[Qgis-developer] QGIS 3 plugins - require implementation as processing algs?

[Denis Rouzaud]

On 09/12/2016 01:34 PM, Alexandre Neto wrote:
> On the other hand, like Vincent said, we don't want to have a bunch of
> angry plugin developers. We also don't want to have thousands of
> misfit and duplicated plugins around. But I'm not sure I would like to
> go back to the time that, as a user, you needed to be fishing
> repositories to get some nice extra functionality. I think that this
> will increase the number of duplicates and make the life of both users
> and developers harder.
>
> Although we should have rules, can we manage part of this impositions
> with the trusted plugins?

I feel a bit puzzled too. To me, it's a bit harsh to force someone to
develop one way. People developing plugins are not necessarily advanced
programmers, and might found a bit too complicated to have to dig into
documentation. I believe many (like me) usually just copy existing code
and try to modify it by small steps. Having to redesign the whole plugin
would sound way too hard/long.

I like the idea of the trusted plugins. From the discussion in Girona
about trusted developers, it could be a nice way to have a short list of
well-made, non redundant, not too specific plugins. These would be
tagged as such. Then, the big list with less quality control.

Plugins are probably one of the reason of QGIS' popularity and
preventing plugins authors from publishing their work easily is not a
good move to me. Having a quality labeled list sounds like the good
deal. It shouldn't be too hard to maintain such list. I wouldn't mind
being involved in this.

My 2 cents.