Re: Tracking errata for ActivityPub, Activity Streams core and Activity Vocabulary

út 14. 4. 2026 v 21:24 odesílatel Darius Kazemi <darius.kazemi@gmail.com>
napsal:

> I agree, Melvin, that we need to explicitly scope any fast tracking, but
> critically Evan has listed two things here:
>
>  - spelling errors
>  - "syntax errors in examples", which are indeed non-normative and
> distinct from syntax errors in specification language.
>
> Both of these are definitionally class 2 changes per W3C process:
> https://www.w3.org/policies/process/#class-2
>
> So Evan is proposing something that is already scoped for us by the W3C,
> and imo entirely appropriate.
>
> Of course even class 2 changes will and should be visible for review, and
> if anyone notices a class 3 or 4 change slipping in under the guise of a
> class 2 change I encourage them to raise a flag to the group! But I expect
> that to be rare if we stick to strict W3C change class definitions.
>

Sounds good. One note from Class 2:

“If there is any doubt or disagreement as to whether a change functionally
affects interpretation, that change does not fall into this class.”

In practice, disagreements on classification do arise, so it may be worth
being explicit that such cases are escalated.


>
> -Darius
>
>
> On Tue, Apr 14, 2026, 1:57 PM Melvin Carvalho <melvincarvalho@gmail.com>
> wrote:
>
>>
>>
>> pá 3. 4. 2026 v 18:52 odesílatel Evan Prodromou <
>> evanp@socialwebfoundation.org> napsal:
>>
>>> For the last few years, the CG has tracked incoming issues for the
>>> ActivityPub and Activity Streams 2.0 specifications, and when errors are
>>> noted, we have maintained a list of errata.
>>>
>>> During our joint discussion today, we proposed adopting the versions of
>>> these docs with errata corrections applied as the basis of the next version
>>> of the documents.
>>>
>>> It seems to me that this is a good point to stop tracking errata for
>>> these documents, and start making those changes directly to the new drafts.
>>> For example, spelling errors or syntax errors in the examples.
>>>
>>
>> One concern is defining “spelling errors or syntax errors” as a class of
>> changes. In practice, once a fast-track category exists, there can be
>> disagreement over whether a change fits that class.
>>
>> At W3C this often becomes a classification question, where proponents
>> view changes as minor but reviewers may not. When in doubt, these typically
>> need to be treated as the higher class.
>>
>> It may help to define a lightweight mechanism for resolving
>> classification disputes, or to explicitly scope this path to clearly
>> non-normative changes only.
>>
>>
>>>
>>> Does this reflect everyone else's understanding of the state of the
>>> documents? Is there a reason I don't see clearly to keep tracking errata
>>> separately?
>>>
>>> In this case, I'd like to refocus the work on issue triage to triage of
>>> proposed changes to the drafts of the WG, rather than as a pipeline to
>>> errata updates.
>>>
>>> Evan
>>>
>>