RFC: RFC process redux

[brson]

This updates #2 to incorporate feedback. I've stuck with sequential numbers and added an unresolved question about what to do with rejected RFCs. Also add three pieces of metadata to the template (start date, RFC PR #, Rust Issue #).

[steveklabnik]

Why open a new PR instead of just updating the text there?

[chris-morgan]

@steveklabnik because GitHub doesn't permit that. (GitHub's pull requests are not really collaborative.)

[steveklabnik]

All you have to do is push to the branch, and it updates the PR. That includes, for example, you sending a PR to @brson 's branch with updated text.

Anyway, this is very off-topic, sorry.

[bstrie]

Shouldn't this be 2014/03/11?

[brson]

Yes.

[bstrie]

On the one hand I think we should absolutely retain rejected RFCs in the archive, as that's what PEP does. However, I think PEP has a higher bar for inclusion in the first place, so it's possible that we could rethink this if we start to become inundated with low-quality RFCs.

[bstrie]

As far as language goes ("accepted"/"rejected"), here are the categories that PEP uses:

• Meta
• Informational
• Accepted (not yet implemented)
• Open (under consideration)
• Finished (accepted and implemented)
• Historical (Meta and Informational PEPs that have become obsolete)
• Deferred
• Abandoned, Withdrawn, and Rejected

[brson]

Do we need to change the terminology for RFC states and decide on whether to keep old RFCs before accepting this?

[chris-morgan]

I have an extremely strong preference for YYYY-MM-DD rather than YYYY/MM/DD (or YYY/MM/DD!). Let's go with the standardised date formats.

[bstrie]

Seconded. ISO 8601 uses dashes, let's not diverge.

[alexcrichton]

I think we should accept this, we can retroactively fill in old RFCs and change terminology.

[bstrie]

Agreed, this is really sort of a meta-RFC since it doesn't serve as a specification for implementing anything (and as such we can freely modify it after-the-fact).

I still vote for making this RFC #0!

[pnkfelix]

I thought having the template be RFC #0 (or 0000 I suppose) made some amount of sense.

@bstrie could you be convinced to change your vote to make this RFC #1 instead?

[huonw]

Do we need an RFC for choosing the number of this RFC RFC? ;P

[bstrie]

@pnkfelix Good point, I abstain. :P