Python’s Steering Council

There are several references in this PEP to the “Steering Council” or “Council”.This refers to the current members of the elected Steering Council describedinPEP 13, in their role as the final authorities on whether or not PEPswill be accepted or rejected.

Python’s Core Developers

There are several references in this PEP to “core developers”. This refers tothe currently active Python core team members described inPEP 13.

Python’s BDFL

Previous versions of this PEP used the title “BDFL-Delegate” for PEP decisionmakers. This was a historical reference to Python’s previous governance model,where all design authority ultimately derived from Guido van Rossum, theoriginal creator of the Python programming language. By contrast, the SteeringCouncil’s design authority derives from their election by the currently activecore developers. Now, PEP-Delegate is used in place of BDFL-Delegate.

PEP Editors

The PEP editors are individuals responsible for managing the administrativeand editorial aspects of the PEP workflow (e.g. assigning PEP numbers andchanging their status). SeePEP Editor Responsibilities & Workflow fordetails.

PEP editorship is by invitation of the current editors, and they can becontacted by mentioning@python/pep-editors on GitHub. All of the PEPworkflow can be conducted via the GitHubPEP repository issues and pullrequests.

Start with an idea for Python

The PEP process begins with a new idea for Python. It is highlyrecommended that a single PEP contain a single key proposal or newidea; the more focused the PEP, the more successful it tends to be.Most enhancements and bug fixes don’t need a PEP andcan be submitted directly to thePython issue tracker.The PEP editors reserve theright to reject PEP proposals if they appear too unfocused or toobroad. If in doubt, split your PEP into several well-focused ones.

Each PEP must have a champion – someone who writes the PEP using the styleand format described below, shepherds the discussions in the appropriateforums, and attempts to build community consensus around the idea. The PEPchampion (a.k.a. Author) should first attempt to ascertain whether the idea isPEP-able. Posting to theIdeas category of thePython Discourse is usuallythe best way to go about this, unless a more specialized venue is appropriate,such as theTyping category (for static typing ideas)orPackaging category (for packaging ideas) on the Python Discourse.

Vetting an idea publicly before going as far as writing a PEP is meantto save the potential author time. Many ideas have been broughtforward for changing Python that have been rejected for variousreasons. Asking the Python community first if an idea is originalhelps prevent too much time being spent on something that isguaranteed to be rejected based on prior discussions (searchingthe internet does not always do the trick). It also helps to make surethe idea is applicable to the entire community and not just the author.Just because an idea sounds good to the author does notmean it will work for most people in most areas where Python is used.

Once the champion has asked the Python community as to whether anidea has any chance of acceptance, a draft PEP should be presented tothe appropriate venue mentioned above.This gives the author a chance to flesh out the draftPEP to make properly formatted, of high quality, and to addressinitial concerns about the proposal.

Submitting a PEP

Following the above initial discussion, the workflow varies based on whetherany of the PEP’s co-authors are core developers. If one or more of the PEP’sco-authors are core developers, they are responsible for following the processoutlined below. Otherwise (i.e. none of the co-authors are core developers),then the PEP author(s) will need to find a sponsor for the PEP.

Ideally, a core developer sponsor is identified, but non-core sponsors may alsobe selected with the approval of the Steering Council. Members of the GitHub“PEP editors” team and members of the Typing Council (PEP 729) arepre-approved to be sponsors. The sponsor’s job is toprovide guidance to the PEP author to help them through the logistics of thePEP process (somewhat acting like a mentor). Being a sponsor doesnotdisqualify that person from becoming a co-author or PEP-Delegate later on (butnot both). The sponsor of a PEP is recorded in the “Sponsor:” field of theheader.

Once the sponsor or the core developer(s) co-authoring the PEP deem the PEPready for submission, the proposal should be submitted as a draft PEP via aGitHub pull request. The draft must be written in PEP style as describedbelow, else it will fail review immediately (although minor errors may becorrected by the editors).

The standard PEP workflow is:

• You, the PEP author, fork thePEP repository, and create a file namedpep-NNNN.rst that contains your new PEP.NNNN should be the nextavailable PEP number not used by a published or in-PR PEP.
• In the “PEP:” header field, enter the PEP number that matches your filenameas your draft PEP number.
• In the “Type:” header field, enter “Standards Track”,“Informational”, or “Process” as appropriate, and for the “Status:”field enter “Draft”. For full details, seePEP Header Preamble.
• Update.github/CODEOWNERS such that any co-author(s) or sponsorswith write access to thePEP repository are listed for your new file.This ensures any future pull requests changing the file will be assignedto them.
• Push this to your GitHub fork and submit a pull request.
• The PEP editors review your PR for structure, formatting, and othererrors. For a reST-formatted PEP,PEP 12 is provided as a template.It also provides a complete introduction to reST markup that is usedin PEPs. Approval criteria are:
  • It is sound and complete. The ideas must make technical sense. Theeditors do not consider whether they seem likely to be accepted.
  • The title accurately describes the content.
  • The PEP’s language (spelling, grammar, sentence structure, etc.)and code style (examples should matchPEP 7 &PEP 8) should becorrect and conformant. The PEP text will be automatically checked forcorrect reStructuredText formatting when the pull request is submitted.PEPs with invalid reST markup will not be approved.

  Editors are generally quite lenient about this initial review,expecting that problems will be corrected by the reviewing process.Note: Approval of the PEP is no guarantee that there are noembarrassing mistakes! Correctness is the responsibility of authorsand reviewers, not the editors.

  If the PEP isn’t ready for approval, an editor will send it back tothe author for revision, with specific instructions.

• Once approved, they will assign your PEP a number.

Once the review process is complete, and the PEP editors approve it (note thatthis isnot the same as accepting your PEP!), they will squash commit yourpull request onto main.

The PEP editors will not unreasonably deny publication of a PEP. Reasons fordenying PEP status include duplication of effort, being technically unsound,not providing proper motivation or addressing backwards compatibility, or notin keeping with the Python philosophy. The Steering Council can be consultedduring the approval phase, and are the final arbiter of a draft’s PEP-ability.

Developers with write access to thePEP repository may claim PEPnumbers directly by creating and committing a new PEP. When doing so, thedeveloper must handle the tasks that would normally be taken care of by thePEP editors (seePEP Editor Responsibilities & Workflow). This includesensuring the initial version meets the expected standards for submitting aPEP. Alternately, even developers should submit PEPs via pull request.When doing so, you are generally expected to handle the process yourself;if you need assistance from PEP editors, mention@python/pep-editorson GitHub.

As updates are necessary, the PEP author can check in new versions if they(or a collaborating developer) have write access to thePEP repository.Getting a PEP number assigned early can be useful for ease ofreference, especially when multiple draft PEPs are being considered at thesame time.

Standards Track PEPs consist of two parts, a design document and areference implementation. It is generally recommended that at least aprototype implementation be co-developed with the PEP, as ideas that soundgood in principle sometimes turn out to be impractical when subjected to thetest of implementation.

Discussing a PEP

As soon as a PEP number has been assignedand the draft PEP is committed to thePEP repository,a discussion thread for the PEP should be createdto provide a central place to discuss and review its contents, and thePEP should be updated so that theDiscussions-To header links to it.

The PEP authors (or sponsor, if applicable) may select any reasonable venuefor the discussion, so long as the the following criteria are met:

• The forum is appropriate to the PEP’s topic.
• The thread is publicly available on the web so that all interested partiescan participate.
• The discussion is subject to thePython Community Code of Conduct.
• A direct link to the current discussion thread is provided in the PEPunder theDiscussions-To header.

ThePEPs category of thePython Discourseis the preferred choice for most new PEPs,whereas historically thePython-Dev mailing list was commonly used.Some specialized topics have specific venues, such astheTyping category and thePackaging category on the PythonDiscourse for typing and packaging PEPs, respectively.If the PEP authors are unsure of the best venue,the PEP Sponsor and PEP editors can advise them accordingly.

If a PEP undergoes a significant re-write or other major, substantivechanges to its proposed specification, a new thread should typically be createdin the chosen venue to solicit additional feedback. If this occurs, theDiscussions-To link must be updated and a newPost-History entry addedpointing to this new thread.

If it is not chosen as the discussion venue,a brief announcement post should be made to thePEPs categorywith at least a link to the rendered PEP and theDiscussions-To threadwhen the draft PEP is committed to the repositoryand if a major-enough change is made to trigger a new thread.

PEP authors are responsible for collecting community feedback on a PEPbefore submitting it for review. However, to avoid long-winded andopen-ended discussions, strategies such as soliciting private or morenarrowly-tailored feedback in the early design phase,collaborating with other community members with expertise in the PEP’ssubject matter, and picking an appropriately-specialized discussion for thePEP’s topic (if applicable) should be considered.PEP authors should use their discretion here.

Once the PEP is assigned a number and committed to the PEP repository,substantive issues should generally be discussed on the canonical publicthread, as opposed to private channels, GitHub pull request reviews orunrelated venues. This ensures everyone can follow and contribute,avoids fragmenting the discussion,and makes sure it is fully considered as part of the PEP review process.Comments, support, concerns and other feedback on this designated threadare a critical part of what the Steering Council or PEP-Delegate willconsider when reviewing the PEP.

PEP Review & Resolution

Once the authors have completed a PEP, they may request a review forstyle and consistency from the PEP editors.However, content review and acceptance of the PEP is ultimately theresponsibility of the Steering Council, which is formally initiated byopening aSteering Council issue once the authors (and sponsor, if any)determine the PEP is ready for final review and resolution.

To expedite the process in selected cases (e.g. when a change is clearlybeneficial and ready to be accepted, but the PEP hasn’t been formally submittedfor review yet), the Steering Council may also initiate a PEP review, firstnotifying the PEP author(s) and giving them a chance to make revisions.

The final authority for PEP approval is the Steering Council. However, whenevera new PEP is put forward, any core developer who believes they are suitablyexperienced to make the final decision on that PEP may offer to serve as itsPEP-Delegate bynotifying the Steering Councilof their intent. If the Steering Council approves their offer,the PEP-Delegate will then have the authority to approve or reject that PEP.For PEPs related to the Python type system, the Typing Council (PEP 729)provides a recommendation to the Steering Council. To request such arecommendation, open an issue on theTyping Council issue tracker.

The term “PEP-Delegate” is used under the Steering Council governance modelfor the PEP’s designated decision maker,who is recorded in the “PEP-Delegate” field in the PEP’s header.The term “BDFL-Delegate” is a deprecated alias for PEP-Delegate, a legacy ofthe time when when Python was led bya BDFL.Any legacy references to “BDFL-Delegate” should be treated as equivalent to“PEP-Delegate”.

An individual offering to nominate themselves as a PEP-Delegate must notifythe relevant authors and (when present) the sponsor for the PEP, and submittheir request to the Steering Council(which can be done via anew issue ).Those taking on this responsibility are free to seekadditional guidance from the Steering Council at any time, and are also expectedto take the advice and perspectives of other core developers into account.

The Steering Council will generally approve such self-nominations by default,but may choose to decline them.Possible reasons for the Steering Council declining aself-nomination as PEP-Delegate include, but are not limited to, perceptions ofa potential conflict of interest (e.g. working for the same organisation as thePEP submitter), or simply considering another potential PEP-Delegate to bemore appropriate. If core developers (or other community members) have concernsregarding the suitability of a PEP-Delegate for any given PEP, they may askthe Steering Council to review the delegation.

If no volunteer steps forward, then the Steering Council will approach coredevelopers (and potentially other Python community members) with relevantexpertise, in an attempt to identify a candidate that is willing to serve asPEP-Delegate for that PEP. If no suitable candidate can be found, then thePEP will be marked as Deferred until one is available.

Previously appointed PEP-Delegates may choose to step down, or be asked to stepdown by the Council, in which case a new PEP-Delegate will be appointed in thesame manner as for a new PEP (including deferral of the PEP if no suitablereplacement can be found). In the event that a PEP-Delegate is asked to stepdown, this will overrule any prior acceptance or rejection of the PEP, and itwill revert to Draft status.

When such standing delegations are put in place, the Steering Council willmaintain sufficient public records to allow subsequent Councils, the coredevelopers, and the wider Python community to understand the delegations thatcurrently exist, why they were put in place, and the circumstances under whichthey may no longer be needed.

For a PEP to be accepted it must meet certain minimum criteria. Itmust be a clear and complete description of the proposed enhancement.The enhancement must represent a net improvement. The proposedimplementation, if applicable, must be solid and must not complicatethe interpreter unduly. Finally, a proposed enhancement must be“pythonic” in order to be accepted by the Steering Council. (However,“pythonic” is an imprecise term; it may be defined as whatever is acceptable tothe Steering Council. This logic is intentionally circular.) SeePEP 2for standard library module acceptance criteria.

Except where otherwise approved by the Steering Council,pronouncements of PEP resolution will be posted to thePEPs category on thePython Discourse.

Once a PEP has been accepted, the reference implementation must becompleted. When the reference implementation is complete and incorporatedinto the main source code repository, the status will be changed to “Final”.

To allow gathering of additional design and interface feedback before committingto long term stability for a language feature or standard library API, a PEPmay also be marked as “Provisional”. This is short for “Provisionally Accepted”,and indicates that the proposal has been accepted for inclusion in the referenceimplementation, but additional user feedback is needed before the full designcan be considered “Final”. Unlike regular accepted PEPs, provisionally acceptedPEPs may still be Rejected or Withdrawneven after the related changes havebeen included in a Python release.

Wherever possible, it is considered preferable to reduce the scope of a proposalto avoid the need to rely on the “Provisional” status (e.g. by deferring somefeatures to later PEPs), as this status can lead to version compatibilitychallenges in the wider Python ecosystem.PEP 411 provides additional detailson potential use cases for the Provisional status.

A PEP can also be assigned the status “Deferred”. The PEP author or aneditor can assign the PEP this status when no progress is being madeon the PEP. Once a PEP is deferred, a PEP editor can reassign itto draft status.

A PEP can also be “Rejected”. Perhaps after all is said and done itwas not a good idea. It is still important to have a record of thisfact. The “Withdrawn” status is similar - it means that the PEP authorthemselves has decided that the PEP is actually a bad idea, or hasaccepted that a competing proposal is a better alternative.

When a PEP is Accepted, Rejected or Withdrawn, the PEP should be updatedaccordingly. In addition to updating the Status field, at the very leastthe Resolution header should be added with a direct linkto the relevant post making a decision on the PEP.

PEPs can also be superseded by a different PEP, rendering the originalobsolete. This is intended for Informational PEPs, where version 2 ofan API can replace version 1.

The possible paths of the status of PEPs are as follows:

While not shown in the diagram, “Accepted” PEPs may technically move to“Rejected” or “Withdrawn” even after acceptance. This will only occur ifthe implementation process reveals fundamental flaws in the design that werenot noticed prior to acceptance of the PEP. Unlike Provisional PEPs, thesetransitions are only permitted if the accepted proposal hasnot been includedin a Python release - released changes must instead go through the regulardeprecation process (which may require a new PEP providing the rationale forthe deprecation).

Some Informational and Process PEPs may also have a status of “Active”if they are never meant to be completed. E.g.PEP 1 (this PEP).

PEP Maintenance

In general, PEPs are no longer substantially modified after they have reachedthe Accepted, Final, Rejected or Superseded state. Once resolution is reached,a PEP is considered a historical document rather than a living specification.Formal documentation of the expected behavior should be maintained elsewhere,such as theLanguage Reference for core features, theLibrary Referencefor standard library modules or thePyPA Specifications for packaging.

If changes based on implementation experience and user feedback are made toStandards track PEPs while in the Provisional or (with SC approval) Acceptedstate, they should be noted in the PEP, such that the PEP accurately describesthe implementation at the point where it is marked Final.

Active (Informational and Process) PEPs may be updated over time to reflectchanges to development practices and other details. The precise processfollowed in these cases will depend on the nature and purpose of the PEPin question.

Occasionally, a Deferred or even a Withdrawn PEP may be resurrectedwith major updates, but it is often better to just propose a new one.