@evan @julian @bengo As chair of the new working group: I want as much work as possible done in the (community oriented, open) CG. I want the CG to bring proposals to the WG and I want the WG to reach out to the CG when we have needs.

The WG exists because W3C policy states that a CG cannot publish normative W3C specifications, only a (Members-only) WG can. I'm going to do the best I can to make the WG run as openly as possible within the framework handed to me.

@darius@friend.camp that makes sense, thanks! I was under the impression that the CG would be dissolved in favor of the WG. It doesn't seem like this is the case.

@julian @darius That's not the case!

@dmitri do you have the link for the meeting notes from last Friday? They'd be really helpful for this conversation.

@julian @darius @dmitri I think @bengo has brought a lot of healthy skepticism to the CG and WG chartering efforts over the last few years. I'm glad he's worked so hard on this, and I hope he continues to make sure that processes stay open and fair.

@darius @evan @julian

Darius, that's only partially true. Per policy, CG cannot publish normative specifications. It's NOT true that "only a WG can". S6.2.6 explicitly describes how AP can be updated without a WG. It requires W3C staff help, and requires backwards compatibility. Clarifying and improving the spec has always been possible without a WG, but not if W3C staff obstructs it, and not when insiders are determined to publish breaking changes despite lack of consensus in CG.

@darius @evan @julian I made the point in SWICG meeting years ago. W3C staff seemed to agree. https://github.com/w3c/strategy/issues/435#issuecomment-2065436813

There's no reason to keep repeating the half truth because it sounds true, but it's not, so it has clearly caused confusion. If that's the only reason we have a WG, we don't need one. There are other reasons.

There has not been CG consensus on a request to republish AP, and this is a way of venue shopping to a much smaller consensus group.

@evan @julian @darius @dmitri This is not only about fairness and openness, but even moreso about backwards compatibility and not changing the conformance classes in a way that will effectively 'fork' ActivityPub. Which is why I emailed you Jan 16, 2024 with concerns about the changes to normative references in your draft (no reply). A WG is ONLY needed to make non normative changes affecting conformance classes like you've authored into your draft.

@evan @julian @darius @dmitri IMHO what's best for fellow implementers and end-users, and most fair to those who took risk to implement the specification as agreed upon, is:
WE DO NOT BREAK USERSPACE[0]

Linus' diction in that footnote is not great, but the principle is. It stands up for all implementers, not just rich/big ones. That's leadership in protocol development. Instead some of our leaders have not only been breaking userspace, but doing it for hire.

[0]: https://lore.kernel.org/all/CA+55aFy98A+LJK4+GWMcbzaa1zsPBRo76q+ioEjbx-uaMKH6Uw@mail.gmail.com/

@bengo @evan @julian The idea is not just to make class 1 and 2 changes per 6.2.6. The idea is to make class 3 and 4 changes as well. That was in scope of what we discussed during the many meetings about the WG charter.

And I think the fact of the charter getting approved by the CG represents consensus? I wouldn't have agreed to be chair of a group I felt was illegitimate. (I know the consensus does not reflect unanimous consensus. I'm okay with that.)

@darius @evan @julian

> The idea is to make class 3 and 4 changes as well.

It's a bad and unfair idea, is what i"m saying. Totally respect your position if you disagree.

> And I think the fact of the charter getting approved by the CG represents consensus?

Another half truth. CG consensus is entirely determined by the CG chair. There could be a vast majority against something, and if the CG Chair says there is consensus, there is. 'consensus' is very malleable due to this.

@darius @evan @julian The CG decision policy, ie the group which ostensibly decided to approve a charter, *requires* the chair be elected. The CG Chair has not been elected EVER. And yet we are talking about what the CG has decided by consensus as determined by a completely different policy than the CG charter's decision policy requires. It's so clear an outcome was decided and all process that made that inconvenient is ignored, so I just can't let this misinfo spread that process requires WG.

@darius @evan @julian And I'm extremely disappointed in w3c staff for completely disregarding and failing to enforce the charter, and moreso in leadership in our community for failing to even have a decision policy, agreeing to one, and then ignoring its requirements when inconvenient.

@bengo @evan @julian I do not agree with your analysis of the situation post, say, January 2025 when I started to get involved in things. Prior to that (esp regarding CG chair selection) I was not involved and can't make any claims. My goal here is to get the standard to a place where its current shortcomings are addressed, and it is more widely implemented. (I'm trying to stay neutral on what those shortcomings are. I want to the CG to figure that out and make proposals alongside the WG.)

@bengo @julian @darius @dmitri That doesn't ring a bell. Checking my email archives, I see your announcement of activitypub-testing on 15 Jan and you declined the AP issue triage on 17 Jan.

@bengo @julian @darius @dmitri I don't think private email is the right way to work on ActivityPub or Activity Streams, anyways. If you think something went wrong with one of the errata or the editors' draft, you should definitely open an issue on GitHub or post to the public-swicg mailing list. You've been active in those discussions, so that's probably a great way to talk about it.

@bengo @julian @darius @dmitri I also think backwards compatibility is important. It would be catastrophic to announce an incompatible next version that doesn't work with the existing millions of users and tens of thousands of servers. If we add new features, like LOLA, they'd have to be compatible with the network as it is now.

@evan @julian @darius @dmitri this isn't the first time you've claimed to not receive an email I sent from gmail and got no undeliverability notice for. You should fix your self hosted mail setup.

Beyond that, I'm finding it hard to reconcile your public position that private email is not the right way to work on these things when you have sent me so many private emails e.g. about how I should sign an NDA with Facebook to come work with you and them on ActivityPub and ActivityStreams.

@bengo I regret losing your email. It was clearly important. I'm sorry.

@bhaugen update: this is the best place for relevant information and to provide feedback. https://lists.w3.org/Archives/Public/public-swicg/2026Jan/0016.html