New package merge conflicts

[mahrud]

Whenever there are multiple PRs for adding packages, after the first one is merged there is a merge conflict.

This is kind of silly and we should be able to fix it. Here are a couple of solutions:

1. get rid of =distributed-packages; distributed all .m2 files in packages directory.
2. alphabetize =distributed-packages so new packages aren't all added at the end.

What's the point of having =distributed-packages and the version#"packages" variable anyway?

[DanGrayson]

Those are both good ideas.

The original reason for the file was so we could have undistributed packages. But now we have the subdirectory undistributed-packages, so that takes care of that.

[mahrud]

A related issue is that there are ghost packages: in 51d352fbdb9f5903c5bedbd5dce0c14d3fc66d2d, 9224486f3fc4b8e00e883570756ab969be351009, and a76b7f70a983a7a0acaf2b0d84bdecc27ba36157, three published packages were added, but the commits are very messed up:

• the one titled "ThreadedGB" also touches MultiprojectiveSpaces,
• the one titled "CodingTheory" also touches ThreadedGB, and
• the one titled "SchurVeronese" also touches CodingTheory and MultiprojectiveSpaces!)

and in the middle of all this SchurVeronese was never added to =distributed-packages! This results in a dead link from the changes, 1.20 page to a ghost package.

The other ghost package is "NCAlgebraV2", which maybe was supposed to be deleted once "NCAlgebra" was added?

[mahrud]

@mikestillman @DanGrayson the package "SchurVeronese" is still not added properly! It's documentation is inaccessible despite it having a gold star.

[mikestillman]

@mahrud @DanGrayson Is this still a problem, or can we close this issue?

[mahrud]

This issue is about eliminating the distributed packages file, which is still there, why close it?

[mahrud]

701 is also about this same problem.