• AstarVienna/ScopeSim_Templates#89
• AstarVienna/speXtra#39
• AstarVienna/HowManyBloodyPhotons#14
• AstarVienna/skycalc_ipy#66
• AstarVienna/astar-utils#37
• AstarVienna/AnisoCADO#13
• AstarVienna/ScopeSim#460

In #9 we moved from markdown-link-spec to linkspector.

There was no github action for linkspector at the time, so we had to setup something ourselves. Now there is an 'official' github action, so we could move to that: https://github.com/UmbrellaDocs/action-linkspector

Also, there now is a default configuration file, so it is not necessary to create one: UmbrellaDocs/linkspector#14

Perhaps we should have a workflow that builds the documentation, so we don't have to rely on RTD to find bugs in the documentation

There is currently only one 'test' in this project, which runs all the tests and notebooks of the other ScopeSim projects, and then creates a pull request for any new files that are fetched by those tests. Normally, there should be no such new files; there should only be new files when there are new tests, e.g. using a new filter or something like that.

This means that under normal circumstances, there is no test that checks whether the servers that e.g. skycalc_ipy tries to reach actually work. It would be useful to know whether those services are still up, so we should create some tests that run without caching data. However, maybe that should be done in the projects themselves?

Perhaps we could do mutation testing, e.g. with mutmut.

That is, we mutate our (passing) tests automatically, to see whether they then fail. This can be useful because many of our tests do not actually assert much, they just verify that the code runs without raising exceptions. As in, the current tests give us coverage, but no confidence that the result is actually correct. Mutation testing would allow us to see which tests could be improved and how.