It looks like this is something that should go upstream. Why not open a merge request?

On a related note, could you put a little more effort into your commit messages?

It is an ad-hoc fix to make the code work with what I am doing in modepy from modepy@47058eb4 to modepy@ad78f308, which is then a part of ongoing effort to make an element that exactly reproduces what volumential uses for volume function discretization.

Name-wise, I found uses of both Gauss-Legendre and Legendre-Gauss, and it is called simply the Gauss quadrature in 25.4.29 of Abramowitz and Stegun.

I thought it is better to consolidate changes that I want to make to different projects before submitting it upstream.

(Most of those meaningless commit messages come from my abusing Gitlab for syncing code between computers, which is admittedly better to be avoided.)

On an unrelated issue, I cannot seem to find a way to assign/subscribe to a commit like this within Gitlab system. Am I supposed to bookmark the link myself and get back to get updates from time to time?

(Most of those meaningless commit messages come from my abusing Gitlab for syncing code between computers, which is admittedly better to be avoided.)

OIC

On an unrelated issue, I cannot seem to find a way to assign/subscribe to a commit like this within Gitlab system. Am I supposed to bookmark the link myself and get back to get updates from time to time?

I believe it works like the other discussions it Gitlab. As the original author of the commit, you're automatically subscribed. If someone @-mentions you, you're also subscribed. (And I'm not sure there's a way to unsubscribe...)

It is an ad-hoc fix to make the code work with what I am doing in modepy from modepy@47058eb4 to modepy@ad78f308, which is then a part of ongoing effort to make an element that exactly reproduces what volumential uses for volume function discretization.

OIC. It looked like a wrong import. Didn't realize GaussLegendre et al were new, sorry.

Name-wise, I found uses of both Gauss-Legendre and Legendre-Gauss, and it is called simply the Gauss quadrature in 25.4.29 of Abramowitz and Stegun.

Not sure I have much of a preference, as long as it's consistent (in modepy).

I thought it is better to consolidate changes that I want to make to different projects before submitting it upstream.

Fact: tiny merge requests are easy to review. Monster merge requests are exponentially harded. No need to batch them up.

But I agree that a set of related changes across projects should likely go in mostly together.