Re: Math package: lcalc-1.23

[tornaria <tornaria@users.noreply.github.com>]

[-- Attachment #1: Type: text/plain, Size: 1822 bytes --]

New comment by tornaria on void-packages repository

https://github.com/void-linux/void-packages/pull/30036#issuecomment-815220499

Comment:
General answer to all comments above:
- This program is kind of old and not maintained by the author. It works and it's useful in what it does, but it should be treated mostly as frozen (so, I doubt example programs will change).
- It's included and used in sagemath, which I used as upstream, they will do maintenance if pari changes (I can do it as well).
- patches came mostly from sagemath (https://git.sagemath.org/sage.git/tree/build/pkgs/lcalc/patches)
- I did this in 2018 and I just noticed they have updated patches, in particular there is some makefile cleanup that might be nice to have (like using $(CXX), etc, so cross build will work as noted above).
-  I could update it to have *exactly* the same patches as in sagemath (either included in or as distfiles from a stable tag of the git repo above) and think of sagemath as the upstream.
- The test is just some quick hack I did to know the program worked ok. I can make it smaller if that makes it more palatable. I think at least knowing that the binary runs and that the library can be linked to is good, but maybe instead of computing 1000 zeros of the Riemann zeta function I can just compute 10, etc.

Let me also point out that, other than sagemath, I don't think there will be any dependency of this program. But when we get to package sagemath, this program will have to be compiled, either as a separate package (like here) or as part of the omnibus sagemath.

Also: I rather spend my time and energy on packaging other -- more generally useful -- math software, than polishing too much this (niche) program. All criticism is still very much welcomed, I'm just starting with the "easy" packages.