C++ Logo


Advanced search

Re: [wg14/wg21 liaison] [isocpp-lib-ext] RFC: C++ needs to support opening files in "exclusive" mode

From: Niall Douglas <s_sourceforge_at_[hidden]>
Date: Thu, 30 Sep 2021 15:19:06 +0000
On 30/09/2021 16:00, Aaron Ballman via Lib-Ext wrote:

>> Yeah the above is the old wording. I thought Robert Secord from WG14 was
>> supposed to have fixed it?
> My liaison senses were tingling. :-D The only paper I'm aware of on
> the topic is: http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2357.htm
> The meeting minutes on the discussion of that paper are in:
> http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2377.pdf
> Some interesting snippets from the minutes:
> *Straw poll: Do we want to incorporate N2357’s second proposed change
> into C2X? 11/1/11 (In favor, Opposed, Abstain)
> *Straw poll: Do we want any changes to p5 regarding the
> “exclusive” in exclusive file access? 11/1/12 (In favor, Opposed,
> Abstain)
> Niall: I can propose alternative wording. I intend to write exactly
> what POSIX does
> So neither poll gained consensus, and I think we're waiting on Niall
> to produce the follow-up paper.

So I *did* produce alternative wording (I judged it at the time it was
too small for a whole paper), which you've already seen copy and pasted
just there, complete with paragraph numbers. I did that back in 2019,
and I sent it in. Robert Secord then followed up with me about it in a
private email chain thereafter, and as far as I was aware until now, not
my exact proposed wording was chosen, but something close enough was.

So *somebody* on WG14 noticed me sending it in and responded, I guess
somehow between then and now something got dropped. It's a very small
change in the wider scheme of things, so it's easy to happen. In
fairness, had I actually put this into a N-paper, it probably wouldn't
have got dropped. C'est la vie.

Can we act on this now or is it too late? I have a feeling you now will
want a paper? A copy and paste of the email I just quoted is probably
sufficient, it has all relevant information needed for WG14 to make a
decision I think?

In the end, as the meeting minutes show, my new proposed normative text
just replicates the POSIX standard. I think that a safe choice for WG14,
and WG21, to follow.


Received on 2021-09-30 10:19:09