Re: Review of Process 2020

On 12/3/19 3:36 AM, Carine Bournez wrote:
> On Mon, Dec 02, 2019 at 04:38:22PM -0500, fantasai wrote:
>>
>> Florian did mention "Ever-extending" as a possibility (referencing
>> https://en.wikipedia.org/wiki/XMDF_(E-book_format)) ), if you're looking for
>> Ever-something. :)
> 
> We won't really always only extend, I suppose (the process already had
> whatever was needed to do that, in fact).

It did not. To add features to a specification, it was required to issue a 
FPWD. See https://www.w3.org/2019/Process-20190301/#revised-rec

“To make changes which introduce a new feature or features, W3C must follow 
the full process of advancing a technical report to Recommendation beginning 
with a new First Public Working Draft.”

Wrt dropping things, at least for Web technology, we're generally unable to 
remove features due to Web-compat.

On a different, but related, point, removing a feature from a Recommendation 
arguably rescinds it (which would affect licensing obligations), so this 
should only be done in very deliberate circumstances and not as a matter of 
general maintenance.

>> The Snapshot has Director's approval and CR-type patent implications
>> (because it triggers an exclusion opportunity), and the Update does not (is
>> equivalent to a WD).
> 
> I understood the distinction, but essentially by reading the other
> documents. The Snapshot is essentially what a CR is currently, so
> I don't see the need for a new term. The 'editorial CRs' and 'CR update|draft'
> are more confusing terms.
> 
>> There's (at least) two reasons for this:
>> 1. The lawyers working on the WHATWG policy concluded that having snapshots
>> was what they were willing to work with, so we're adopting the same PP model
>> 2. This is a close model to the process we have today: CR does not change,
>> other than to change its name to "CR Snapshot", and the ED gets published on
>> /TR as the "CR Update".
> 
> IMHO, that's our current model. The addition of non-CfE-triggering CRs
> was already done a few years ago, even if the process was clarified
> after the facts. >
> The real addition of Process2020 is that some CRs with substantial changes
> will be non-CfE-triggering. "Draft" seemed a good word to describe
> the unfinished state of those. "CR Update" is way too loose, it seems
> it could apply to anything published as CR (CfE-triggering, editorial or
> draft).

Happy to switch it to CR Draft. Done.

>> As for working under the assumption of an updated PP: we don't have time to
>> wait for PSIG to adopt a new PP and then start working on Process edits to
>> match. We have to work in parallel. If for some reason we can't update the
>> PP, it's easy to remove the section. It's less easy to add it later when
>> we're under a time crunch.
> 
> Maybe I'm just too pessimistic on the possibility to get the PP changed
> in a timely manner, but I doubt it will be changed before this process
> version.

Sure, and in that case we can remove the section. :)

~fantasai
who is trying to be prepared for optimism as well as pessimism

Received on Thursday, 5 December 2019 04:55:03 UTC