Open germa89 opened 2 months ago
Pinging @AlejandroFernandezLuces for feedback.
Hmm that's an interesting problem... If I'm understanding the issue right, I would separate the visualizer to update individually instead of being included in a group. Since updating this involves more changes than just bumping the dependency, I think it being independent is justified and it solves this issue.
In any case I'm not sure why the cache update is being triggered there. It seems that the update is somehow linked to the group to the single dependency.
On https://github.com/ansys/pymapdl/pull/3386 the image cache got updated, but "in theory" no plotting dependencies were updated.
Well... in that PR, the visualizer was updated. Later I remove that update.
So it seems we are not pinging properly the important dependencies for testing. Since pyvista started to be a "subdependency" (dependency of a dependency, visualizer), I have seen this twice.... I did fix it first time in: https://github.com/ansys/pymapdl/pull/3371/files
I believe I should add pyvista as a test dependency so we make sure we are getting the same stuff everytime.