On Sat, Jun 27, 2026 at 3:25 AM Andrzej Krzemieński via Paper-reviews < paper-reviews@lists.wg21.org> wrote:
It is not clear what the purpose of the paper is.
The purpose of the paper is to enter information into WG21's historical record. It is institutional memory. This is indicated by the "Intent: inform" metadata field.
The paper has section "Abstract" but it doesn't help figure out the purpose. "This paper traces the path from one to the other." -- yes, but why? Why do you feel it is important to communicate this to WG21 via a paper?
I never put in the abstract the rationale for the paper's existence. My abstracts are always compressed restatements of the entire paper. The first sentence is the brutal one-line summary of the entire paper. The paragraph that follows is an expanded version which explains what the paper does, not why it does it. The purpose of all info-papers is to provide analysis.
Section "Diagnosis". The title is confusing. What is being diagnosed? The sequencing of facts leads the reader to believe that the objection "no way to signal failure from execute()" is the reason why WG21 switched from executors to S/R. There is no evidence for this, but this conclusion, even though not stated in plain text, will be imprinted in the reader's head.
Yes, "no way to signal failure from execute" is exactly the diagnosis. This is what happened. The Networking TS was set aside for this reason.
Section "The Framing" makes a claim that when you adopt the "schedule work" view then no-failure-reporting is a problem, but when you adopt the "schedule continuation" view then no-failure-reporting is not a problem. I expected to find the explanation for why this is the case. I didn't find any. This is a statement (a key statement, apparently) without a supporting evidence/proof.
The opening of Section 3 cites P4094, P4095, and P4096, which contain the supporting evidence. They trace the history of this decision through the papers that led up to it.
Section "The Coroutine Executor": I may remember this incorrectly, but it looks like the intrusive pointer implementation indeed allows you to schedule the continuation without any resource allocation: it just cannot fail. It is not that you are concealing failures: you just guarantee with the design that there will be no failures. If I am right about this, this should be stressed as the crucial point, addressing the objection in P2464R0.
This misses the entire point. The objection in P2464 does not need addressing, because P2464 analyzes the executor through the work-framing which is incorrect. A suspended coroutine is not work. It does not need to have a work scheduler "schedule" it. Work is the thing that the operating system performs while the coroutine is suspended. When the work is finished, then the coroutine is resumed with the result. The resumption of the coroutine itself is not the result.
Section "The Symmetry". It is unclear what the table compares and *why*. "The committee rediscovered that continuations need a continuation-framed executor." -- I do not see how what has been written earlier in the paper would lead to this conclusion. The author assumes that the reader is familiar with the term "continuation-framed executor". This assumption is unfounded.
The paper assumes the reader reads the citations, yes. Otherwise, every paper would have to restate everything it cites, which would become quite cumbersome. Mungo was kind enough to enumerate the cited papers in the original post. The table demonstrates that the Networking TS executor, the P3941R4 infallible scheduler, and the coroutine executor are all the same thing.
Section "The Return" misses a very important observation. the Executor model just didn't allow reporting errors from scheduling continuations. S/R with std::exec::taks on top gives user an option: either the S/R verbose model with scheduling error reporting, or a "nice" interface without the reporting. This is a huge difference.
No, the user does not have an option. P3941R4 makes this clear, and it is quoted in Section 4. Schedulers used with std::execution::task must be infallible. Or to put it differently, they must be used with schedulers which are continuation-framed. For scheduling a sender pipeline, the user does have options. This is acknowledged in the paragraph of Section 6. But sender pipelines are not coroutines. The paper is scoped to the coroutine executor, and the comparison holds there.
The fact that Mungo had to explain in this thread what the paper does, is an indication that the paper itself didn't do its good job at explaining its purpose.
Maybe. Information-only papers are a novelty so this could be just a matter of people getting used to it.
I find something unethical about this "AI-generated paper + human review requests" setup. Traditionally, a review means:
The author made an effort to communicate something that they feel is important, to make sure that the information is accurate and complete, and that it is presented in a way that it will be consumable by the reader. Reviewers are then requested to check if some cracks still exist.
In this AI setup this changes to:
AI just generates some text, indistinguishable at the first glance from reasoning, intention and intellectual honesty. Reviewers are now expected to engage in an effort to put the product to a shape where the paper communicates something they feal important, to make sure that the information is accurate and complete, and that it it presented in a way that it will be consumable by the reader.
The real effort of making sure that a true idea is conveyed cleanly to the readers is moved from the author onto the reviewers. Yet, the author is credited.
No that's not fair at all. I put work into this paper and it reflects the feedback I have received. The machine does what I tell it to do, and in this case the parameters of the paper included being very short. I read it over several times and I also made edits. The paper reflects exactly what I wanted it to say, and it leaves out the things I don't want said. If you take issue with the paper then the fault is mine
I cannot be sure that the content is the AI product. I infer this from my impression where:
1. It has section "Abstract" but after reading it I still don't know what this paper is about and for.
The abstracts of my papers reflect what I think should be in an abstract. I could be wrong on what should be in it, but as written the papers reflect what I believe is important. If you think my definition of what goes in the abstract is wrong, then we can talk about that and I can make adjustments.
2. It has sentences with the shape of logical inferences (there is a premise cause, there is a conclusion clause), but after reading them I do not know what is communicated and why.
Actually, you do know what is communicated. You said it yourself in your review: "Section "The Framing" makes a claim..." This was correct. You didn't read the cited papers, and I'm not sure how that can be fixed. Thanks