If there are no more changes required to the pull request it could do with a bunch of likes from racketeers and maybe a supportive comment from maintainers of racket/racket to attract attention of the maintainers of https://github.com/github/gitignore
Oh, don't get me wrong, I thumbs-upped the pull request; I wouldn't torpedo the existing PR by expressing doubt on that github issue's comment chain. This is something that can be edited later or never; the proposal is a clear improvement on what's there now (that is, nothing at all).
I agree -- I think the PR looks good! Even if the maintainers think that e.g. .DS_Store shouldn't be in there because they prefer that stuff to be in /Global, or whatever, I'm sure they'll give you that feedback.
For what it's worth, omitting # files works better for me than not omitting them. An unsaved change is rarely something that I want to keep, because I've normally just run a build/tests that use the saved files. Instead, a # file is usually because I accidentally typed < instead of Meta-< at some point when trying to jump to the start of something that I'm merely reading (and I haven't quit the editor). Omitting # files almost certainly saves me more trouble than having them added in the racket/racket repo. Also, putting rules in a machine-specific configuration is not a good solution for me, because I spend a lot of time on short-lived virtual machines or Docker containers, and it's sometimes easiest to push repairs from those environments.
All that said, I don't have a strong opinion for raco pkg new.
I think this is about right as it means new users will get a sensible default .gitignore that will hold them over until they are doing their own /Global gitignore configuration.
More feedback is encouraged.
PS
If anyone knows a way to get the attention of the maintainers of https://github.com/github/gitignore so the PR can be merged that would be appreciated