chore: setup changelog #443
Conversation
| trim_trailing_whitespace = true | ||
|
|
||
| [*.{md,markdown}] | ||
| trim_trailing_whitespace = false |
There was a problem hiding this comment.
Standardized editor config options. Added as markdown files use two spaces at the end of a line for a line break, but in all other files, you want whitespace trimmed.
|
This PR brings up a larger point, v1.0.0. Technically, when used in production you are at 1.0. I think we're there and should release a 1.0 as the next release. The biggest advantage here is that breaking changes are properly noted by the major version number. I'm updating some projects with shelljs right now and I'm not sure what 0.6 to 0.7 may have broke or if it did. Lastly, we could take a note from facebook/reactjs. They recently went from |
|
LGTM. With regards to v1.0.0, semver defines v0 as a beta, so that unstable packages can be developed easily. I personally think that we need to start progressing to 1.0.0, but I have a few things I want to work out before doing so, I'll create a milestone when I'm off mobile. |
|
Big picture ShellJS discussion in #444. |
|
@nfischer we good on this PR then? Assuming you'll make script updates on another PR. |
|
LGTM, but why has this not passed CI? |
|
It did pass on my last commit. But the auto changelog commit
|
|
If that's the case, we should probably wait to merge until we can figure that out (don't want to accidentally freeze our CI) |
|
I just mean only that commit is not reporting for some reason. Tests are still running, and passing as seen here: https://travis-ci.org/shelljs/shelljs/builds/130655225 I'll try to remove that commit and re-add another one to see if that works. |
levithomason
force-pushed
the
feature/changelog
branch
from
3ffab5e
to
2bb193f
Compare
|
If this doesn't work, we can actually remove the [ ci skip ] for our setup. |
levithomason
force-pushed
the
feature/changelog
branch
from
2bb193f
to
5160f2e
Compare
|
OK, removed the ci skip so the statuses show through on the PR. Tests were still running, but they never reported back. Suppose the extra test for changelog-only commits doesn't hurt.
|
This PR adds the same changelog flow added to
shx. Also generates the first changelog.