The hash in this commit seems to be wrong. Did the tarball change again?

hash mismatch in fixed-output derivation '/nix/store/q91m00l6dkljka4rqqrm606w4wq0ycps-nim-1.4.0.tar.xz':
  wanted: sha256:0pch14xkymzhrkmyai8i5fcjaxicgjgbzyk3vph6qzvkdfmj0cdy
  got:    sha256:0gf2lqkqzai6mg7mf4y04gdy1ddiavans09i8aisa88ssfza5ywx

@ehmry do you still have the source tarball with sha256 0pch14xkymzhrkmyai8i5fcjaxicgjgbzyk3vph6qzvkdfmj0cdy in your store? If so, can you diff the contents of this and the new one?
If they keep changing it we should switch to pulling the sources directly from a git checkout/tarball maybe? CC @dom96

@petabyteboy I do not. If this is how they are going to do releases then we should stay one version behind.

Hi! @ehmry @petabyteboy is there a chance this will get merged, or not until next version is out (which may be unspecified time in the future)? would there be higher chance if this expression was changed to build from github sources? I kinda miss Nim 1.4, and I wonder if there's a way I (or someone else) could help things advance...

Seems like they haven't switched out the release tarball since I last checked a month ago, so we could probably go ahead with nim. I pushed the new hash to this branch. Still need to package nimble seperately.

Can we leave nimble as is here, and extract it later?

Can we leave nimble as is here, and extract it later?

After this change nimble isn't packaged at all. This would be very confusing to me as a user when I'm used to having nimble installed as part of the nim package.

I think packaging nimble separately is momentarily confusing but inevitable.

Actually not a problem, I will package nimble seperately from the unwrapped compiler and combine it into the wrapper.