-
-
Notifications
You must be signed in to change notification settings - Fork 4.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
1.7.1 Release #2887
Comments
Sgtm |
Notice that we should uncomment 68d1b56 before releasing.
|
Any objections to merging #2844 before 1.7.1? |
To clarify... #2846 would be a breaking change, too. I'm concerned about |
By the way, I just want to mention that I'm heads down on my new team and won't have the time to be as involved in the next few months. I just clicked the "Unwatch" button, please cc me if anything needs my attention. I trust @azz, @lydell, @suchipi and co to make the project move forward in a good way :) |
Hi guys, What are your thoughts around including the This could give users of editor integrations a bit more context on why things may break, which is important since we do CLI args validations. The current model of CLI args validation forces any editor integration that rely on the cli to do major releases (breaking changes) since by passing the new args it completely break compatibility with older prettier. The reason I raised this discussion in here is due to the work that has been done around validation of CLI args to 1.7.1 |
@mitermayer Could you give an example? Prettier only warns about unknown options, so passing new options to an old Prettier should not be a problem? What work around the CLI are you talking about specifically? |
We didn't add additional validations for CLI, just add missing validations for configs from New flags will be always ignored and show warnings, so it shouldn't be a problem, the only breakings will be new choices since the old version didn't know what to do with it, e.g. |
The problem happens because, when prettier release new features we want to add support to it on editors. We do that by passing extra arguments to the cli. And the most common form of editor integration is by leveraging "stdin" and parsing the values of "stdout". The error 'unknown arg blah' can cause things like this: prettier/vim-prettier#53 Another possible solution is if there was a way to mute this errors from being printed when executing the CLI |
Maybe when an error occurs, you could run |
Unless there's something I don't understand, I'm not sure what the difference between that and including the version number in the error messages would be. |
@mitermayer Can't you hide unwanted warnings? |
@suchipi, running prettier again to get the version is an option but there are some problems with it lets imagine this scenario
The above logic is not ideal for editor integrations for 2 reasons, have an extra version call to the client (which could in theory also fail) and also means that editor integrations would have to keep up to date on how prettier spits errors on the stdout in order to parse it correctly. (we already do that by handling syntax errors, having other types of errors like that just increase complexity) My suggestion is if we could have either:
or
What are your thouhts around it ? |
The warnings/errors are always printed to stderr, only valid output will be printed to stdout, and you can also catch the exit code to determine if it's an error. Or maybe we can allow |
Redirect stderr to /dev/null |
@ikatyang so you're saying you could include |
@suchipi Just like what |
@mitermayer Could we move this conversation to a new issue? I don't think it needs to block a patch/bugfix release. |
@suchipi sounds good to me! |
@suchipi Are you up for writing the change log again? Or should we look if anyone else is interested? |
I can do it. Since it's just a point release, I was thinking that something like what we did for 1.6.1 would be enough, rather than a dedicated release document (like we did for 1.6.0 and 1.7.0). But I haven't reviewed all the merged PRs yet, so if there's enough stuff, maybe we should write a larger changelog doc. My hope is to get the changelog done this evening and then publish the new release tomorrow morning. I could publish it in the evening, too, but last time around we published in the morning so that stuff wouldn't break while people were asleep; I think it'd be good to do it that way again. |
Okay, here's a proposed changelog and list of all the TODOs we left for ourselves to take care of post->1.7.0 Commits: 1.7.0...master Changelog (based on this PR search):
Prerelease action item: uncomment this
Postrelease action items (things that need to be removed/uncommented):
I was hoping to get the release out in the morning, but I have to go into the doctor in the morning, so it will probably not go out until midday. |
Hmm, actually, some of those |
Okay, sounds good 👍 I'll just uncomment the one @ikatyang mentioned then (the yml thing in the README) |
Released: c2bc33b |
I've done the postrelease stuff around postcss and the playground now, and it seems to work out. |
Cool, thank you for doing that @lydell |
There have been a lot of CSS lowercase fixes and we are starting to get duplicate reports on some of them. We should probably publish a patch release containing those fixes.
The text was updated successfully, but these errors were encountered: