Use consistent notation for complexity annotation

[alexfmpe]

Fix #722

[treeowl]

One of these modules somewhere uses MathJax. Let's make them all do that.

[sjakobi]

One of these modules somewhere uses MathJax. Let's make them all do that.

@treeowl Could you clarify what you mean here?

FWIW, what I was looking for in this PR was a proper fix for #722. Any other possible haddock improvements don't need to happen in this PR IMHO.

[treeowl]

One of these modules somewhere uses MathJax. Let's make them all do that.

@treeowl Could you clarify what you mean here?

Look how all the math is formatted in Data.Sequence. Let's do that everywhere.

[sjakobi]

Look how all the math is formatted in Data.Sequence. Let's do that everywhere.

Oh yeah. That does look nicer! I'd argue that this is a different issue than what this PR was about though.

[sjakobi]

Two more tweaks, otherwise LGTM.

[alexfmpe]

Look how all the math is formatted in Data.Sequence. Let's do that everywhere.

If that's the convention going forward, I might as well make these particular annotations conform to it now

[sjakobi]

Look how all the math is formatted in Data.Sequence. Let's do that everywhere.

If that's the convention going forward, I might as well make these particular annotations conform to it now

I wouldn't mind. I don't currently volunteer to update the remaining annotations though.

[sjakobi]

Could you share another screenshot, @alexfmpe? I think you're somehow using fewer backslashes than the annotations in Data.Sequence tend to have.

Here are the docs BTW: https://haskell-haddock.readthedocs.io/en/latest/markup.html#mathematics-latex

[alexfmpe]

Good catch, the log/min functions were being rendered as variables

[alexfmpe]

[sjakobi]

Gorgeous! Many thanks! :)

[sjakobi]

I've made #731 where we can track the remaining annotations to convert.

[treeowl]

Why did the complexity info move?

[sjakobi]

Why did the complexity info move?

That was my idea. It's much longer now, so it would have distracted from the function description if we had kept it in the old spot.

[treeowl]

We need a good and consistent approach to that. As I recall, @m-renaud wants it always at the bottom. I'm skeptical about that, because the complexity of an operation seems like one of the most important things to know about it. But it's also true that a long complexity statement before the basic info about what an operation does is problematic. So I dunno.

[sjakobi]

Putting it consistently at the bottom (above the since-annotation?) sounds fine to me.

A closed PR doesn't seem like a good place to discuss this further though. Consider opening an issue?

[m-renaud]

I was actually thinking about this the other day, we have the @since metadata which has a consistent rendering across packages, could we introduce standard @complexity and @complexityInfo annotations and do something similar. Agreed that this is probably not the best place to discuss though.