Move big paper trilogy and update p numbers

vinniefalco · vinniefalco · commit 0ea72dbf7066 · 2026-03-15T14:33:20.000-07:00
diff --git a/source/d2583-symmetric-transfer.md b/source/d2583-symmetric-transfer.md
@@ -1,6 +1,6 @@
 ---
 title: "Symmetric Transfer and Sender Composition"
-document: D2583R3
+document: P2583R3
 date: 2026-03-10
 reply-to:
   - "Mungo Gill <mungo.gill@me.com>"
diff --git a/source/d4050-task-diversity.md b/source/d4050-task-diversity.md
@@ -11,7 +11,7 @@ audience: LEWG
 
 `std::execution::task<T, Environment>` ([P3552R3](https://wg21.link/p3552r3)<sup>[1]</sup>) is proposed as a lingua franca for coroutine-based asynchronous code. The `Environment` parameter is an open query-response protocol whose interoperability surface is defined by a single concept: `queryable`, which is `destructible`. This paper asks a simple question: when two libraries define different environments, how does one task `co_await` the other? The answer, traced step by step through the specification, is that no general conversion exists. The query set is open by design, and the only adaptation mechanism - `write_env` - requires the caller to know every missing query by name. The risk to the ecosystem is structural, documented by the specification itself, by NVIDIA's reference implementation, by the only production precedent (Boost.Asio), and by `task`'s own author.
 
-This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4053R0](https://wg21.link/p4053r0)<sup>[9]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[10]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[11]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[12]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[13]</sup>, "The Cost of `std::execution` For Networking."
+This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4053R0](https://wg21.link/p4053r0)<sup>[9]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[10]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[11]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[12]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[13]</sup>, "The Case for Coroutines."
 
 ---
 
@@ -405,7 +405,7 @@ The author thanks Gor Nishanov for the coroutine model's explicit support for ta
 10. [P4054R0](https://wg21.link/p4054r0) - "Two Error Models" (Vinnie Falco, 2026). https://wg21.link/p4054r0
 11. [P4055R0](https://wg21.link/p4055r0) - "Consuming Senders from Coroutine-Native Code" (Vinnie Falco, Steve Gerbino, 2026). https://wg21.link/p4055r0
 12. [P4056R0](https://wg21.link/p4056r0) - "Producing Senders from Coroutine-Native Code" (Vinnie Falco, Steve Gerbino, 2026). https://wg21.link/p4056r0
-13. [P4058R0](https://wg21.link/p4058r0) - "The Cost of `std::execution` For Networking" (Vinnie Falco, 2026). https://wg21.link/p4058r0
+13. [P4058R0](https://wg21.link/p4058r0) - "The Case for Coroutines" (Vinnie Falco, 2026). https://wg21.link/p4058r0
 
 ### StackOverflow and GitHub
 
diff --git a/source/d4053-net-comparison.md b/source/d4053-net-comparison.md
@@ -12,7 +12,7 @@ audience: LEWG
 
 Four sender-based TCP echo servers are constructed from [P2300R10](https://wg21.link/p2300r10)<sup>[1]</sup> and [P3552R3](https://wg21.link/p3552r3)<sup>[2]</sup> and compared against a coroutine-native echo server. The sender composition algebra - `when_all` cancellation, `upon_error`, `retry` - does not apply to compound I/O results without losing data, requiring shared state, or converting routine errors to exceptions. One construction - "just use `set_value`" - preserves all data by bypassing the composition algebra entirely, producing code nearly identical to the coroutine version. Both paradigms are equivalent when compound results stay on the value channel. The composition algebra is the sender model's value proposition over coroutines - the reason to accept its additional complexity. If it does not apply to compound I/O results, the programmer pays the cost of the sender model and receives the coroutine model's behavior. The finding is about cost, not defect.
 
-This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[15]</sup>, "On Task Type Diversity"; [P4054R0](https://wg21.link/p4054r0)<sup>[7]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[13]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[14]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[17]</sup>, "The Cost of `std::execution` For Networking."
+This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[15]</sup>, "On Task Type Diversity"; [P4054R0](https://wg21.link/p4054r0)<sup>[7]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[13]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[14]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[17]</sup>, "The Case for Coroutines."
 
 ---
 
@@ -621,7 +621,7 @@ Any person quoted in this paper who believes their words have been presented out
 
 16. Ville Voutilainen, [libunifex-with-qt](https://git.qt.io/vivoutil/libunifex-with-qt) - Qt/stdexec integration examples (2024). https://git.qt.io/vivoutil/libunifex-with-qt
 
-17. [P4058R0](https://wg21.link/p4058r0) - "The Cost of `std::execution` For Networking" (Vinnie Falco, 2026). https://wg21.link/p4058r0
+17. [P4058R0](https://wg21.link/p4058r0) - "The Case for Coroutines" (Vinnie Falco, 2026). https://wg21.link/p4058r0
 
 18. [P3796R1](https://wg21.link/p3796r1) - "Coroutine Task Issues" (Dietmar K&uuml;hl, 2025). https://wg21.link/p3796r1
 
diff --git a/source/d4054-two-error-models.md b/source/d4054-two-error-models.md
@@ -17,7 +17,7 @@ The question launched a discussion. Ville Voutilainen, Ian Petersen, Jens Maurer
 
 The finding: both coroutines and senders have an abstraction floor - a boundary where compound results are reduced to a single value. The coroutine floor is `throw`. The sender floor is `set_error`. Both destroy compound data when crossed. The difference is where the floor sits relative to composition. In coroutines, the floor is opt-in. In senders, the composition algebra lives above it.
 
-This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[14]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[6]</sup>, "Sender I/O: A Constructed Comparison"; [P4055R0](https://wg21.link/p4055r0)<sup>[12]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[13]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[18]</sup>, "The Cost of `std::execution` For Networking."
+This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[14]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[6]</sup>, "Sender I/O: A Constructed Comparison"; [P4055R0](https://wg21.link/p4055r0)<sup>[12]</sup>, "Consuming Senders from Coroutine-Native Code"; [P4056R0](https://wg21.link/p4056r0)<sup>[13]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[18]</sup>, "The Case for Coroutines."
 
 ---
 
@@ -437,6 +437,6 @@ Any quoted participant who wishes a passage retracted or revised may contact the
 
 17. Ian Petersen, four sender implementations of channel dispatch, March 2026. https://godbolt.org/z/7W51hYE7c
 
-18. [P4058R0](https://wg21.link/p4058r0) - "The Cost of `std::execution` For Networking" (Vinnie Falco, 2026). https://wg21.link/p4058r0
+18. [P4058R0](https://wg21.link/p4058r0) - "The Case for Coroutines" (Vinnie Falco, 2026). https://wg21.link/p4058r0
 
 19. [P1525R1](https://wg21.link/p1525r1) - "One-Way execute is a Poor Basis Operation" (Eric Niebler, Kirk Shoop, Lewis Baker, Lee Howes et al., 2020). https://wg21.link/p1525r1
diff --git a/source/d4055-sender-bridge.md b/source/d4055-sender-bridge.md
@@ -12,7 +12,7 @@ audience: LEWG
 
 An `IoAwaitable` bridge ([P4003R0](https://wg21.link/p4003r0)<sup>[3]</sup>) consumes `std::execution` ([P2300R10](https://wg21.link/p2300r10)<sup>[1]</sup>) senders with inline operation state, correct stop token propagation, and automatic executor dispatch-back. The bridge is one class template. The complete implementation is in Appendix A.
 
-This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[13]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[2]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[11]</sup>, "Two Error Models"; [P4056R0](https://wg21.link/p4056r0)<sup>[12]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[14]</sup>, "The Cost of `std::execution` For Networking."
+This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[13]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[2]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[11]</sup>, "Two Error Models"; [P4056R0](https://wg21.link/p4056r0)<sup>[12]</sup>, "Producing Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[14]</sup>, "The Case for Coroutines."
 
 ---
 
@@ -151,7 +151,7 @@ The authors thank Dietmar K&uuml;hl for `beman::execution`<sup>[5]</sup> and for
 
 13. [P4050R0](https://wg21.link/p4050r0) - "On Task Type Diversity" (Vinnie Falco, 2026). https://wg21.link/p4050r0
 
-14. [P4058R0](https://wg21.link/p4058r0) - "The Cost of `std::execution` For Networking" (Vinnie Falco, 2026). https://wg21.link/p4058r0
+14. [P4058R0](https://wg21.link/p4058r0) - "The Case for Coroutines" (Vinnie Falco, 2026). https://wg21.link/p4058r0
 
 ---
 
diff --git a/source/d4056-awaitable-sender.md b/source/d4056-awaitable-sender.md
@@ -12,7 +12,7 @@ audience: LEWG
 
 An `IoAwaitable` ([P4003R0](https://wg21.link/p4003r0)<sup>[2]</sup>) can be wrapped as a `std::execution` sender. Awaitables returning `void` or a single value map to `set_value`. Awaitables returning `error_code` map to `set_value()` on success and `set_error(ec)` on failure - no exceptions. Awaitables returning compound I/O results - any tuple-like whose first element is `error_code` with additional elements - are rejected at compile time. The coroutine body is the translation layer: it inspects the compound result, reduces it to an `error_code`, and returns that. The bridge routes the `error_code` through the three channels without exceptions.
 
-This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[14]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[7]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[13]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[6]</sup>, "Consuming Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[15]</sup>, "The Cost of `std::execution` For Networking."
+This paper is one of a suite of six that examines the relationship between compound I/O results and the sender three-channel model. The companion papers are [P4050R0](https://wg21.link/p4050r0)<sup>[14]</sup>, "On Task Type Diversity"; [P4053R0](https://wg21.link/p4053r0)<sup>[7]</sup>, "Sender I/O: A Constructed Comparison"; [P4054R0](https://wg21.link/p4054r0)<sup>[13]</sup>, "Two Error Models"; [P4055R0](https://wg21.link/p4055r0)<sup>[6]</sup>, "Consuming Senders from Coroutine-Native Code"; and [P4058R0](https://wg21.link/p4058r0)<sup>[15]</sup>, "The Case for Coroutines."
 
 ---
 
@@ -241,7 +241,7 @@ The authors thank Dietmar K&uuml;hl for `beman::execution`<sup>[4]</sup> and for
 
 14. [P4050R0](https://wg21.link/p4050r0) - "On Task Type Diversity" (Vinnie Falco, 2026). https://wg21.link/p4050r0
 
-15. [P4058R0](https://wg21.link/p4058r0) - "The Cost of `std::execution` For Networking" (Vinnie Falco, 2026). https://wg21.link/p4058r0
+15. [P4058R0](https://wg21.link/p4058r0) - "The Case for Coroutines" (Vinnie Falco, 2026). https://wg21.link/p4058r0
 
 ---
 
diff --git a/source/d4059-big-claims.md b/source/d4059-big-claims.md
diff --git a/source/d4060-big-reputation.md b/source/d4060-big-reputation.md
diff --git a/source/d4061-big-humility.md b/source/d4061-big-humility.md
