Restrict "other" interfaces in devicetype-library PR's going forward

[DanSheps]

What would everyone's thoughts be on restricting the usage of "other" for a interface (or other connection type) in the devicetype library.

The main driver is:

#9098 in netbox, which is potentially going to open up "other" for more types, and thinking about it

In the device type library I think it would make more sense to have well-defined types being used instead of other so we don't just end up with a bunch of "other" interfaces/ports/etc

One option would be to adjust the CI to see if we can flag a warning or maybe have an optional check to see if there are types with "other" in them and throw a warning but not block merge.

Just some thoughts, feel free to add any comments

[jeremystretch]

IMO the "other" classifications are a stop-gap, and we should discourage its use in the "official" definitions. I'm in favor of restricting the use of "other" for all component types in the absence of a solid argument otherwise.