C++ Logo


Advanced search

Re: Meeting in Varna?

From: Jens Maurer <jens.maurer_at_[hidden]>
Date: Mon, 22 May 2023 19:20:23 +0200
On 22/05/2023 14.19, René Ferdinand Rivera Morell wrote:
> On Mon, May 22, 2023 at 2:23 AM Jens Maurer <jens.maurer_at_[hidden]> wrote:
>> (copying Herb)
>> On 22/05/2023 04.47, René Ferdinand Rivera Morell via SG15 wrote:
>>> I haven't seen any mention of SG15 meeting in Varna. And I don't see a
>>> room reservation in the schedule. Is SG15 meeting in Varna? Wondering
>>> because we are supposed to get a vote in to approve a work item for a
>>> formal start of the Ecosystem IS.
>> Herb, I thought we wanted New Work Item Proposals only when we had something
>> resembling a Working Draft with fairly fleshed-out content, to avoid the
>> embarrassment of cancelling a project at the ISO level. What's the current
>> approach here?
> What was communicated to us in Kona22 was that we needed a list of
> items for the working draft and at least one item fully in a working
> draft.

Ok. That means we would have a minimum shippable thing, even if things
go down afterwards.

> Although honestly the directions were not that clear. What we
> have currently is the list of items (see
> https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2023/p2656r2.html#_goals),


> and a PR for the draft with the contents of the introspection item
> (see https://github.com/cplusplus/ecosystem-is/pull/2 -- and
> attached).

Well, that needs very serious work. For instance, it seems


is a valid /version-range/ per the current wording.
Is 01.00002 supposed to be a valid version number?
(Note leading zeros.)

Also, all mentions of grammar non-terminals seem to use regular italics
in the text, not the \grammarterm slanted font.

Also, I recommend to use Unicode character designations for non-alphanumeric
characters that want to be highlighted, avoiding ambiguities with surrounding

"separated by a period (.)" -> "separated by U+002E FULL STOP".

> And there's been some discussion on the possibility of
> moving the contents of the Modules TR into this. We could add
> placeholders for the goal items in the draft if that's what's needed.

That would be a good idea; the Working Draft is what's distributed
alongside the New Work Item Proposal. On the other hand, promising
more than you can deliver isn't good, either.

> Like C++ we are operating with the train model. I.e. something will
> get published, the contents will vary on effort.

Hm... It would be good if the description of the "Boostrap" process
in https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2023/p2656r2.html#_bootstrap
would clearly point out that LWG and CWG are involved; after all
EWG and LEWG are also explicitly mentioned.

It is unclear to me whether CWG (and even more so LWG) would be the
right group to review wording for the Ecosystem IS. On the other hand,
I'm not seeing WG21 passing a straw poll without one of those groups
having reviewed the material.

At what point do you anticipate that wording review for CWG and/or LWG
will be on the agenda? Please add that explicitly to the "Timeline";
this review might generate substantial load for CWG.

Oh, and I'm seeing "--std-info" as a command-line option. I thought
Microsoft, for instance, is using the /xxx option format. Has the use
of "--xxx" been socialized with Microsoft?


Received on 2023-05-22 17:20:29