Re: #backgroundColor + <image>

Let me take a closer look. It may mean we need to take an approach similar
to what we did for #opacity-{block,inline}, which weren't defined in TTML1.

On Thu, Feb 7, 2019 at 9:22 AM Pierre-Anthony Lemieux <pal@sandflow.com>
wrote:

> Hi Glenn,
>
> The specification is unclear at best: the #backgroundColor definition
> defers to TTML1, which has no concept of <image>.
>
> Unambiguous feature designators are important since profiles like IMSC use
> them to impose conformance requirements.
>
> The approach taken with #padding, which adds capabilities introduced in
> TTML2 as a separate feature designator, seems clear and explicit.
>
> Best,
>
> -- Pierre
>
>
> On Tue, Feb 5, 2019 at 8:16 PM Glenn Adams <glenn@skynav.com> wrote:
>
>> I should clarify. I think it is a bad idea to insist on combination
>> feature definitions as a general rule, even when the combination comes into
>> play in a new version that was not present in a previous version. In
>> particular, I think if an implementation claims support for #backgroundColor
>> (already) and add a claim for support for #image, then it should
>> implement support for #backgroundColor on image as a natural
>> interpretation of the specification language.
>>
>> What I hear (behind your question) is a desire to not support background
>> color on an image (and to have a feature to control this). This sounds like
>> a new use case and should be justified by making a business case for not
>> supporting background color on image when an implementation otherwise
>> claims support for #backgroundColor and #image.
>>
>> Have you made such a business case for non-support?
>>
>>
>> On Tue, Feb 5, 2019 at 8:46 PM Glenn Adams <glenn@skynav.com> wrote:
>>
>>> Feel free to file an issue. As for myself, I'm ok with the current state
>>> of affairs.
>>>
>>> On Tue, Feb 5, 2019 at 8:32 PM Pierre-Anthony Lemieux <pal@sandflow.com>
>>> wrote:
>>>
>>>> Hi Glenn,
>>>>
>>>> This would conflict with the objective of not expanding the conformance
>>>> requirements of individual features defined in TTML1.
>>>>
>>>> For instance, #padding remains unchanged from TTML1,
>>>> while #padding-version-2 includes both #padding and padding capabilities
>>>> introduced in TTML2.
>>>>
>>>> Best,
>>>>
>>>> -- Pierre
>>>>
>>>> On Tue, Feb 5, 2019 at 5:22 PM Glenn Adams <glenn@skynav.com> wrote:
>>>>
>>>>> My reading says it is #backgroundColor, which, as written (now and in
>>>>> TTML1), did not distinguish which elements it applied to, so, by virtue of
>>>>> the upgraded *applies to* in TTML2 for tts:backgroundColor, it would
>>>>> effectively apply (now) to image as well.
>>>>>
>>>>> On Tue, Feb 5, 2019 at 1:50 PM Pierre-Anthony Lemieux <
>>>>> pal@sandflow.com> wrote:
>>>>>
>>>>>> Hi all,
>>>>>>
>>>>>> Which TTML2 feature covers tts:backgroundColor when applied to
>>>>>> <image>?
>>>>>>
>>>>>> Perhaps it is included in #image?
>>>>>>
>>>>>> Best,
>>>>>>
>>>>>> -- Pierre
>>>>>>
>>>>>