[QGIS-Developer] What to do with evis?

[Nyall Dawson]

Hi list,

With Denis' recent PR adding native support for browsing
features/tables  (see https://github.com/qgis/QGIS/pull/9992), I think
it's a good time to open discussion about the future of the evis
plugin.

In my understanding, browsing tables was one of the remaining
compelling features which evis offered, and if we have a native
integrated support for this then I doubt there's a good reason to keep
dragging around the evis baggage with us.

My 2c:

- The plugin is well and truly abandoned and unmaintained. It's only
seen fixes to keep it compiling for the last 5+ years. There's
obviously no demand from the user or development community to extend
this plugin. This either means the plugin is unused, or it's "feature
complete". My gut feeling is the former, or we'd at least be seeing
bug reports still flow in.

- It's a substantial amount of code, and it does require work to keep
it functional. If this work isn't needed, we're just wasting our time

- It's a core c++ plugin, which we are trying to reduce. Again,
there's been no movement from anyone I know of to port this to a
not-installed-by-default python plugin (which is where it really
belongs)

If I'm not the only one thinking evis has reached it's eol, I would propose we:

- Add a messagebar warning when the plugin is enabled/loaded on QGIS
3.8, stating that it is deprecated and will be removed in QGIS 3.10.

- Link this message to a QGIS blog post detailing the reason for
removal, which can clearly state that if users require the
functionality they should contact the QGIS project to discuss funding
a community maintained, Python version of the plugin

- Remove it in 3.10.

Thoughts?

Nyall