@starting-style doesn't nest correctly

[mariusGundersen]

To reproduce:

Playground link

[popover]:popover-open {
  opacity: 1;
  transform: scaleX(1);

  @starting-style {
    opacity: 0;
    transform: scaleX(0);
  }
}

Current behavior:

This produces the following output, which is clearly incorrect:

[popover]:popover-open {
  opacity: 1;
  transform: scaleX(1);
}
@starting-style {
  opacity: 0;
  transform: scaleX(0);
}

Expected behavior:

It should produce the following css:

[popover]:popover-open {
  opacity: 1;
  transform: scaleX(1);

  @starting-style {
    opacity: 0;
    transform: scaleX(0);
  }
}

Environment information:

• less version: 4.2.0

This is based on the sample code from mdn: https://developer.mozilla.org/en-US/docs/Web/CSS/@starting-style

[matthew-dean]

@mariusGundersen Experimental CSS, by definition, is generally not supported by Less. That said, Less's default behavior of bubbling any at-rule that is undefined should probably be re-thought, or an escape-hatch provided. So this is more a feature request than a bug.

[rejhgadellaa]

With more and more @-rules being added (e.g. @starting-style and @position-try) and browsers supporting nested css (and thus @-rules), I think the default behavior of Less should indeed be changed. Should we move this into a separate ticket?

[rejhgadellaa]

Additional information:

@starting-style is now supported on Chromium and WebKit. Firefox has it implemented and plans to enable it soon.

[shrpne]

@starting-style is supported in Firefox now from 129 version

[mariusGundersen]

Experimental CSS, by definition, is generally not supported by Less.

Starting style is not experimental anymore

[ogdakke]

Yup, this should be supported now. Any timeline on this?

[puckowski]

I have working code for @starting-style support on top of Less 4.2.1 (4.2.1 is yet to be published to npm) on my machine. I'll get the support into my fork https://github.com/puckowski/less.js. The fork also has support for the @scope at-rule. master branch will be clear when @starting-style support is ready.

If things go well I'll prepare a PR to this repo, but no guarantee of code merge or that a release will be published.

I maintain my fork since I use it for some of my projects, so PRs are welcome there.

[puckowski]

Less.js 4.2.1 fork release with support for CSS Container Queries, Media Queries Level 4, starting-style at-rule, and scope at-rule support with all tests passing: https://github.com/puckowski/less.js/releases/tag/4.2.1-container-mq4-tr2-scope-1

I'll likely get around to submitting a PR to this repo just for @starting-style next weekend.

If you use the fork and run into an issue, feel free to open an issue on the fork.

[matthew-dean]

@puckowski v4.2.1 released - https://www.npmjs.com/package/less - it may have some pulls your fork does not.

Btw, this repo could definitely use more core contributors.

[puckowski]

@matthew-dean

Thank you for publishing Less.js 4.2.1.

I do not have access otherwise I would help, but release 4.2.1 doesn't look published here on GitHub Releases, just FYI. I know some users have asked about that in the past.

[puckowski]

PR #4289 should resolve this issue. I'll support any requested revisions until it gets merged.

[jpattishall-ebay]

Hi @puckowski, out of curiosity, will this also address @container in @media not properly nesting?

Currently I see in 4.2.1 Playground I see this:

@media only screen and (min-width: 768px) {
  @container (min-width: 500px) {
    .primary-content {
      font-size: 1rem;
    }
  }
}

results in:

@container only screen and (min-width: 768px) and (min-width: 500px) {
  .primary-content {
    font-size: 1rem;
  }
}

But I see 4.1.3 produced the expected output:

@media only screen and (min-width: 768px) {
  @container (min-width: 500px) {
    .primary-content {
      font-size: 1rem;
    }
  }
}

Thanks!

[puckowski]

@jpattishall-ebay PR #4289 does not resolve/address the issue you are describing. Thank you for your comment and for raising that separate issue. I was not previously aware of that issue.

I've taken a look at your issue on Less.js 4.2.1 and I believe I have a fix, but I need some time for further testing and revision. I'll try to prepare a separate PR by Friday 12/20/24.

I'll likely get my fix into my fork and cut a release before I file the PR, so I will notify you if you'd like to use that while waiting on the official Less.js release.