Add @show_name invocations

[fingolfin]

Also update ConstPolyRing show method

[lgoettgens]

@show_name requires the object to have attribute storage enabled

[fingolfin]

I don't think so? We specifically wrote it to also work for non-attribute storing types.

But @show_special does require it. However I wonder if we should change that, so that one can write show methods w/o having to worry about attribute storage being available or not.

But that's for another PR, if at all. For now, perhaps we don't want matrix spaces to print by name because they are supposed to be uniquely identified by their base ring plus dimensions. So perhaps I'll instead leave a comment there indicating we deliberately don't use show_name and show_special.

[lgoettgens]

I don't think so? We specifically wrote it to also work for non-attribute storing types.

But @show_special does require it. However I wonder if we should change that, so that one can write show methods w/o having to worry about attribute storage being available or not.

I'll provide a PR that makes show_special a noop in these cases

[fingolfin]

With the ConstPolyRing changes moved to PR #1894 this is now rather small: on the one hand MatRing and MatSpace as mentioned before -- but perhaps we ought to retain those as they are?
And then for NewRing in the tests. That show method is specifically for testing the three printing modes, so perhaps it should also stay as it is.

So in principle I'd be OK with closing this...