Closed fregante closed 4 years ago
I’m kind of hoping to restore Firefox support in v6 before releasing this in v7, sadly we depend on another package to fix that. The only way to expedite this is for me to publish a temporary fork of browser-launcher
and use it as a dependency in #149. What do you think?
Btw, I created https://github.com/juliangruber/just-launch as a replacement for browser-launcher, but haven't gotten around rewriting this module on top of it. Maybe you could pick that up?
@juliangruber that's a nice little module. I think it would still be a breaking release though. I'd rather make v6 first for existing users before moving onto the next version.
What if we release v6.0.1 with:
And then later publish v7 with:
Drop PhantomJS support
just-launch
In short:
git reset b46c09c --hard
undoes this PR locallygit push --force
undoes the PR on this reponpm version 6.0.1
npm publish
git pick d353cb5
so this PR is applied on topgit push
to push everything to the repoThen, when just-launch
is added (or whatever the next release will be) you can publish it as v7
@juliangruber that's a nice little module. I think it would still be a breaking release though. I'd rather make v6 first for existing users before moving onto the next version.
I agree that it would be a breaking change and let's finish other work first.
What if we release v6.0.1 with:
Left a change request, otherwise :+1:
In short:
Why do you want to undo this PR? And please don't force push on a shared repo.
I felt it would be easier to explain like that, but the point is releasing 6.0.1 without the PhantomJS commit. How you do that is up to you
Why don't we just release 7.0.0 with the phantomjs commit and continue from there? It has landed anyway, people should be ok with upgrading.
Because it fixes existing installation and doesn't require users to update (nor you'll have to release a new breaking release of tape-run
)
Multiple release channels 🤷♂ Removing phantomjs is good, let's just let them upgrade. I don't want to go through the work of setting up a multi release channel process yet
It’s really not that hard. Let’s not complicate things. It’s a single push to fix existing installations, not a continued thing.
I’m just suggesting this to make it easy for everyone who uses this module. If publishing a single release is too hard to justify, then release as you wish
7.0.0
is out 🎆
PhantomJS was discontinued in March 2018; the last release was in 2016.
It's currently breaking the tests for unknown reasons and it's not worth investigating.
When merged, this will be released as a breaking.