Rename is_vertical to passes_transversality_checks and execute related checks

[HereAround]

No description provided.

[apturner]

This looks good to me after those 2 changes, thanks @HereAround

[HereAround]

@aaruni96 I cannot seem to find the label now, but could this PR please also be backported to 1.3.1?

[lgoettgens]

@aaruni96 I cannot seem to find the label now, but could this PR please also be backported to 1.3.1?

If I see correctly, this PR renames a function. If this wasn't in experimental this would be a breaking change, thus needing a major bump in the version number. IMO this is therefore not a bugfix that should be backported, but of course the last word is to @aaruni96 and @benlorenz

[benlorenz]

Renaming for backports seems wrong, but if I see this correctly this was just renamed in the other direction about a month ago (#4562, i.e. between 1.2.2 to 1.3.0), also without any deprecation or alias. Maybe you can explain this a bit?

We could probably add the new name but with an alias or a deprecation ?
(And I don't really know what execute related checks means)

Edit: I now realize the other renaming was slightly different, but nonetheless the is_vertical name existed only for a very short time.

[HereAround]

@benlorenz Yes, let me elaborate.

About maybe 15 years ago, people in the F-theory community were talking about "fluxes being vertical" (whatever a flux is and whatever this may mean - happy to elaborate, but this is beyond the scope of this answer). By now, the phrase for a flux having this very property is instead "flux passes transversality checks". On top of this wording change, today "flux being vertical" means something entirely different than 15 years ago. Sadly, I have mirrored this historic shift of terminology/meaning in OSCAR.

When I brought in is_vertical, I was in some places of OSCAR using it to mean that it did about 15 years ago. However, in other places, is_vertical was used in today's meaning and some sanity checks, that actually should be executed, were not done. This PR does two things:

1. The first thing that this PR does is to disentangle the terminology. What we currently have as check in OSCAR is only passes_transversality_checks. In particular, no check is_vertical should currently exist in OSCAR.

2. With this clarification, some default output must always satisfy passes_transversality_checks. Likewise, some sanity checks must include the transversality check. This latter change is what my cryptic message "execute related checks" refers to. (Btw - this change is what caused the extra long tests to fail earlier this week.)

[benlorenz]

This this is in experimental and only exists in 1.3.0 (considering releases), I would be fine with renaming this and adding an alias or deprecation for 1.3.1.

2. Btw - this change is what caused the extra long tests to fail earlier this week.

This means backporting this PR will break these tests on the release branch?

[HereAround]

Oh, I did not think of this. Indeed, this would break the extra long tests on the release branch. So presumably not a good idea to backport this change.

[HereAround]

Unless we also backport a fix for the extra long tests, which I am happy to provide. But then, maybe I should just wait for OSCAR 1.4.0?