W3C home > Mailing lists > Public > www-style@w3.org > September 2013

Re: [filter-effects] filter different components of an element

From: Dirk Schulze <dschulze@adobe.com>
Date: Sat, 7 Sep 2013 01:03:43 -0700
To: Lea Verou <lea@verou.me>
CC: "www-style@w3.org list" <www-style@w3.org>
Message-ID: <87387AC9-5F69-4839-9831-C700E6AA0C0C@adobe.com>

On Sep 6, 2013, at 10:52 PM, Lea Verou <lea@verou.me> wrote:

> On Sep 6, 2013, at 12:23, Dirk Schulze <dschulze@adobe.com> wrote:
>> With the CSS filter() image function I see a lot of the use cases already addressed. You can filter background-image, border-image-source, mask-image, mask-image-source, list-item-image and apply the filter property on the content itself. There are still things you can not do with the filter property nor with the filter() function, but I didn't hear any request from authors so far.
>> 
>> I would be in favor for removing this issue or (if people feel strongly about it) at least delay it to the next level of the spec.
>> 
>> Any opinions?
> 
> Are there any actual use cases that filter() doesn’t cover?

filter() can just filter CSS Images. You can not filter border: solid black 1px; for example.

> 
> Although it might be worth it to design a generic mechanism for styling backgrounds, borders etc separately. Many of the FXTF specs need this: Blending modes, Compositing, filters, masking etc. If this is exposed, it’s seems unwise to approach every case separately. Perhaps we could do it by adding a bunch of pseudo-elements? But then it gets complicated to define which properties are available there…

I agree. If we want to have special mechanism like "filter the border" or "filter just the content" with a property on the element itself, we should have global, property independent mechanism to define which paint phase should be affected. This is quite a challenge especially with blending where you may want to blend different paint phases with each other. But solvable.

Greetings,
Dirk

> 
> ~Lea
Received on Saturday, 7 September 2013 08:04:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:34 UTC