PEP 1: Update main PEP discussion venue from Python-Dev to Discourse

[CAM-Gerlach]

The Steering Council recently changed the canonical PEP announcement venue from Python-Dev to Discourse. Therefore, we should update PEP 1 to reflect this.

[warsaw]

FWIW, I don't read Petr's email as a pronouncement, but just that the SC would like to move official discussions to Discourse.

The Steering Council would like to switch from python-dev to
discuss.python.org.

and

However, we would like to know if this will pose an undue burden to
anyone, if there are workflows or usage problems that we are not aware
of. As mentioned, this is something the Steering Council thinks is a
good idea, but we want to make sure we're aware of all the impact when
we make the final decision.

So before this is merged, I would like to see that "final decision" be pronounced by the SC.

[CAM-Gerlach]

Great point, thanks @warsaw — I marked this as draft to block it from being merged until the SC does so (and at least one SC member approves this PR — I'd tagged @encukou since he was the one who announced it originally, and is the most involved here).

[encukou]

The blocker is updating docs. I filed python/devguide#944 for it.
Please keep this open until the switch :)

[gvanrossum]

I'm confused. I thought the blocker was the SC actually deciding. But now you (@encukou) claim that the (only?) blocker is updating the docs, so IIUC the SC has already decided. If that's the case, why isn't PEP 1 the "source of truth"? Why does PEP 1 have to wait for the devguide to be updated?

[AA-Turner]

• I think we should be careful about saying Discourse in general versus the PEPs category specifically, especially if we direct people to subscribe to only a handful of categories on Discourse.

• Other changes are cutting for length

A

[AA-Turner]

PEP 1 is very long, I'd suggest cutting the second part of this sentence for brevity

Suggested change

	is a good choice for most new PEPs,
	whereas historically the `Python-Dev`_ mailing list was commonly used.
	is a good choice for most new PEPs.

[CAM-Gerlach]

I certainly agree that PEP 1 is rather long, and could do with being more shorter and focused at the macro level (and have some much more impactful ideas for this). That said, this is the micro level of less than a dozen words out of 6500; whether we cut out a relatively concise mention is more a question of whether it serves a purpose consummate with its fairly brief length.

In this case, I retained this line, the only reference left to Python-Dev in the PEP, since I felt it provided important context for historical PEPs, a large majority of which have their discussions and resolutions on that venue, many without one or both being explicitly mentioned/linked in the PEP in question.

However, if others feel that doesn't justify even the relatively brief mention, I can elide it as you suggest.

[warsaw]

I'm confused. I thought the blocker was the SC actually deciding. But now you (@encukou) claim that the (only?) blocker is updating the docs, so IIUC the SC has already decided. If that's the case, why isn't PEP 1 the "source of truth"? Why does PEP 1 have to wait for the devguide to be updated?

I agree with @gvanrossum. The devguide will get updated, but I think the SC needs to make a public pronouncement before PEP 1 can be changed.

[encukou]

The SC decided long ago, but started a thread on the list to find any unknown blockers. It turns out that there are issues, but no real blockers.
The SC didn't set a time when it'll stop gathering the feedback, but it's been long enough now. The only blocker is actually making the changes.

I think the SC should make the pronouncement as the docs changes (PEP 1 and devguide) are merged. Just like with PEPs, it's not “final” until the implementation is merged.
Not that the order of events is terribly important. And there's no rush to do this change.

(I'm writing as a SC member, not as the SC as a whole.)

[CAM-Gerlach]

To note, the primary devguide sections discussing Python and core dev communication channels were already updated in python/devguide#938 . I've opened python/devguide#945 to update the remaining sections.

If that's the case, why isn't PEP 1 the "source of truth"? Why does PEP 1 have to wait for the devguide to be updated?

I suppose it makes sense in ensuring the documentation in sync with the specification, but I can (and did) just update the devguide too (and its already partially updated). There were also at least a couple sections that in large part duplicate PEP 1's description of parts of the PEP process (and each other), which had (predictably) drifted out of date with current PEP 1 guidance; I simplified those to minimize the duplication and focus on the unique elements for the specific scenarios at hand, and refer to PEP 1 for the rest.

[encukou]

(speaking as a single SC member, not as the whole)

[CAM-Gerlach]

Updated to fully address @AA-Turner and @encukou 's feedback.

@encukou I'm assuming you'll want to add this to the SC agenda in concert with python/devguide#945 ?

[encukou]

Already there – the agenda lists topics rather than individual PRs.

[encukou]

The switch was announced, so I'll merge.