Is a dot-import a bad practice?

[magic tundra]

yes, quite frankly it’s typically considered bad practice outside of testing code, but if you want to then nobody’s going to stop you 🤷‍♀️

[knotty glen]

one way to solve that is making most of your stuff receivers

that way only place it shows is like v := mymath.NewVector() and then it's v.This() v.That()

btw... I don't know the syntax... what's dot import looking like (8 years of go, never used it)

oh interesting

type aliases are fine, but maybe a smoke that the name of your package is not great

vector ?

or vect?

what's the f2 standing for?

fortran 2 ?

but you're on the right track, naming is hard... but "better names"

oh... that's not great name I think

well I was about to say Point but... realized it could be 2d or 3d

why not just
type Vector struct {
xi []float64
}
(though maybe you care about memory foot print)

PlanePoint vs VolPoint?

Plane implies 2 dimensional

but maybe just Point (for 2d) and Vector (for 3d)

[reef gorge]

Personally I don’t see a problem with it for some math/vector related stuff. Oftentimes math libraries can be a bit nonidiomatic to make usage and semantics easier. Just make sure to not pollute the namespace too much

[lapis juniper]

If my names conflict with go names I follow my madeup convention of prefixing std packages with go- , works fine for me

gocrypto gomath gorand etc

Packages like fhttp are nearly 1:1 std ones and they always shadow std packages