Re: [mathml-tests] Import MathML tests for the mathvariant attribute. (#4021)

Thanks Karl, some comments:

> I haven't seen anything in MathML3 clearly requiring renderers to extend this to the Arabic Mathematical Alphabetic Symbols introduced in 6.1 and/or indefinitely to new characters added in Mathematical Alphanumeric Symbols or similarly named blocks.

Mmh, that's probably an issue in the MathML recommendation. I had already reported this in the past and Arabic characters are now mentioned in chapter 3. It seems that chapter 7 have not been updated, though.

> I assume that support for SMP characters is now sufficiently advanced that this mechanism need not be extended further.

I guess so. For Arabic characters there were already new mathvariant values for them, though.

> http://www.mathml-association.org/MathMLinHTML5/S2.html#SS3.SSS1.tab2 is nice and precise. I don't know what authority should be given to this document, but if this document describes and these tests align with the current behaviour of more than one implementation, then the tests are serving a useful purpose of testing interoperability, thanks.

As a general rule of thumb, the tests in mathml/ are going to verify this MathML implementation note, not the official MathML3 recommendation. I already discussed with @jgraham and @youennf and it seems that it is fine to do so, even if it is not an official W3C standard. The MathML4 threads on the Math WG mailing list have not done much progress so far (@kohlhase and @davidcarlisle may know more about the internal discussions and status) so we don't need to wait any further to import the interoperability tests into that W3C repository.

In this particular case of mathvariant, the implementation note just describes the core content of https://www.w3.org/TR/MathML3/chapter3.html#presm.commatt and avoids ambiguous statements from the official recommendation such as:

"Renderers have complete freedom in mapping mathematics style attributes to specific rendering properties. However, in practice, the mathematics style attribute names and values suggest obvious typographical properties, and renderers should attempt to respect these natural interpretations as far as possible. For example, it is reasonable to render a token with the mathvariant attribute set to "sans-serif" in Helvetica or Arial. However, rendering the token in a Times Roman font could be seriously misleading and should be avoided. "

"Renderers should support those combinations of character data and mathvariant values that correspond to Unicode characters, and that they can visually distinguish using available font characters. Renderers may ignore or support those combinations of character data and mathvariant values that do not correspond to an assigned Unicode code point, and authors should recognize that support for mathematical symbols that do not correspond to assigned Unicode code points may vary widely from one renderer to another."

Such statements have led to serious inconsistent implementations in the past because some mathvariant values were implemented with CSS properties. I hope it will be fixed in a future Math WG document...

View on GitHub: https://github.com/w3c/web-platform-tests/pull/4021#issuecomment-257245038

Received on Monday, 31 October 2016 09:10:12 UTC