Re: [css3-writing-modes] inconsistent handling of 'Tr' codepoints in 'text-orientation'



On 10/2/13 4:26 AM, "Koji Ishii" <kojiishi@gluesoft.co.jp> wrote:

>>Unless there's stronger justification for the optional fallback scheme
>>you've described, I
>> think we should resolve to omit it entirely.
>> Omitting fallback in no way affects any form of compliance with Unicode.
>
>I can't connect previous paragraph and this paragraph. Yes, we agreed and
>spec updated to allow your proposed behavior over a year ago. Now you're
>asking to omit the Unicode-defined behavior. I do not understand how
>these two connect to each other.
>
>This is the point of discussion, right?
>
>You want to reject one optional behavior currently in the spec. James and
>I in this ML are against. Murakami-san was against as well if I remember
>correctly. Can you explain why you think we should reject the behavior?

Which email of James are you referring to?

I can't make any sense of this compliance argument *at all*, and from
recent side discussions I'm not alone. Let's recap:

1. You give implementors a choice between two options. This means they can
implement either and be conformant with css-writing-modes. Let's call them
A and B for short.

2. You deem one of these options - say A - to be 'Unicode-defined' and
'compliant'. Further, you assert that not having A in the spec makes it
somehow violate or even fork Unicode (!). If we can't specify B alone
without breaking Unicode this of course implies B does not comply with
Unicode. 

Combining #1 and #2 with your explicit claim that CSS is required to
promote and enforce Unicode compliance (however defined) we then find that
you have written a specification that tells implementors they can either:

 Implement A and comply with Unicode

OR
 Implement B which violates Unicode

…and that doing either will be conformant!

If 'Unicode compliance' were a requirement *and* implementors needed
alternatives then all these alternatives must be Unicode-compliant. We
can't both say CSS's role is to enforce some Unicode feature *and* tell
implementors it is conformant to ignore it.

Received on Wednesday, 2 October 2013 14:33:23 UTC