It looks like
brew has removed support for the
--with-default-names argument. So when trying to install GNU
sed Mac users will get
Error: invalid option: --with-default-names
There are a bunch of commits from January like this one -
https://github.com/Homebrew/homebrew-core/commit/43f2e19d93f04eddf24ec19187b3648f594b80e2
Can see that they've removed support for the option, and now print a caveat on install instructing the user to edit
bashrc if they want the programs to run from their original names (i.e.
sed instead of
gsed).
I'm not sure that's particularly an improvement in user experience, but it is what it is.
Need to update the
README to detail the new process. Did toy with the idea of including a test in the script to see if
gsed exists (and use that if it does), but I don't really like the idea of having hacks in there for the sake of one (albeit popular) OS.
In Github #30 (where this was identified), user has provided the output from running `brew install grep`
brew install grep
==> Downloading https://homebrew.bintray.com/bottles/grep-3.3.mojave.bottle.2.tar.gz
######################################################################## 100.0%
==> Pouring grep-3.3.mojave.bottle.2.tar.gz
==> Caveats
All commands have been installed with the prefix "g".
If you need to use these commands with their normal names, you
can add a "gnubin" directory to your PATH from your bashrc like:
PATH="/usr/local/opt/grep/libexec/gnubin:$PATH"
That path can differ depending on flags passed in, but that should be the default.
Also need to find out whether this post (
https://scriptingosx.com/2017/04/about-bash_profile-and-bashrc-on-macos/) about Macs being a bit stupid when it comes to
bashrc and
bash_profile is still accurate.
Activity
2019-03-26 17:21:21
Webhook User-Agent
View Commit
2019-03-26 17:25:21
Webhook User-Agent
View Commit
2019-03-26 17:29:45
There's a test encode running at the moment (seemingly successfully, but time will tell).
I'm going to update the README to remove the
2019-03-26 17:33:21
Webhook User-Agent
View Commit
2019-03-26 17:39:46
Have asked the user who originally reported issues to try pulling the latest revision to run on their Mac. Other than creating aliases I don't have a good way to test how a Mac will behave
2019-03-28 16:09:00
The user has provided output which shows the test should work (in the sense that
I wondered if perhaps it doesn't like the ternary operator used on that line in the script, so have asked for the output of the following
If exit statuses are set correctly then it'd be worth adjusting to use those. If not, might need to revisit the original decision to include logic in the script to handle it. Sodding Macs.
2019-03-28 17:52:21
Webhook User-Agent
View Commit
2019-03-29 09:02:02
2019-03-29 09:54:36
The key line isn't being written into the manifest, but there's also no error being thrown to
and that's not present here. Suggests
2019-03-29 11:12:21
Webhook User-Agent
View Commit
2019-03-29 11:13:12
First run yielded some errors not visible in the video - I accidentally set the wrong variable for the
Test run following the commit to fix that resulted in the key being inserted correctly.
2019-03-29 11:39:32
2019-03-29 11:49:39
2019-03-29 11:56:21
Webhook User-Agent
View Commit
2019-03-29 13:09:13
2019-03-29 13:09:13
2019-03-29 13:09:17