oil 0.7.0 (new formula)

[waldyrious]

• Have you followed the guidelines for contributing?
• Have you checked that there aren't other open pull requests for the same formula update/change?
• Have you built your formula locally with brew install --build-from-source <formula>, where <formula> is the name of the formula you're submitting?
• Is your test running fine brew test <formula>, where <formula> is the name of the formula you're submitting?
• Does your build pass brew audit --strict <formula> (after doing brew install <formula>)?

[waldyrious]

@andychu, for the formula's test block I used the "smoke test" section of https://www.oilshell.org/release/latest/doc/INSTALL.html for inspiration, but at some point during local testing, the produced AST for the demo script I was using changed as follows:

-       spids: [5 18]
+       spids: [4 17]

Any insights as to what could explain this change? I don't want to introduce a test that isn't deterministic.

[andychu]

-n is more of an informal demo for users -- I wouldn't use that as the test. That format is subject to change at any time (and maybe I should make that more clear somehow)

Also despite the versioning, the later releases are almost certainly strictly better than earlier ones, so

https://www.oilshell.org/release/0.8.pre2/

is worth packaging.

Thanks for packaging Oil! Please update the wiki when it's merged and let me know if you have more questions.

https://github.com/oilshell/oil/wiki/Oil-Deployments

[andychu]

Oh also eventually Oil will be distributed with tests you can run. That's this issue:

oils-for-unix/oils#167

If you want to help let me know :) There's been some big progress recently at http://travis-ci.oilshell.org/jobs/ (details on Zulip)

[chenrui333]

@andychu just curious why this oilshell depends on Python 2, not Python 3. Also, depends on python 2.7.13 (not 2.7.16)

[chenrui333]

Thanks for bring this shell into my attention, good to know there is always another one exist. :)

[andychu]

The goal is to remove the CPython dependency altogether. Some details:

http://www.oilshell.org/blog/2020/01/parser-benchmarks.html

http://www.oilshell.org/blog/2019/06/17.html#why-is-it-written-in-python

[SMillerDev]

If it depends on Python it should define uses_from_macos "python@2"

[andychu]

It doesn't require Python to be installed: From the packager's perspective it looks like a C program.

[SMillerDev]

Does it require python to run?

[waldyrious]

-n is more of an informal demo for users -- I wouldn't use that as the test.

What would you suggest instead, then? Perhaps one of the spec tests?

despite the versioning, the later releases are almost certainly strictly better than earlier ones, so oilshell.org/release/0.8.pre2 is worth packaging.

Sure, I can do that. @SMillerDev, @chenrui333, is that OK from your side as well?

[andychu]

I would just remove it -- -c "echo hi" is a good sanity check already. And as mentioned Oil issue 167 is for people who want more than that.

@SMillerDev It doesn't require Python to run

[waldyrious]

Thanks for packaging Oil! Please update the wiki when it's merged

No problem, and will do 👍

[SMillerDev]

is that OK from your side as well?

Nope, only stable releases are allowed.

@SMillerDev It doesn't require Python to run

Okay.

[waldyrious]

Another question: is there any way to link to the source code repository of the formula when it has a dedicated website? I know I could use head, but IIUC that implies that the formula should support building from the master branch, which AFAIK is not the case — according to the README:

Developer builds don't work on OS X, so use the release tarballs on OS X.

[SMillerDev]

You can use head for that. While it allows people to try and build the software, that configuration means that you're on your own if it breaks.

[waldyrious]

I've added the head block and simplified the test. I also made a minor adjustment to the head block of another formula, but I can submit that as a separate PR if it's preferable.

[SMillerDev]

Please make this a separate PR

[waldyrious]

Will do 👍

[waldyrious]

@SMillerDev I suspect I'll need to also fix this if I am to submit that change, but I'm not sure exactly how to do that.

Can you clarify what adjustment I would need to do to gobby/0.5.0.patch, and whether that would need to be done before or after the PR in this repo?

[waldyrious]

Fixed the audit error and removed the change to an unrelated formula.

[SMillerDev]

If you can just squash the commits into one <name> <version> (new formula) now?

[waldyrious]

Sure, I wasn't sure if force-pushing was OK. Will squash now.

[chenrui333]

@waldyrious Thanks for your contribution to Homebrew! 🎉 🥇

Without awesome contributors like you, it would be impossible to maintain Homebrew to the high level of quality users have come to expect. Thank you!!!!