ybiquitous Github contribution chart
ybiquitous Github Stats
ybiquitous Most Used Languages

Activity

29 Sep 2022

Ybiquitous

Refactor writeOutputFile() to give func name and use async/await

Pushed On 29 Sep 2022 at 10:18:06

Ybiquitous

Refactor to add function names

Which issue, if any, is this issue related to?

None.

Is there anything in the PR that needs further explanation?

Giving function names (avoiding anonymous) helps debugging.

This refactoring relies on the ESLint func-names rule.

Also, it seems better to add the rule to our eslint-config-stylelint config.

Merged On 29 Sep 2022 at 10:09:14

Ybiquitous

Commented On 29 Sep 2022 at 10:09:14

Ybiquitous

Refactor to add function names

Which issue, if any, is this issue related to?

None.

Is there anything in the PR that needs further explanation?

Giving function names (avoiding anonymous) helps debugging.

This refactoring relies on the ESLint func-names rule.

Also, it seems better to add the rule to our eslint-config-stylelint config.

Forked On 29 Sep 2022 at 10:08:21

Ybiquitous

[note] I've simplified the logic to avoid using the same local variable name as the function name.
On 29 Sep 2022 at 10:08:21

Ybiquitous

Refactor to add function names

Which issue, if any, is this issue related to?

None.

Is there anything in the PR that needs further explanation?

Giving function names (avoiding anonymous) helps debugging.

This refactoring relies on the ESLint func-names rule.

Also, it seems better to add the rule to our eslint-config-stylelint config.

Merged On 29 Sep 2022 at 10:09:15

Ybiquitous

Commented On 29 Sep 2022 at 10:09:15

Ybiquitous

Refactor to add function names

Created On 29 Sep 2022 at 10:06:38
Create Branch
Ybiquitous In stylelint/stylelint Create Branchadd-func-names

Ybiquitous

A mighty, modern linter that helps you avoid errors and enforce conventions in your styles.

On 29 Sep 2022 at 10:04:57
Issue Comment

Ybiquitous

Account for potentially breaking `value-no-vendor-prefix` behavior with `display: -webkit-box;`

What is the problem you're trying to solve?

I recently applied a fairly straightforward stylelint config to a legacy project and found that several instances of display: -webkit-box; got changed to display: box;, as I would expect with the value-no-vendor-prefix rule enabled.

However, to my surprise not only is there not an equivalent non-prefixed display value available, display: -webkit-box; and the related -webkit-line-clamp property have been added to the CSS Overflow Module 3 spec in that same vendor-prefixed syntax. As far as I can tell, display: -webkit-box; is the only valid way (even in non-webkit browsers) to produce the line -webkit-line-clamp behavior at the time of writing.

I was able to avoid this autofixing by adding

"value-no-vendor-prefix": [
    true,
    {
        "ignoreValues": ["box"]
    }
] 

to my .stylelintrc.json, as the docs correctly suggest.

What solution would you like to see?

Have "ignoreValues": ["box"] be the default value of value-no-vendor-prefix

Forked On 29 Sep 2022 at 08:51:51

Ybiquitous

If people want to ignore more values in addition to the default one, they will need to include it like this:

{
  "ignoreValues": [
    "box", // default
    "other",
    "values"
  ]
} 

this way may take some doing, but indeed, it is more flexible than the latter way; "ignore -webkit-box regardless of ignoreValues".

I agree with the way "change the default value of ignoreValues". 👍🏼

Commented On 29 Sep 2022 at 08:51:51
Issue Comment

Ybiquitous

Account for potentially breaking `value-no-vendor-prefix` behavior with `display: -webkit-box;`

What is the problem you're trying to solve?

I recently applied a fairly straightforward stylelint config to a legacy project and found that several instances of display: -webkit-box; got changed to display: box;, as I would expect with the value-no-vendor-prefix rule enabled.

However, to my surprise not only is there not an equivalent non-prefixed display value available, display: -webkit-box; and the related -webkit-line-clamp property have been added to the CSS Overflow Module 3 spec in that same vendor-prefixed syntax. As far as I can tell, display: -webkit-box; is the only valid way (even in non-webkit browsers) to produce the line -webkit-line-clamp behavior at the time of writing.

I was able to avoid this autofixing by adding

"value-no-vendor-prefix": [
    true,
    {
        "ignoreValues": ["box"]
    }
] 

to my .stylelintrc.json, as the docs correctly suggest.

What solution would you like to see?

Have "ignoreValues": ["box"] be the default value of value-no-vendor-prefix

Forked On 29 Sep 2022 at 08:31:40

Ybiquitous

I see. The next problem seems to be how to change the default behavior:

  • change the default value of ignoreValues as proposed
  • or, ignore -webkit-box regardless of ignoreValues

Commented On 29 Sep 2022 at 08:31:40
Issue Comment

Ybiquitous

Automate GitHub Release

Which issue, if any, is this issue related to?

Ref: https://github.com/stylelint/stylelint/issues/6343#issuecomment-1250353973 Ref: #6253

Is there anything in the PR that needs further explanation?

I believe this PR can omit the following step of releasing:

https://github.com/stylelint/stylelint/blob/eee9deb35190f46c5f964769d521fef5d7d1d7a8/docs/maintainer-guide/releases.md?plain=1#L16

This is not just a copy-and-paste of CHANGELOG.md. Instead, it rewrites user links to user mentions (e.g. @ybiquitous) so that the "Contributors" section should be displayed on a release page like this.

But, one downside. This PR increase complexity: a remark script and a GitHub action. It may be too much for this automation. I'll close the PR if anyone expresses a concern.

Forked On 29 Sep 2022 at 08:26:11

Ybiquitous

If this new workflow is not as expected in the next release, let's recover it manually.

Commented On 29 Sep 2022 at 08:26:11

Ybiquitous

Automate GitHub Release (#6371)

Pushed On 29 Sep 2022 at 08:23:16

Ybiquitous

Automate GitHub Release

Created On 29 Sep 2022 at 08:23:16
Issue Comment

Ybiquitous

Automate GitHub Release

Which issue, if any, is this issue related to?

Ref: https://github.com/stylelint/stylelint/issues/6343#issuecomment-1250353973 Ref: #6253

Is there anything in the PR that needs further explanation?

I believe this PR can omit the following step of releasing:

https://github.com/stylelint/stylelint/blob/eee9deb35190f46c5f964769d521fef5d7d1d7a8/docs/maintainer-guide/releases.md?plain=1#L16

This is not just a copy-and-paste of CHANGELOG.md. Instead, it rewrites user links to user mentions (e.g. @ybiquitous) so that the "Contributors" section should be displayed on a release page like this.

But, one downside. This PR increase complexity: a remark script and a GitHub action. It may be too much for this automation. I'll close the PR if anyone expresses a concern.

Forked On 29 Sep 2022 at 08:16:50

Ybiquitous

@jeddy3 Thanks for the feedback! I've updated the releasing guide via fddc0da3b.

Commented On 29 Sep 2022 at 08:16:50

Ybiquitous

Remove "Copy and paste ..." from releasing guide

Pushed On 29 Sep 2022 at 08:14:36

Ybiquitous

Fix unit-no-unknown false positives for font-relative length units (#6374)

See https://w3c.github.io/csswg-drafts/css-values-4/#font-relative-lengths

Pushed On 29 Sep 2022 at 08:10:56

Ybiquitous

Fix `unit-no-unknown` false positives for font-relative length units

Created On 29 Sep 2022 at 08:10:54
Issue Comment

Ybiquitous

Fix `unit-no-unknown` false positives for font-relative length units

Which issue, if any, is this issue related to?

Closes #6373

Is there anything in the PR that needs further explanation?

I've updated the unit list based on the "CSS Values and Units Module Level 4" spec.

Forked On 29 Sep 2022 at 08:10:48

Ybiquitous

Thanks for the quick review!

Commented On 29 Sep 2022 at 08:10:48
Issue Comment

Ybiquitous

Use automation for a changelog and release management

Describe the documentation issue

With https://github.com/changesets/changesets/tree/main/packages/changelog-github, we'll have thanks mentions for contributors in our changelog and releases.

What solution would you like to see?

Enable changesets


Alternative, manage thanks mentions manually.

Forked On 29 Sep 2022 at 08:07:09

Ybiquitous

  • during a release, we're in the terminal anyway to perform other updates (e.g., to the website and demo)

That makes sense. I have no objections. 👍🏼

Commented On 29 Sep 2022 at 08:07:09
Issue Comment

Ybiquitous

Use automation for a changelog and release management

Describe the documentation issue

With https://github.com/changesets/changesets/tree/main/packages/changelog-github, we'll have thanks mentions for contributors in our changelog and releases.

What solution would you like to see?

Enable changesets


Alternative, manage thanks mentions manually.

Forked On 29 Sep 2022 at 07:10:18

Ybiquitous

@JounQin Thanks for the explanation. I understand well. 👍🏼

I'd like to hear @jeddy3's idea.

Commented On 29 Sep 2022 at 07:10:18
Issue Comment

Ybiquitous

Use automation for a changelog and release management

Describe the documentation issue

With https://github.com/changesets/changesets/tree/main/packages/changelog-github, we'll have thanks mentions for contributors in our changelog and releases.

What solution would you like to see?

Enable changesets


Alternative, manage thanks mentions manually.

Forked On 29 Sep 2022 at 07:01:44

Ybiquitous

I'm not sure to understand this part, if the CI job fails, we can just retry it and when the npm version is not published, changesets will still recognize it and try to release again

I'm worried about a situation where tagging is successful but publishing is failed. In such a case, it seems harder to recover than manually releasing. Although, it should be rare.

Commented On 29 Sep 2022 at 07:01:44

Ybiquitous

Fix `unit-no-unknown` false positives for font-relative length units

Which issue, if any, is this issue related to?

Closes #6373

Is there anything in the PR that needs further explanation?

I've updated the unit list based on the "CSS Values and Units Module Level 4" spec.

Forked On 29 Sep 2022 at 06:31:57

Ybiquitous

Sort alphabetically.
On 29 Sep 2022 at 06:31:57

Ybiquitous

Fix `unit-no-unknown` false positives for font-relative length units

Which issue, if any, is this issue related to?

Closes #6373

Is there anything in the PR that needs further explanation?

I've updated the unit list based on the "CSS Values and Units Module Level 4" spec.

Merged On 29 Sep 2022 at 06:31:57

Ybiquitous

Commented On 29 Sep 2022 at 06:31:57

Ybiquitous

Fix `unit-no-unknown` false positives for font-relative length units

Created On 29 Sep 2022 at 06:31:21
Create Branch
Ybiquitous In stylelint/stylelint Create Branchupdate-font-relative-length-units

Ybiquitous

A mighty, modern linter that helps you avoid errors and enforce conventions in your styles.

On 29 Sep 2022 at 06:30:00
Issue Comment

Ybiquitous

Account for potentially breaking `value-no-vendor-prefix` behavior with `display: -webkit-box;`

What is the problem you're trying to solve?

I recently applied a fairly straightforward stylelint config to a legacy project and found that several instances of display: -webkit-box; got changed to display: box;, as I would expect with the value-no-vendor-prefix rule enabled.

However, to my surprise not only is there not an equivalent non-prefixed display value available, display: -webkit-box; and the related -webkit-line-clamp property have been added to the CSS Overflow Module 3 spec in that same vendor-prefixed syntax. As far as I can tell, display: -webkit-box; is the only valid way (even in non-webkit browsers) to produce the line -webkit-line-clamp behavior at the time of writing.

I was able to avoid this autofixing by adding

"value-no-vendor-prefix": [
    true,
    {
        "ignoreValues": ["box"]
    }
] 

to my .stylelintrc.json, as the docs correctly suggest.

What solution would you like to see?

Have "ignoreValues": ["box"] be the default value of value-no-vendor-prefix

Forked On 29 Sep 2022 at 12:54:12

Ybiquitous

@CurrrBell Thanks for using the template.

This issue seems related to not only value-no-vendor-prefix but also property-no-vendor-prefix because the -webkit-line-clamp property has also been standardized for legacy compatibility.

I'm concerned that giving built-in rules some exceptions would lead to less flexibility. So it looks better to me to change stylelint-config-standard instead. What do you think?

Commented On 29 Sep 2022 at 12:54:12
Issue Comment

Ybiquitous

Use automation for a changelog and release management

Describe the documentation issue

With https://github.com/changesets/changesets/tree/main/packages/changelog-github, we'll have thanks mentions for contributors in our changelog and releases.

What solution would you like to see?

Enable changesets


Alternative, manage thanks mentions manually.

Forked On 29 Sep 2022 at 12:17:44

Ybiquitous

@JounQin Thanks for the answer.

generate GitHub release (But the release content could be different with our format)

Regarding this, it seems possible to prevent generating a GitHub release via the createGithubReleases: false option of changesets/action.

It sounds good to me to publish on CI for easier workflow, but perhaps it may be tough to recover if a publishing CI job would fail. As of now, we don't have a problem with continuing to use np manually.

Commented On 29 Sep 2022 at 12:17:44

Ybiquitous

Make preparing GitHub Release easier

Which issue, if any, is this issue related to?

Ref: https://github.com/stylelint/stylelint/issues/6343#issuecomment-1250353973 Ref: #6253

Is there anything in the PR that needs further explanation?

I believe this PR can omit the following step of releasing:

https://github.com/stylelint/stylelint/blob/eee9deb35190f46c5f964769d521fef5d7d1d7a8/docs/maintainer-guide/releases.md?plain=1#L16

This is not just a copy-and-paste of CHANGELOG.md. Instead, it rewrites user links to user mentions (e.g. @ybiquitous) so that the "Contributors" section should be displayed on a release page like this.

But, one downside. This PR increase complexity: a remark script and a GitHub action. It may be too much for this automation. I'll close the PR if anyone expresses a concern.

Forked On 28 Sep 2022 at 04:20:15

Ybiquitous

See also https://cli.github.com/manual/gh_release_create
On 28 Sep 2022 at 04:20:15