Deprecate plugincheck in favor of plugincheck2
tolzhabayev opened this issue · 5 comments
- Remove
plugincheck
from the readme and only keepplugincheck2
- Show a message to migrate to
plugincheck2
every time it's ran. Do not remove code yet. - Add a deprecation notice to the
readme.md
of theplugincheck
pkg - Mark code as deprecated https://stackoverflow.com/questions/7849663/how-do-you-mark-code-as-deprecated-in-go
- Search for usages of
plugincheck
in the wild: -
- At least 14 plugins are currently using it
@tolzhabayev since both of them are compatible can we start putting the plugincheck2 binary in place of plugincheck after a while?
Are they really replaceable / a drop-in replacement?
Are they really replaceable / a drop-in replacement?
The parameters are the same and the share the same configuration. I am not really sure why they were made as different projects. I'll ask around.
Ideally we should end up with one project that is versioned (like our go sdk) - we need to figure out what is better:
- Keep
plugincheck
and release a new version of it and deprecateplugincheck2
- Means we would have to migrate everyone and their CI's, etc. to
plugincheck
- Means we would have to migrate everyone and their CI's, etc. to
- Keep
plugincheck2
and deprecateplugincheck
which I would not change at all, also not place another binary into it - just leave as is- Means we would have to keep
plugincheck2
name forever and iterate on it using versioning (e.g.plugincheck2 version 0.312
). Naming might be a bit confusing. - Migrate everyone to use
plugincheck2
which is at least already the case inplugin-tools
and plugin CI pipelines generated from it
- Means we would have to keep
@tolzhabayev we are almost in the second version. The only problem we have with deprecating plugincheck
is that we can't control who is using it. We could delete the code for it and break the few places in some third party plugins that are using it and expect they realize they are using the wrong version.
I don't see option 1 as a viable option because it'd be considerable more effort to move everyone