emacs 27.1 #59448

CamJN · 2020-08-11T02:42:49Z

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>)?

there were issues with the audit, listed below, not sure how to fix them:

emacs:
  * Formula emacs contains deprecated SPDX licenses: ["GPL-3.0"].
  * Files were found with references to the Homebrew shims directory.
    The offending files are:
      libexec/emacs/27.1/x86_64-apple-darwin19.6.0/emacs.pdmp

miccal · 2020-08-11T03:11:44Z

I can answer one of your questions:

* Formula emacs contains deprecated SPDX licenses: ["GPL-3.0"].

The license should be:

license "GPL-3.0-or-later"

aconchillo · 2020-08-14T22:43:14Z

Seems to be affected by #58984.

CamJN · 2020-08-14T23:42:18Z

@aconchillo can you elaborate how so? I looked at the issue you linked and it looks like a Haskell sandbox issue?

aconchillo · 2020-08-15T05:46:32Z

Yes, there's been other packages unrelated to Haskell AFAICT with the same issue. For example: #58683. I don't know how agda has nothing to do with all this, but I'm experiencing the same issue in this PR #59490.

gromgit · 2020-08-15T06:31:22Z

Version-bumping any formula requires testing all the formulae that depend on it:

$ brew uses --include-build --recursive emacs
agda                   cask                   cedille                doxymacs               eless                  emacs-dracula          gforth                 lfe                    mu                     nesc                   proof-general          rtags

aconchillo · 2020-08-15T08:16:29Z

Thanks @gromgit. That makes sense and I guess it explains the long times of the check.

aconchillo · 2020-08-17T22:29:36Z

This PR #59709 will fix the issue once merged.

chenrui333 · 2020-08-23T17:44:46Z

@CamJN can you rebased your PR, I just merged the Agda fix PR.

CamJN · 2020-08-23T18:39:14Z

Still getting this:

=> FAILED
40
Error: 1 problem in 1 formula detected
41
emacs:
42
  * Files were found with references to the Homebrew shims directory.
43
    The offending files are:
44
      libexec/emacs/27.1/x86_64-apple-darwin19.5.0/emacs.pdmp
45

I think this is due to emacs now using a new portable dumper instead of the abomination that was unexec.

CamJN · 2020-08-24T00:12:13Z

Indeed, the portable dump file does refer to /usr/local/Homebrew/Library/Homebrew/shims/mac/super according to strings.

How does homebrew usually deal with things like this?

CamJN · 2020-08-24T00:20:15Z

Some discussion here: https://archive.casouri.cat/note/2020/painless-transition-to-portable-dumper/index.html

CamJN · 2020-08-24T01:27:02Z

I rebuilt using the --with-dumping=unexec configure option, and ran find . -type f -exec fgrep -ai '/shims' {} \; -print which seems to not include a reference to the shims dir, except in the config.log file.

That said, that's not a viable strategy going forward as unexec is going away eventually.

SMillerDev · 2020-08-24T07:09:43Z

What files does it reference?

CamJN · 2020-08-24T17:01:58Z

@SMillerDev it just refers to the path to the shims super env dir not a file in there, in fact I believe (but don't know for sure yet) it is just capturing the PATH env var at the time it dumps the new executable.

CamJN · 2020-08-24T18:58:05Z

@SMillerDev here's the strings output that makes me think it's just the PATH:

$ strings /usr/local/opt/emacs/libexec/emacs/27.1/x86_64-apple-darwin19.6.0/emacs.pdmp | rg -B 13 shims
/sbin
/usr/sbin
/bin
/usr/bin
/usr/local/opt/gnutls/bin
/usr/local/opt/libevent/bin
/usr/local/opt/openssl@1.1/bin
/usr/local/opt/p11-kit/bin
/usr/local/opt/nettle/bin
/usr/local/opt/libtasn1/bin
/usr/local/opt/libidn2/bin
/usr/local/opt/gettext/bin
/usr/local/opt/pkg-config/bin
/usr/local/Homebrew/Library/Homebrew/shims/mac/super

aconchillo · 2020-09-02T00:58:56Z

This is not directly related but in #60307 (comment) I believe I explained why Homebrew does this type of checks. Basically, when a bottle is built Homebrew doesn't want references to Homebrew prefix hardcoded in files, this is so at installation time a new prefix can be used.

Homebrew takes care of this in text files when building a bottle by replacing Homebrew path references with placeholders and replaces them back with the Homebrew prefix at installation time, but in this case it's a binary file so it can't do any of that.

I don't really know enough about Homebrew but this seems a tough one. Without knowing much, the only thing that comes to my mind is creating a link to the shims directory:

/usr/local/opt/homebrew-shims/ -> /usr/local/Homebrew/Library/Homebrew/shims/mac/super

or something like that and add the /usr/local/opt/... directory to PATH instead of /usr/local/Homebrew/.... This way the formula shims check will succeed and bottles will be able to install in different Homebrew prefixes as long as /usr/local/opt is fixed (i.e. the same in system where bottle is built and system where bottle is installed).

This of course involves changing Homebrew, so hopefully there's a much simpler solution. I use emacs in the terminal on my day to day work, so I'm very interested in this one 😄 .

CamJN · 2020-09-02T01:04:25Z

I've been busy, but I have a theory that we can edit the PATH by loading some elisp before the executable is dumped. I'm going to try but I can't do it right now.

aconchillo · 2020-09-02T01:14:33Z

I've been busy, but I have a theory that we can edit the PATH by loading some elisp before the executable is dumped. I'm going to try but I can't do it right now.

Great! Would that remove /usr/local/Homebrew/Library/Homebrew/shims/mac/super?

CamJN · 2020-09-02T01:48:45Z

@aconchillo that is the plan, yes.

aconchillo · 2020-09-02T07:38:50Z

I was curious and tried combinations of the following (this is just a test):

;; Remove Homebrew shims path from PATH since Homebrew doesn't
;; like having hardcoded paths to the Homebrew directories.
;; See https://github.com/Homebrew/homebrew-core/pull/59448#issue-465829006
(let ((env-path (getenv "PATH")))
  (message "Before PATH is %s" env-path)
  (message "Process environment is %s"
           (mapconcat 'identity initial-environment " "))
  (if env-path
      (setenv "PATH"
              (mapconcat 'identity
                         (delete "/usr/local/Homebrew/Library/Homebrew/shims/mac/super"
                                 (split-string env-path ":"))
                         ":")))
  (message "After PATH is %s" (getenv "PATH")))

in different parts of loadup.el without success. loadup.el is where emacs.pdmp is created, but (getenv "PATH") returns nil. I'll investigate more.

CamJN · 2020-09-02T15:13:24Z

Rather than edit loadup.el, you want to add a site-load.el to the lisp/ dir. And yes, the getenv is where I last had issues.

CamJN · 2020-09-03T16:31:28Z

Ok, I got rid of the reference to the homebrew shims dir, someone who is up to date on what homebrew wants stylistically should chime in on how they want this file added.

Formula/emacs.rb

aconchillo · 2020-09-03T18:23:51Z

@CamJN Oh... so it was exec-path. Nice!

Formula/emacs.rb

CamJN · 2020-09-03T19:47:17Z

Sweet, all checks passed. Do I need to squash or can you guys use the squash and merge button?

SMillerDev · 2020-09-03T19:48:17Z

Not without breaking things. Please squash the commits.

CamJN · 2020-09-03T19:51:13Z

Done

CamJN · 2020-09-04T00:57:09Z

agda timed out on 10.15 but the formula change is the same that passed everything last time.

CamJN · 2020-09-04T15:50:47Z

@SMillerDev is there a way to run just the failed test again? Or kick off all the tests? Not sure what to do about the agda test since that was supposed to be fixed and I rebased on master so I should have that commit included.

CamJN · 2020-09-04T15:55:58Z

@chenrui333 any idea how to speed up the agda tests? They're timing out: https://github.com/Homebrew/homebrew-core/pull/59448/checks?check_run_id=1070317115#step:5:699

aconchillo · 2020-09-04T16:29:49Z

@SMillerDev is there a way to run just the failed test again? Or kick off all the tests? Not sure what to do about the agda test since that was supposed to be fixed and I rebased on master so I should have that commit included.

Rebasing and pushing again should kick off all tests.

CamJN · 2020-09-06T16:35:09Z

All checks passed again.

SMillerDev

Thanks @CamJN ! Without contributions like yours it'd be impossible to keep homebrew going with the high standards that users have come to expect from the project. You can feel good knowing that you've made the world a tiny bit better for homebrew users around the world! 👍 🎉

BrewTestBot · 2020-09-06T18:14:57Z

🤖 A scheduled task has triggered a merge.

BrewTestBot added the deprecated license Formula uses a deprecated SPDX license which should be updated label Aug 11, 2020

chenrui333 removed the deprecated license Formula uses a deprecated SPDX license which should be updated label Aug 11, 2020

CamJN force-pushed the master branch from ac78fd3 to 2b694cd Compare August 23, 2020 18:07

CamJN force-pushed the master branch from 2b694cd to 2b8600a Compare September 3, 2020 16:30

SMillerDev reviewed Sep 3, 2020

View reviewed changes

Formula/emacs.rb Outdated Show resolved Hide resolved

Formula/emacs.rb Show resolved Hide resolved

aconchillo reviewed Sep 3, 2020

View reviewed changes

Formula/emacs.rb Outdated Show resolved Hide resolved

CamJN force-pushed the master branch from 6b888e7 to 626a37a Compare September 3, 2020 19:14

CamJN force-pushed the master branch from 626a37a to 4978567 Compare September 3, 2020 19:50

emacs 27.1

719ad3c

CamJN force-pushed the master branch from 4978567 to 719ad3c Compare September 4, 2020 16:31

SMillerDev approved these changes Sep 6, 2020

View reviewed changes

BrewTestBot closed this in 451f028 Sep 6, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

emacs 27.1 #59448

emacs 27.1 #59448

CamJN commented Aug 11, 2020

miccal commented Aug 11, 2020

aconchillo commented Aug 14, 2020

CamJN commented Aug 14, 2020

aconchillo commented Aug 15, 2020

gromgit commented Aug 15, 2020

aconchillo commented Aug 15, 2020

aconchillo commented Aug 17, 2020

chenrui333 commented Aug 23, 2020

CamJN commented Aug 23, 2020

CamJN commented Aug 24, 2020 •

edited

CamJN commented Aug 24, 2020

CamJN commented Aug 24, 2020

SMillerDev commented Aug 24, 2020

CamJN commented Aug 24, 2020 •

edited

CamJN commented Aug 24, 2020

aconchillo commented Sep 2, 2020 •

edited

CamJN commented Sep 2, 2020

aconchillo commented Sep 2, 2020

CamJN commented Sep 2, 2020

aconchillo commented Sep 2, 2020

CamJN commented Sep 2, 2020

CamJN commented Sep 3, 2020

aconchillo commented Sep 3, 2020

CamJN commented Sep 3, 2020

SMillerDev commented Sep 3, 2020

CamJN commented Sep 3, 2020

CamJN commented Sep 4, 2020

CamJN commented Sep 4, 2020

CamJN commented Sep 4, 2020

aconchillo commented Sep 4, 2020

CamJN commented Sep 6, 2020

SMillerDev left a comment

BrewTestBot commented Sep 6, 2020

emacs 27.1 #59448

emacs 27.1 #59448

Conversation

CamJN commented Aug 11, 2020

miccal commented Aug 11, 2020

aconchillo commented Aug 14, 2020

CamJN commented Aug 14, 2020

aconchillo commented Aug 15, 2020

gromgit commented Aug 15, 2020

aconchillo commented Aug 15, 2020

aconchillo commented Aug 17, 2020

chenrui333 commented Aug 23, 2020

CamJN commented Aug 23, 2020

CamJN commented Aug 24, 2020 • edited

CamJN commented Aug 24, 2020

CamJN commented Aug 24, 2020

SMillerDev commented Aug 24, 2020

CamJN commented Aug 24, 2020 • edited

CamJN commented Aug 24, 2020

aconchillo commented Sep 2, 2020 • edited

CamJN commented Sep 2, 2020

aconchillo commented Sep 2, 2020

CamJN commented Sep 2, 2020

aconchillo commented Sep 2, 2020

CamJN commented Sep 2, 2020

CamJN commented Sep 3, 2020

aconchillo commented Sep 3, 2020

CamJN commented Sep 3, 2020

SMillerDev commented Sep 3, 2020

CamJN commented Sep 3, 2020

CamJN commented Sep 4, 2020

CamJN commented Sep 4, 2020

CamJN commented Sep 4, 2020

aconchillo commented Sep 4, 2020

CamJN commented Sep 6, 2020

SMillerDev left a comment

Choose a reason for hiding this comment

BrewTestBot commented Sep 6, 2020

CamJN commented Aug 24, 2020 •

edited

CamJN commented Aug 24, 2020 •

edited

aconchillo commented Sep 2, 2020 •

edited