WG21 Croydon Trip Report
WG21 ISO C++ Standards Committee
Meeting: WG21 ISO C++ Standards Committee Dates: 23-28 March 2026 Location: Hilton London Croydon Sessions: EWG, LEWG, LEWGI Author: Vinnie Falco
This was my second in-person WG21 meeting. It exceeded every expectation I had. The committee is full of brilliant, dedicated people, and I came away from the week energized about what is ahead.
Sunday, 22 March - Arrival
I arrived at the Hilton London Croydon on Sunday afternoon. Harry Bott, our CEO at C++ Alliance, had gotten in earlier - he shared an Uber from Gatwick with Daveed Vandevoorde. This was the first time Harry and I met in person. After months of working together remotely, shaking hands in the hotel lobby was a moment I will not forget.
We took an Uber to a local Croydon mall so Harry could pick up a light coat. He had flown in from Florida and the English weather was not cooperating. A practical start to an extraordinary week.
The C++ Alliance had strong representation at this meeting: Harry Bott, Mungo Gill, Matheus Izvekov, Matthias Wippich, and our assistant Emma, who was on site from Tuesday through Friday.
Monday, 23 March
The week opened with the plenary session, followed by LEWG.
I had prepared personal letters for several members of the committee whose work and leadership I admire. On Monday I hand-delivered letters to Bjarne Stroustrup, Guy Davidson, Nina Ranns, Andrzej Krzemieński, and Jeff Garland. Dietmar Kühl received his later in the week. John Spicer and Ville Voutilainen were unable to attend Croydon, so their letters will find them another way.
LEWG reviewed P3373, P3425/P3986, and P3669 during the day. In the evening, the P3962R0 session on implementation reality drew a lively crowd.
I had lunch at Tai Tung, a Chinese restaurant near the hotel that became our regular spot throughout the week. Michael Wong made introductions and I had my first real conversation with Bjarne Stroustrup - about Profiles, the direction of safety in C++, and the work ahead. A great start.
Tuesday, 24 March
LEWG morning: P4031 and P3940.
The afternoon block covered papers I have been deeply involved with: P4007R0 (Senders and Coroutines), P3552R3, and P2583R3 (Symmetric Transfer and Sender Composition). Ian Petersen, joining on Zoom, provided an honest and constructive assessment of aspects of P3552R3. The discussion was substantive and collegial throughout.
I did not vote against anything this week. On nearly every vote I cast neutral. I did not block std::execution::task. I came to Croydon to listen, learn, and collaborate.
In the evening, Guy Davidson sponsored a session on P3874R1 - memory-safe language design. The discussion was thoughtful and engaged.
Wednesday, 25 March
The busiest day of the week.
LEWG in the morning, then the C++26 committee photo before lunch in the main conference room. This is the photo that will accompany the standard. It was a good moment to be part of.
I attended SG23 in the late morning. The room was packed. I offered my seat to someone who turned out to be Peter Bindels - I did not recognize him at the time. A small friendly moment that led to a productive collaboration later in the week.
Wednesday afternoon I presented in SG18. The session went well - positive reception and engaged discussion.
LEWG’s C++29 afternoon block looked ahead to what the next standard should prioritize. In the evening, Jon Bauman led a session asking “What is a memory-safe language?” - a question the committee is working through carefully.
Thursday, 26 March
LEWG all day. Peter Bindels presented P3655 (cstring_view) in the morning. During the session, I updated my Escape Hatches paper (P4035R0) in real time to support his work. Co-author Marco Foco responded with P3566, additional material to incorporate. This kind of real-time cross-proposal collaboration - updating papers during presentations to strengthen each other’s work - is exactly what the committee should be doing more of.
I also continued supporting Profiles through the PAVE paper (P4137R0), which proposes evidence methodology for measuring profile coverage.
Nina Ranns brought me into LEWGI, where I presented P4133R0 (”What Every Proposal Must Contain”) as a slideshow. Michael Wong was also in the room. The presentation stimulated over an hour of robust discussion on proposal evaluation criteria and mitigations. The room showed genuine interest in the evaluation model, and there was meaningful movement toward supporting AI-assisted workflows in the paper process - a direction that could benefit the entire committee’s productivity.
Friday, 27 March
A productive morning meeting with Bjarne Stroustrup, continuing our Monday conversation about Profiles and how the PAVE paper (P4137R0) can serve as supporting evidence. Bjarne sees value in the approach, and I am looking forward to continuing this collaboration.
Dietmar Kühl and I had several substantive conversations throughout the week on allocator design, sender/receiver topics, and the broader async landscape. I appreciate his technical depth and generosity with his time.
I met with Ian Sandoe, a GCC implementor whose work is directly relevant to P4126. Mungo also met with him separately.
Vietnamese lunch with Roger Orr and Harry. LEWG continued through the day, with the motions deadline at 8 PM.
One note from the week: a Code of Conduct concern was raised in a public forum. I addressed it directly and personally, and the matter was resolved through a private conversation. We agreed to work together going forward.
The evening was beer and pizza with Matthias Wippich, Jan Schultke, and Matheus Izvekov. No politics. Just unwinding after a long, productive week.
Saturday, 28 March - Closing
Saturday morning I developed P4162 (”To boldly suggest an overall plan for C++29”) and emailed it to Bjarne and Nina. I also emailed Bjarne D4163 (”What Civilizations Remember”), which Harry and I wrote on Friday. Bjarne replied by email: “Very good.”
Closing plenary at 8:30 AM. The national body votes on C++26 took place today. After years of work by hundreds of people, C++26 is moving to international ballot.
During the closing plenary, Nevin Liber (U.S. National Body representative) announced that the paper system is being improved to allow papers to be marked “information only.”
Continued discussions with Ian Sandoe before heading out.
Then it was time to go home.
The Network Endeavor
P4003R0 (”Coroutines for I/O”) is published and targeting its first LEWG review at Brno in June 2026. The full Network Endeavor informational paper series will appear in the April mailing. Of those, only P4003R0 requests floor time. The rest are informational reference material that the committee can consult at their own pace.
I had productive discussions with WG21 leadership about modernizing committee infrastructure and tooling - an area where C++ Alliance is well positioned to contribute.
Reflections
This trip exceeded every expectation. The committee is full of people who care deeply about getting C++ right. I enjoyed the technical discussions, the meals, the hallway conversations, and the late-evening sessions. I came in hoping to be useful. I left feeling like I belong.
I love these meetings. I am already looking forward to Brno.