[Plugin] CSSValidator - Offline W3C CSS Validator

[KevinH]

Jon, please stop attacking everything that is epub3. CSS3 is valid in epub3 and rem is different than em in epubs that use different font sizes in different places as it always references the size of the font in the root (html) element.

Epub2 is now totally dead. Deprecated almost 15 years ago (first approved epub3 spec was dated, October 11, 2011). Nothing new should ever be created in epub2.

Good, spec compliant Epub3 readers are now broadly available and making epub3 backwards compatible with the old epub2 spec is possible.

This forum's love and support for a long dead specification is really misguided.

Fair warning: At some point, Sigil will eventually drop support for creating new epub2s and when asked to load one, it will eventually use integrated code (based in my ePub3-itizer plugin) to generate an epub3-only version for editing and saving.

[Doitsu]

Originally Posted by hayaku

IE, unit : “rem” is an incorrect unit : padding: 0.25rem; when I have a CSS entry like "padding: 0.25rem;".

As KevinH has already pointed out, you'll only get this message when CSS2.1 (or lower) is selected as the CSS spec.
You can test this yourself with the online validator:
https://jigsaw.w3.org/css-validator/

If you're working on epub3 books, you'll need to change the following plugin preferences:

  "css_epub2_validation_spec": "css21",
  "css_epub3_validation_spec": "css3",

For more information, see the very first topic.

[JSWolf]

Originally Posted by KevinH

Jon, please stop attacking everything that is epub3. CSS3 is valid in epub3 and rem is different than em in epubs that use different font sizes in different places as it always references the size of the font in the root (html) element.

Epub2 is now totally dead. Deprecated almost 15 years ago (first approved epub3 spec was dated, October 11, 2011). Nothing new should ever be created in epub2.

Good, spec compliant Epub3 readers are now broadly available and making epub3 backwards compatible with the old epub2 spec is possible.

This forum's love and support for a long dead specification is really misguided.

Fair warning: At some point, Sigil will eventually drop support for creating new epub2s and when asked to load one, it will eventually use integrated code (based in my ePub3-itizer plugin) to generate an epub3-only version for editing and saving.

This is not attacking ePub 3. There's a reason not to use rem.

If you are creating an ePub for sale, it could get read on a Kobo as ePub and rem will not work. It could get read on a nook or an old Sony Reader where rem does not work.

So if you want to make sure you eBook does not break us em. ePub 3 should be made as backwards compatible as possible and rem is not going to help do that.

[KevinH]

Old readers that do not support epub3 and css3 are a dying breed, just like B&W TVs, followed by Colour, followed by LCD, followed by .... Tying yourself to old dead epub2 only e-readers is not the solution. As more and more epub3s are generated, the old readers wll just get the junk pile.

Using epub3 with full css support is the present and future.

[Turtle91]

Originally Posted by JSWolf

This is not attacking ePub 3. There's a reason not to use rem.

If you are creating an ePub for sale, it could get read on a Kobo as ePub and rem will not work. It could get read on a nook or an old Sony Reader where rem does not work.

So if you want to make sure you eBook does not break us em. ePub 3 should be made as backwards compatible as possible and rem is not going to help do that.

So, Jon, you really need to understand how css coding works. If a reader/device does not support a setting it is supposed to ignore it. So, you can easily set your css to take advantage of this…

p.largetext {
     font-size:2em
     font-size:2rem
}

Will use rem because it is the ‘most recent style’. If the reader/device doesn’t support rem then it ignores it and reverts to the next ‘most recent style’ which is em.

The coder who builds the epub should code for epub 3.x/css3.x but provide fallback coding if their target market has some ancient relic devices still in their inventory.

These are VERY basic concepts.

There are more advanced techniques where you can even check if a property is supported (@supports) or if the device has specific properties (@media). Your assumption that an epub coder MUST only use ePub2/css2 to be backwards compatible is just plain incorrect.