Re: New MathML test suite

Hi Vicent.

I'm working on the testsuite corrections today.  Here is what happened
to the ones you reported.

>  - miequivalents11.mml is not well-formed : I guess the code of characters
>    is given in hex, but the 'x' is missing   &#1D400; should be 𝐀
>    for instance.

Done.

>  - mnAtoken6: rendering issue. Sould'nt the first '1' be italic in the
>    proposed rendering? If this is true, then some other tests are wrong too.

Done.

>  - mnAtoken7: the fontcolor attribute used to paint the '0' in green
>    is not defined in MathML 2.0 (it should be color)

Done.

>  - mnSscriptlevel11: I guess the second green '10000' should be in Times.

Done.  I had to change the font name to TimesRoman, since that is what
it is called on my system.  I also made the green a darker shade so it
is easier to see.

>  - what the diffrence between mo2 and mo6?

None so far as I can tell.  I deleted one and renumbered the section.

>  - mo7: rendering issue: why is there a space between "exists" and "delta"
>    and between "that" and 'f'?

WebEQ bug.  Actually, the bug is the lack of space to the left of the
<mo>'s.  MML2 says the default lspace and rspace for an operator not
in the dictionary should be 'thickmathspace'.  Fixed.

>  - moAaccent11.mml: not well-formed. Somme attribute values are not quoted.
> 
>  - moAlargeop13.mml: same problem
> 
>  - moAlrspace14.mml: ditto

Fixed.

I expect to get the corrected tests up at W3C by the end of the day.

--Robert

----------------------------------------------------------------
Robert Miner                          email: RobertM@dessci.com
Design Science, Inc.                  phone: 651-223-2883
http://www.dessci.com
----------------------------------------------------------------

Received on Wednesday, 27 December 2000 11:37:10 UTC