Entry-points don't comply with the standards

[bastonero]

Hello,
As I am adding some basic testing for the entire package, I noticed that the entry-point names choice is not on the same line as in aiida-quantumespresso(-hp).

What do you think?
Shall we rename them before a public release?
What about all the databases run with the old entry-points?

[sphuber]

Technically, the use of quantumespresso_ph is actually consistent with the guidelines. It is aiida-quantumespresso-hp that is "illegally" hijacking the quantumespresso namespace. I did this back when I started the plugin because I thought I would be merging it back into aiida-quantumespresso at some point and didn't want to have to change the entry points. I am not sure whether we want to keep the repositories separate, but that is a separate discussion.

The reason for the guideline is twofold:

1. It prevents packages having entry points with the same name
2. It makes it easily visible to a user which package an entry point comes from

For the first one, we simply have to make sure that the entry points don't clash with aiida-quantumespresso, but since we are also the maintainers, that should not be a problem. The second point is affected negatively a bit by reusing the quantumespresso namespace, but a user only has to look at the actual path of the value that the entry point points to, and it will contain aiida_quantumespresso_ph, making it clear.

So conclusion, for consistency with aiida-quantumespresso-hp I think it would be best to also take quantumespresso as the namespace for the entry points in this package. The consequences of violating the guidelines are not really that bad I would say.

[bastonero]

Thanks for the reply! I didn't want to say it is bad, it is just maybe a little confusing at first, as I would have expected all entry-points to start with quantumespresso, but some were and some didn't.