2  OPEN

Forward the request to the SSWG real-time group 


XSHft6 ERN 3 Accept

Send this down the interpretations track, it is a defect,
proposed change as suggested.

XSHft6 ERN 4 OPEN

Forward to the SSWG real-time working group.

There was some debate about whether you would not want to check
the timeout flag under some conditions even if it were invalid.
There appear to be two choices:

(ed note i suspect my notes need assistance here...)

1.Add a shall fail error and indicate that if it would block
and the time needs to be checked if the time value was
bad, and a may fail, if ithe function would not block and the 
time value was bad

or 

2. Only look at the time value if you don't need to block.

This should probably apply also to 

pthread_rwlock_timedwrlock
pthread_rwlock_timedrdlock

The change would be similar to that of the EINVAL error
on the sigtimedwait() page.


XSHft6 ERN 5 Accept
Send this down the interpretations track, it is a defect,
proposed change as suggested.

XSHft6 ERN 6 OPEN

wait and see on the decision for XSH ERN 4


XSHft6 ERN 7 Accept
Send this down the interpretations track, it is a defect,
proposed change as suggested.



XCUft6 ERN 1 Accept
Send this down the interpretations track, it is a defect,
proposed change as suggested.


XCUft6 ERN 2 Accept
Send this down the interpretations track, it is a defect,
proposed change as suggested.


XBDft6 ERN 1  Accept
Send this down the interpretations track, it is a defect,
proposed change as suggested.



Other business:

Andrew noted that he had received the "balance report" from the IEEE
for the formation of the TC1 ballot group, as follows:

Interest Category  Eligible Voters Percent of Group 

User 		     15                 39 

Producer 	      4                 10 

General Interest     17                 44 

Government 	      2                  5 

Total Eligible Voters 38

Since no group has over 50% of the ballot group this is  thought to be
ok for balance.

Andrew reported that his top priority was to get the draft 5 out,
and to commence the IEEE ballot as soon as possible, there is a cutoff
at November 1st for the December Standards Board meeting, ideally
we need to complete the ballot before then, get over 75% 
approval and commence a recirculation, if necessary, before November 1st
(which is tight). At the moment we are in the hands of the IEEE,
whether they can start the ballot promptly.