In addition to the Rust statement, I would like to explicitly apologize and take responsibility for my part in this. We need to be transparent about how things operate, both as an essential step to improving how we operate, and as an essential part of being accountable and responsible.

I apologize for my own role in what led to the removal of a RustConf keynote speaker, at great harm to the speaker, the conference, and Rust.

The below is a full account of my own involvement in this and all the details I'm aware of. (I am not speaking for anyone else.) That includes mistakes and harm I'm personally responsible for that I'm aware of, followed by the steps I'm personally taking to avoid making such mistakes and prevent such harm in the future. I'm speaking for myself as an individual here; this is separate from any steps that groups or other individuals may take to avoid mistakes and prevent harm in the future.

I make no excuses for the result that I contributed to here. Any details given by way of explanation or context do not excuse the outcome.

-----

Back in December, Sage on the Foundation side reached out about a proposed project from Shepherd's Oasis involving compile-time introspection, doing due diligence with the team(s) it'd most likely intersect with. In parallel, the proposer of that project (JeanHeyd) reached out to Niko and to me in a joint DM. I talked to him at length about the project, expressed encouragement about exploring possible approaches for the underlying problem, and discussed possible approaches and solutions with him. I subsequently reported back to Foundation folks that we didn't see any issue with the project going forward, and that they should evaluate it as they saw fit.

(I didn't realize the connection until after all of this happened between this project proposal from December and the keynote. But I want to be thorough about all aspects of this I was involved in.)

In February, Sage, on behalf of RustConf, talked to me (as a project member and program committee member) about RustConf wanting the project to propose a couple of keynotes. (RustConf runs the program but has always asked the project to propose keynote speakers.) They wanted that to happen ideally by late April or May, concurrently with selecting talks, so that the whole schedule could be announced at the same time. I brought that to the interim "leadership chat" (created in 2021); at the time, we hoped that the Council might possibly be up and running in time to decide on that.

Starting on April 18 (while I was traveling), after hearing that the RustConf schedule was likely to get announced in May, and given that the Council wasn't going to be up and running by then, I brought up the keynote selection again on leadership chat. Various people, myself included, put forward suggestions for candidates. (As with most things on leadership chat, there was no defined process here.) Some things occurred in parallel here, such as people who knew some of the folks on that list reaching out to them directly to find out who would be willing to speak. Over the course of discussion, we came up with 7 names, including JeanHeyd, for possible keynote speakers.

Starting on May 5, I noted that nobody had made any concrete proposals for selecting a pair of keynote speakers. (There was no defined process here either.) Someone proposed a specific pair of keynote speakers, including JeanHeyd. I supported that proposal myself, as did a few others: a total of five people (out of 18 in leadership chat) ended up responding to the proposed pair of speakers. One person proposed a different pair of speakers. Nobody raised any issues. Two people, including myself, conveyed the two proposed keynote speakers to Sage.

For my part, I didn't personally have any direct experience with or understanding of JeanHeyd's ongoing work to base any evaluation on; however, I did know he was an editor of the C standard, which already seemed to me like it would make for any number of amazing keynote possibilities, as RustConf has sometimes had one keynote from someone with a substantially different perspective.

Based on the information I have, I thought, and still think, he would have make a good keynote speaker, with a variety of possible topics to speak on.

Later (May 10) a different conference speaker cancelled, leading to us needing a closing keynote as well to fill the gap. This again took place via an ad-hoc discussion on leadership chat. A total of three people participated in that part of the discussion, before someone mentioned the proposed candidate to Sage less than a day later, then two more people added support *after* that suggestion was taken to Sage.

On May 18, I received several complaints from a few Rust project members, about various aspects of the compile-time reflection project and the associated blog post that had recently come out, and about the RustConf keynote selection. (I had previously skimmed the blog post and had no context for this; this was the first I heard anything like it.) The only portion of this that I personally chimed in on was to agree that the compile-time reflection work, specifically, would probably not make a great keynote; not for any reasons of its quality, but solely because of its experimental nature. (I had had the assumption that any number of other possible topics of JeanHeyd's considerable expertise would be the keynote topic.)

Someone asked about how the process for keynote selection worked, which I described. At least one person asked whether there was anything that could be done to change the selection at this point.

At this point, having heard a set of emphatic complaints which I had no context to evaluate myself, I stated that I didn't know whether the schedule or keynotes had been announced yet, and promptly posted an initial message to leadership chat mentioning these complaints. A couple of people on leadership chat chimed in to this discussion, expressing some negative sentiments; one person mentioned the idea that the topic would make a great invited talk rather than a keynote. (Further comments and discussion on this, including substantially more positive responses and disagreement about taking any actions, happened after I reached out to Sage.)

I and another person (separately) reached out to Sage shortly afterwards. I asked Sage if keynotes had been announced yet, attempted to provide a heads-up about the complaints, asked if they could hold off, and conveyed that some people on the project side were expressing concerns. This was one of many mistakes I made.

In this discussion on leadership chat, as with many others, we didn't follow any process. No consensus emerged, and no decisions were actually reached. In addition, I treated this conversation as rushed (based on perceived time-sensitivity). Other people followed up later on in the discussion on leadership chat, expressing various different points of view.

In subsequent conversation with Sage, I provided details from the complaints I had received from a few project members, and (compounding my mistakes here) discussed "options". Sage expressed, and I agreed, that the invitation to speak at RustConf must not be withdrawn. (People expressed the same sentiment in leadership chat.) I raised the possibility of the topic being a talk, rather than a keynote. This was again a mistake, and I was thoughtless to not consider that that was still incredibly hurtful.

-----

To Sage, I did a poor job of communicating the nature and context of who was raising concerns, how widespread those concerns were, the degree of consensus, and **most importantly**, that this was a set of concerns from one team rather than any kind of consensus or decision by the project or leadership chat. (Sage has already spoken up in detail about their perspective on how this came across.) I unintentionally portrayed the situation that, in conjunction with others, led to Sage making a bad decision (their own words) based on incomplete information. The fact that this was unintentional doesn't change the fact that my poor communication contributed to someone getting hurt and the Rust Project taking an unprofessional and inappropriate action, and I am deeply sorry for that.

I did not make it at all clear that I wasn't serving as a conduit for leadership chat, or that leadership chat had reached no consensus and no decision. I did not make it at all clear exactly who I *was* representing, nor should I have been representing them to Sage in the first place. (I had previously been a point of contact for Sage for several other aspects of RustConf, but that's a role I should have been far more careful about.)

I failed to pull Sage directly into a conversation and then *get out* of the middle of that conversation in which I did not then have anything to offer.

I should not have been acting as a telephone (having separate smaller conversations going on with a subset of people relaying between them, rather than one group conversation). If anything, I should have pulled everyone into one group conversation, which would have avoided incomplete information.

I continued operated by the same ad-hoc partial discussion processes that leadership chat regularly operated by, and while I have contributed to developing replacements/successors with better processes (the Leadership Council), I failed to recognize the great potential for harm caused by operating in an ad-hoc manner in the interim.

I should have recognized that we were making a different kind of decision, one with a much greater potential to hurt someone, yet using similarly loose ad-hoc processes. I should have noticed that we weren't successfully capturing consensus, and personally taken steps to better improve that and control for that.

I should have put more effort into reaching more people, rather than the default of "whoever chimes in".

I should have stopped to think, rather than just being a pass-through for the complaints of others, no matter how those were conveyed to me. While I do tend to place importance on amplifying the voices of others, particularly those who seem like they haven't been heard, I also should have immediately recognized the potential for harm here, and acted accordingly.

I should have put more thought into the harm that the "options" being explored could cause, and to the extent I took part in this at all, I should have recognized that "downgrading" a keynote to a talk was just as hurtful as I recognized rescinding an invitation entirely would be.

I need to pay close attention to potential miscommunication, in multiple directions. I need to watch for cases where my understanding may not match what someone is saying, and I need to watch for cases where what someone else may be understanding is not what I'm saying.

And all of these things become more of a problem when in less formal, more ad-hoc roles with little structure or process, and thus no particular checks on actions. I say that not to excuse anything I've done or to blame it on a lack of processes to catch mistakes, only to recognize that *I should have recognized the much higher chance and cost of mistakes and taken due care accordingly*.

-----

I need to improve substantially in all of these areas in the future. But I also consider it critical to not make a mistake like this again; I cannot simply say "I'm going to do better" and think that sufficient.

Until I *have* improved substantially, I don't want to put myself in less well-specified, more ad-hoc roles, especially those that don't have well-established and well-tested mechanisms to handle consensus-building and catch potential mistakes.

For that reason, I'm taking the following steps as well:
- I've decided to leave "leadership chat". This also means I have decided to not participate in making any top-level governance decisions, whether ad-hoc or with any new processes in place.
- I'm declining the nomination to serve on the new Leadership Council.
- I will not be speaking at RustConf. (RustConf already decided and announced this.)
- I have decided not to lead the RustConf unconference I had been one of the planned staff members for.
- I've decided to step down from the co-leadership of the language team.

I apologize to JeanHeyd for taking thoughtless actions that led to him being removed as a keynote speaker, and to the Rust project and community for contributing to the culture of mistrust and lack of transparency.