Remove some `not future css` plugins from next major release

[Semigradsky]

1. postcss-color-rgba-fallback
   rgba(153, 221, 153, 0.8) -> #99DD99
   http://caniuse.com/#feat=css3-colors (IE 8-)

2. postcss-pseudoelements
   :before {} -> ::before {}
   http://caniuse.com/#feat=css-gencontent (IE 8-)

3. pixrem
   2.5rem -> 80px
   http://caniuse.com/#feat=rem (IE 10-)

Good plugin pack for this plugins: https://github.com/jonathantneal/oldie

[MoOx]

Why not. Curious to know what people want.

https://twitter.com/cssnext/status/861801417719382016

[jonathantneal]

I did not get to participate in this poll, but I would have voted in favor of removing these not future css plugins. Congrats on getting so many participants. It looks like a large majority wanted them gone.

[jonathantneal]

Also, you might want to remove pleeease-filters from #374 (comment). Firefox has supported gradients since early 2015 in version 35 (we are on 53).

From the pleeease-filters docs:

Firefox support SVG filters (and CSS shorthand since FF35)

Confirmed by caniuse.

[Semigradsky]

Firefox has supported gradients since early 2015 in version 35 (we are on 53).

How about IE 11 support?

[timkelty]

The future of @apply is also not looking so good: http://www.xanthir.com/b4o00
I'd consider removing postcss-apply.

[jonathantneal]

I’ve pinged the csswg or anyone who can spare time to update this db with accurate information.

https://twitter.com/jon_neal/status/863108278666047488

If we get that repo in order as a source of truth, we can very intelligently update the feature set in cssnext, and even add stages/env in the future.

[jonathantneal]

How about IE 11 support?

IE does not support SVGs filters.

[Jessidhia]

I'd also suggest removing color from postcss-cssnext. Supporting the current css-color-function plugin not only (incorrectly) implements a spec that is considered dead by its author, but also prevents adding support for a possible real transform for the actual color function.

See comments by tabatkins on ianstormtaylor/css-color-function#35.

[jonathantneal]

If the name is locked to color-mod so that it matches the current spec, then I don’t see a problem in keeping it around.