Subj : Re: Non-strictly-conforming and unspecified versus undefined behavior
To   : comp.programming.threads,comp.std.c
From : Antoine Leca
Date : Thu Mar 03 2005 04:38 pm

Follow-up to comp.std.c, obviously.

En cv6a2a$2oh$1@nntp.webmaster.com, David Schwartz va escriure:
>> I didn't know until just recently that POSIX also uses the term
>> [strictly conforming] (with a different
>> meaning, of course); I suspect most of the comp.std.c regulars
>> weren't aware of that either.
>
>     It has to.

Sorry, but no. If you are to come to France, you probably will have to learn
a bit how to read French, perhaps how to say "bonjour" (hello).

However, if you do _not_ come here, nobody will ask you to learn French. I
personnaly will find it a pity, but I cannot object to it.
So it is inapropriate to object to the lack of knowledge of the Posix
specific vocabulary from part of people which do have to deal with this
Standard.

You could raise the point that it was inappropriate from part of the people
from comp.std.c to stick to C's vision of s.c. to your points, and I might
agree with you; I would just remark that it seems to me that using the same
term for two things so opposed seems to me a mistake from part of
standardizing authorities; occurring both uses in the same time frame, same
spacial locations, and by a number of persons standing in both committees is
only adding the insult to the injury, in my eyes.


> The C definition of "strictly-conforming" is useless
> in a multithreaded case.

Sure. As you certainly realize now, however, this expression is highly
praised in the (peculiar) comp.std.c forum; the resulting quiproquo (hope
you get this word) is in fact funny to read back again.
I realize it was not funny at all when you first wrote.


>     Please remove further followups to comp.std.c. Anything having
> to do with multithreading is not strictly-conforming to the C
> standard because it invokes undefined behavior, end of story.

You could have done that yourself (with the interface you are using, you
should toggle on the View/AllHeaders, then using the revealed FollowupTo
field) and in fact it should have been done by all the people who understood
at some point of the discussion that it was crossposted.
It is explained in the netiquette for Usenet that crossposting should be
redirected to a single forum, where the interested parties should move to;
here, it is clear to understand why such a rule developped in the first
place.


Antoine

.