[#964] Allow installing seq from GNU ELPA

[raxod502]

I think this is the right solution to #964 - per Jonas, the version of seq.el on GNU ELPA is newer than what is in Emacs now, and it is a legit package that people need to be able to install as a dependency. I'm not sure why this was in the list to begin with, maybe things were different in the past.

[progfolio]

I'm not sure why this was in the list to begin with, maybe things were different in the past.

Git blame says seq was ignored in light of this: #538

[raxod502]

Blargh, guess I should have checked that.

But I still think this is the right change, actually, and the addition of seq in that PR was only a workaround to another bug, which I think may have been fixed since.

I had a read through doomemacs/doomemacs#3406, and it looks from the backtrace in doomemacs/doomemacs#3406 (comment) like the issue is cl-subseq being defined using seq-subseq, and seq-subseq also being defined using cl-subseq.

Now looking at the current version of seq, it seems the implementation of seq-subseq no longer relies on cl-subseq. That changed in this 2020-09-04 commit in GNU ELPA Mirror, which corresponded to this 2020-09-04 commit in GNU ELPA, which corresponded to this 2019-10-27 commit in Emacs core.

The Doom bug was reported 2020-06-17, which precedes all of this by a couple months, so I think it is safe to say that with three years having gone by since the functions no longer have a mutual dependency, we can probably remove this workaround. Does that seem reasonable to you?

[progfolio]

Sounds good to me