[cgl_discussion] POC 11/19/03 8:30 AM PST Agenda

Steven Dake sdake at mvista.com
Tue Nov 18 14:58:17 PST 2003

*AR* Please read the errata notes below and form opinions about them
before we meet Wednesday.

* Determine what to do about new errata posted to mailing list.
* opens

The new errata notes I have are:
minor typo on page 50:  AVL 7.1 Process-level Non-intrusive" should be
"AVL 8.1 Process-level Non-intrusive".

SCTP requirement (STD4.) lists RFC 1112 which may be a mistake.  Discuss
adding RFC3309 as a seperate developer note.

Errata from John Mehaffey:
Page 29, item STD.2.20 (thread safe functions) claims to be
useful for POSIX timers, but I believe that it should read
POSIX threads.

Julie's response:
You're right that that sentence is incorrect.  I'd propose we remove the
sentence "This functionality is used in POSIX Timers."  When the
requirement was written, the sentence was referring to the functions
asctime_r, ctime_r, gmtime_r, and localtime_r; however, those actually
aren't part of POSIX timers per se (not part of the TMR tag at least).

Another typo, in PLT.1.0 (Page 32) it states that this was item
2.1.0 in v1.1, but it was item 2.1.3.

Another, larger issue is what to call the items that were brought
forward from the v1.1 spec unchanged (Page 94).
For instance, category platform, item 2.2 (Remote boot support)
Does (or will) this have a PLT.x[.y] number? I don't think we can
just call it PLT.2, as there is already a PLT.2.0 that is on a
different topic.

What happened to v1.1 items and  Were these
subsumed into PLT.1.0, carried forward unchanged, or deleted?

"This functionality should be integrated with, or work in conjunction
with, the functionality specified by requirement 3.3 Application
Heartbeat Monitor listed on page 64."

page 64 doesn't list anything about application heartbeat monitor. In
fact, the OSDL 2.0 specs don't list the invidual requirement of
application heartbeat monitor (or I can't find it).  The true reference
is page 13 of "CGL Requirements Definition 1.1"

More information about the cgl_discussion mailing list