Plugin versioning pattern/semver adoption #8017
Replies: 3 comments 1 reply
-
|
Aunque a priori no se me ocurre ningún problema en cambiar de 4 dígitos a semver, comentar (por ver las ramificaciones) que las imágenes docker usan los mismos nombres de versión que las herramientas OxS. |
Beta Was this translation helpful? Give feedback.
-
|
Note that our If the plugin's |
Beta Was this translation helpful? Give feedback.
-
|
I’d like to undust this conversation to propose a solution that I believe resolves the mentioned issues (and hopefully doesn’t create new ones). :-P
From discussions on Mattermost, it’s clear that:
In this sense, why not use branches to indicate the OJS version and tags to indicate the plugin version? The solution:
From what I’ve seen, some developers have already started adopting this approach in their new plugins. Can we reach a consensus to recommend this practice to new developers? |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Small discussion about versioning, here's the context: https://github.com/pkp/pln/blob/main/version.xml#L16
Sparing a range to avoid the versions to overlap (looks like we already have some cases) across different branches ends up breaking the versioning meaning.
I'm used to have a unified version/branch and adapt the code to work on other places with the help of "
#ifdefs", but that's not suitable for our use-case, the version for 3.4 might have more features than the one for 3.3 for example, and they can co-exist for a long time.Ideas:
version.xmlstating its compatibility, which is useful when installing the plugin, and can be also re-used by the plugin gallery (instead of adding the such information on the plugins.xml):w.x.y-z) in favor ofsemver, it has room for pre-release and metadata tags.ctgraham
Yes to plugins declaring their compatibility within
version.xmlas a dedicated tag!Note that this will soon be much easier as we can now support composer/semver constraint statements: #7793
p.s.: rescued from a Slack conversation
Beta Was this translation helpful? Give feedback.
All reactions