C++26 Fails Its ISO Ballot, With 9 of 22 National Bodies Voting No

The draft international standard for C++26 was disapproved in its ISO ballot, according to a post by Robert Seacord on LinkedIn. He reports that 9 of 22 national member bodies voted no, against a limit of 25 percent negative votes.

Seacord's post gives two tests and a result for each. On the second, the individual-expert vote, 12 of 20 voted yes, or 60 percent, against a required 66.66 percent. He called it "a sort of a sad day for the C++ community" and said he hoped it could "heal and move on". The post says nothing about what the committee does next.

Test in the ballot	Required	Result reported
Individual experts voting yes	At least 66.66%	12 of 20 (60%)
National bodies voting no	25% or fewer	9 of 22 (41%)

What the 41 percent is made of

The post does not explain the no votes. Commenter José Daniel García wrote under it that 12 countries abstained, some because they could not agree a common position. That is a comment on a post, not a ballot record, and no ISO or WG21 page listing the votes by country had turned up by Oct. 25.

The reading most people will reach for is contracts, the feature that lets code declare preconditions and postconditions the compiler can check. Herb Sutter, who says he chaired the ISO C++ committee from 2002 to 2025 and writes in a personal capacity, reported that the March 2026 plenary vote to finalize C++26 was 114 in favor, 12 opposed and 3 abstaining. He called it non-unanimous "primarily because of the sustained concerns about contracts" and concluded that "the ISO C++ committee still wants contracts, and so contracts have stayed in C++26".

That vote was of the committee, not of national bodies. A paper dated Aug. 14, 2026 and titled Returning C++26 for the Evaluation It Skipped, by Vinnie Falco, Ville Voutilainen, José Daniel García Sánchez and John Spicer, cites the contracts papers P2899 and P2900 among its references. The text of the PDF could not be read in this reporting, so what it asks national bodies to do is not reported here. Nothing fetched connects any country's no vote to that paper.

The step before the ballot

Sutter's March trip report says the committee was "now producing the final document to be sent out for its international approval ballot". It does not say what happens if that ballot fails. In February 2025 the committee had voted 100 in favor, 14 opposed and 12 abstaining to add contracts to the working draft.

Steve Downey questioned the arithmetic in Seacord's comments. The expert-vote figures imply 8 opposed, while 9 is the separate count of negative national bodies, so the two numbers describe different electorates.

What is not yet known

The post does not say whether the committee will revise the draft and ballot again, and Sutter's report does not cover a failed ballot. Commenter Roger Scott asked whether cutting scope and trying again is the way forward, a question he did not answer.

Seacord's counts imply 20 individual experts and 22 national bodies voting. Projects such as the linker in Wild 0.11.0 and the benchmark dispute over Tov ship without any ballot at all.

The next fixed point is whatever ISO and the committee publish about the ballot. Neither had done so in any page fetched in this reporting.