I like ;)

On Mon, Aug 3, 2026 at 11:37 AM Wido den Hollander via dev
<[email protected]> wrote:
>
> Hello,
>
> Over the years there have been many proposals regarding the versioning
> of CloudStack. The biggest hurdle has been the "4" as a prefix, which is
> useless at the moment.
>
> I've looked up many of the proposals and discussions (see below), but
> none of them went anywhere, they silently died. I want this to change.
>
> Ubuntu and Apple us a versioning system where they prefix with the year,
> in 2026 this is "26" and next year it will be "27".

I don't think we need to follow years, I would support it, but the
basic number would be good enough to me
>
> Proposal in the past where to drop the "4" and that would be 20 in the
> 4.20 era (2024!!) and this would now be "24" as we are approaching
> version 4.24.
>
> I think this would be the easiest change by making our next release "25"
> and not 4.25 anymore, with one note: 25 is very close to (20)27 and it
> might be better to jump from 25 to 27 as our next release will be in
> 2027 after we release 4.24.

Here lies my only concern. This will restrict us to at most one major
per year. I like our custom of late where we have a non-LTS in spring
and an LTS in autumn (around CCC)
If we address this . (i.e. allow for experimental features in some way
that will not harm more conservative users) I am 100% on board.

> We can go over many details, but in general I'd like to get consensus
> and get things moving on this topic.
>
> Any major objections before I start a VOTE? The VOTE would be that there
> will be no 4.25 in 2027, but we will have version "27".
>
> Thanks,
>
> Wido
>
>
>
> <2024
> - Jun 2016 - [DISCUSS] 5.0.0 and 6.0.0 (John Burwell): 5.0 for cruft
>    removal/breaking refactors, 6.0 for architectural redesign.
>    https://lists.apache.org/thread/lcmvvyy098oo07rxzf8gzk56oz1lns95
> - Jan 2019 - Why CloudStack 5 (Ivan Kudryavtsev): no radical change to
>    justify a "5"; if it is marketing, just drop the leading "4.".
>    https://lists.apache.org/thread/lwlxs8xgz4glocctf7dv89k5nqqsxmlb
>
> 2024
> - Jan 2024 - [PROPOSAL] version naming : drop the 4. (Daan): the second
>    digit is de facto the major. Variants raised: 5.0, 20.0, and
>    Ubuntu-style YYYY.MM.
>    https://lists.apache.org/thread/lh45w55c3jmhm7w2w0xgdvlw78pd4p87
> - Jan 2024 - [VOTE] drop first version number and continue with semantic
>    versioning: closed same day, no conclusion, more discussion requested.
>    https://lists.apache.org/thread/59m575f9vcvl8gdj9c9v5336gmj3v330
> - Feb 2024 - [VOTE] next version 20 instead of 4.20: several +1, but
>    blocked. Paul Angus -1 (vote conflated dropping the 4 with adopting
>    semver; digit semantics undefined). Guto -1 from the other side (only
>    worthwhile if it comes with a breaking-change mechanism). No result.
>    https://lists.apache.org/thread/4zs8d15ghvvwwro46ry5zjf8fn8x0t88
> - Mar 2024 - Joao's alternative: keep X.Y.Z.N with a fixed cadence
>    (major/2y, minor/6m, patch/2-3m, N for security).
>    https://lists.apache.org/thread/o6o9h3qp8gqrpq4v7o81tl6vp51tkjhg
>    https://github.com/apache/cloudstack/discussions/8970
> - Dec 2024 - CloudStack 5? (Rene Moser): reopened the same question.
>    https://lists.apache.org/thread/hnzp6hnsjyj8593cf6tbgryt1s8z5glq
>
> 2025
> - Apr 2025 - [Discussion] Versioning (Joao): three rules proposed -
>    API-breaking changes, DB schema changes and feature removal only in
>    major versions; naming to be voted separately.
>    https://lists.apache.org/thread/4jk31krsjl8cbp5n8wbt7ypwl65g364j
> - May 2025 - [VOTE] Versioning process: not carried. Daan -0 pending
>    exceptions (hypervisor support, security releases); Rohit -1 (binding)
>    mainly on restricting DB schema changes to majors, while being
>    supportive of dropping the "4." itself.
>    https://lists.apache.org/thread/wf8910ln7wqn9g535ob0n04docbn2jzd
> - Sep 2025 - [LTS] Extend LTS support to 24 months: adjacent, explicitly
>    scoped away from the versioning scheme.
>    https://lists.apache.org/thread/fkl74js7vxml9c29jqmhl2ctsxsq868o



-- 
Daan

Reply via email to