P2992 R1 Attribute [[discard("reason")]]

[wg21bot]

P2992R0 Attribute [[discard]] and attributes on expressions (Giuseppe D'Angelo)

[erichkeane]

EWGI discussed P2992R0 during the Monday Afternoon session in Kona. The following polls were taken:

EWGI believes the application of [[discard]] on expressions is well motivated, and should be included in future versions of the paper.

SF	F	N	A	SA
0	2	4	4	0

Result: Not Consensus

Given the committee's limited time, EWGI believes P2992R0 is sufficiently developed and motivated to forward to EWG.

SF	F	N	A	SA
1	5	3	2	0

Result: Consensus

[wg21bot]

P2992R1 Attribute [[discard("reason")]] (Giuseppe D'Angelo)

[erichkeane]

Note that we shouldn't see this paper until #1761 develops further as that provides the infrastructure to allow that to happen. That said, the 'split' of the papers should likely go the 'other way'. As this is the paper that was forwarded, THIS should contain the "attribute on expressions" paper (as procedurally, that was what was forwarded), and the [[discard("reason")]] should happen separately/needs to go to EWGI.

[dangelog]

Hi,

I might have misunderstood the results of the polls. This paper's R0 proposed both the [[discard]] attribute, as well as attributes on expressions. In there, [[discard]] was proposed as an attribute on statements or expressions.

The first poll got no consensus for keeping [[discard]] on expressions. So in R1 I reduced the paper to just proposing the attribute, and propose it on statements only. This is self-contained -- it doesn't depend on the other paper, although this results in a smaller applicability.

Proposing attributes on expressions (in "general") got instead split in P3093. Of course that feature has a direct use-case in the context of the [[discard]] attribute; but, as I said, I interpreted the poll as a lack of appetite for it. Clearly, if both features get individually merged, I'm going to ask to re-evaluate that poll and make [[discard]] work on expressions as well.

Did I get the split wrong? Possibly; I'm happy to rework this if it makes things easier, or present both papers at the same time.

[jensmaurer]

This paper tracker is for administrative purposes only. If there are questions about the meaning of polls and the next steps of a paper, please discuss them directly with the chair of the respective subgroup.