CEP 8 - Conda Release Schedule
| Title | Conda Release Schedule |
| Status | Accepted |
| Author(s) | Ken Odegard <kodegard@anaconda.com>, Jannis Leidel <jleidel@anaconda.com> |
| Created | May 20, 2022 |
| Updated | August 22, 2022 |
| Discussion | https://github.com/conda-incubator/ceps/pull/26 |
| Implementation | NA |
Abstract
This CEP describes a release cadence for all future conda versions starting with 22.9.0.
Specification
We propose regularly scheduled bi-monthly (every two months) releases where the tagging/release occurs during the week (per ISO 8601, Monday through Thursday) of the second Monday of the month.
In a nod to our many different kinds of users, we will also propose a deprecation policy (to be defined in a later CEP) that allows for a slower adoption rate (i.e., users could update every 3-4 months instead).
To accomplish better, more predictable versioning, we will adopt CalVer:
YY: the major version will be the shortened year (22+)MM: the minor version will be the shortened month (1-12)MICRO: the micro/patch version is reset to zero every month and incremented for every additional release that occurs during that month (0+)
This scheme will start with the September 2022 release of 22.9.0, resulting in the following regular release schedule:
| Version | Release Week |
|---|---|
| 22.9.0 | 2022-09-12 |
| 22.11.0 | 2022-11-14 |
| 23.1.0 | 2023-01-09 |
| 23.3.0 | 2023-03-13 |
| 23.5.0 | 2023-05-08 |
| 23.7.0 | 2023-07-10 |
| 23.9.0 | 2023-09-11 |
| ... | ... |
Note Despite following a bimonthly release schedule, we will permit releases to occur at any time between regular releases for hotfixes, security releases, or other high-priority changes that require immediate release.
To distinguish between the bi-monthly release schedule and other optional releases, we define the following release types:
- Regular release: the regularly scheduled bi-monthly release
- Optional release: releases that may occur on alternating months from regular releases
- Hotfix release: extra releases that may occur during any month to patch an earlier release of the same month (whether a regular or optional release)
So, it's entirely feasible to see the following releases:
| Version | Release Type |
|---|---|
| 22.9.0 | regular |
| 22.9.1 | hotfix |
| 22.9.2 | hotfix |
| 22.10.0 | optional |
| 22.10.1 | hotfix |
| 22.11.0 | regular |
| 22.12.0 | optional |
| 22.12.1 | hotfix |
| ... | ... |
Note In case no or few significant changes have been made since the last release, the conda maintainer team may decide to skip a regular release, as long as the decision is made public and documented in the changelog.
Motivation
Our goal with this CEP is to remove ambiguity/maintainer guesswork of when and what warrants a release.
Backwards Compatibility
Adopting this release schedule does not break any existing processes or schemes.
We will eliminate confusion regarding years of rumored changes slated for a conda 5.0 release since we will skip from conda 4.14.0 to conda 22.9.0. Avoiding conda 5.0 is not to say that these changes won't occur, but instead that it's unrealistic for these changes to be successfully (and safely) rolled out in a single release.
Finally, since SemVer is semantically interchangeable with CalVer and we propose switching from conda 4.14.0 to conda 22.9.0, both version parsing and version ordering will be unaffected.
Alternatives
- Do nothing. Continue with ad hoc releases.
- Rejected for being too inconsistent and challenging for roadmap/planning purposes.
- Follow a monthly schedule.
- Rejected for being too fast for maintainers and multi-user installs.
- Follow a quarterly schedule.
- Rejected for being too slow for maintainers and the average user who wants to see improvements.
Resolution
This was a standard, non-timed-out vote.
- Among Steering Council members there are 10 "yes", 0 "no", and no abstentions.
- No Emeritus Steering member voted.
This vote has reached quorum (10 + 0 = 10 which is at least 60% of 16).
It has also passed since it recorded 10 "yes" votes and 0 "no" votes giving 10/10 which is greater than 60% of 15.
It should be noted that a request for change was recorded in the pull request about minor implementation details that do not invalidate the previous votes. The author made the requested change.
Reference
Copyright
All CEPs are explicitly CC0 1.0 Universal.