Adds setoid export to Algebra.Bundles.SemiringWithoutOne (possible solution to #1917)

[jamesmckinna]

What I don't understand is why/how this solution is to be preferred over others (such as?), or whether this may break anything subsequently... nor how this problem came about in the first place... nor what, if any, CHANGELOG entry might be suitable to account for the fix.

[Taneb]

As an aside I've been wondering if there's a good way to either test that this sort of thing is done right, or rewrite the library so all the re-exports are there by construction.

[jamesmckinna]

Well, hence my mention of #1917 on #1967... and my hope that you will publish your big diagram at some point, as the source for exploring your "good way to test"/"rewrite the library"... but I'm glad it's not just me who feels this kind of things is too difficult to reliably debug 'by hand'!

UPDATED: What remains odd for me is that the explicit reexport of _≉_ is still present in a number of the Algebra.Bundles, yet does not give rise to the same ambiguity of names. So, for me at least, even the 'solution' offered here leaves questions as to whether there is a better, more systematic one, out there.

[jamesmckinna]

status:ready modulo CHANGELOG of course...