Re: Some feedback from Chemistry CG

They don’t want us to do any conversion of chemical symbols to the chemical names.  That doesn’t mean that a 3rd party plug-in / feature of their screen reader could add this functionality.  But without knowing the semantics that this is a chemical equation that won’t be possible.  So I think knowing the semantics is still very important.  They just don’t was MathML to dictate how things will be spoken.


Thanks
EOM
Charles LaPierre
Technical Lead, DIAGRAM and Born Accessible
Imageshare Product Manager
Twitter: @CLaPierreA11Y
Skype: charles_lapierre


On Sep 14, 2020, at 2:44 PM, David Farmer <farmer@aimath.org<mailto:farmer@aimath.org>> wrote:


Is it possible that the Chemistry people completely missed the point
that we are asking about encoding the meaning of what appears in
the MathML?

I can see their comments are making sense in the context of advising
Neil how assistive technology might pronounce "H_2 O".  I do not
see a reasonable argument against marking up the meaning of the
symbols in a chemical formula.  And I can see a lot of reasons for wanting
to do so.

I can also imagine a hierarchy of ways of pronouncing chemical formulas,
just like there is for reading mathematics.  At one extreme is
what was written in the shared email:  “cap C cap H sub 4” for methane.
Next might be "C H 4".  After that might be "methane".  And just like
in math, the user gets to choose, if the source has been properly constructed.

You are reduced to guessing, the exact thing we are trying to avoid,
if the meaning is not encoded.

On Mon, 14 Sep 2020, Neil Soiffer wrote:

Since subscripts and superscripts should not be read in a default way, at least msub/msup need to be tagged in
some way if the base isn't tagged. Because something needs tagging, I do think Chemistry needs to be in level 1.
    Neil
On Mon, Sep 14, 2020 at 1:50 PM Deyan Ginev <deyan.ginev@gmail.com<mailto:deyan.ginev@gmail.com>> wrote:
     Thanks for forwarding Neil!

     It is indeed curious that they are basically asking for a
     presentation-tree readout, no need for semantics at all -- even the
     subscripts are read as-is. If we are to agree with that feedback
     unconditionally, one wonders if we should be excluding chemistry from
     level 1 entirely, as it won't have any material impact compared to raw
     pMML, and it may get the acceptable level of speech without any
     additional annotation work.

     Greetings,
     Deyan

     On Mon, Sep 14, 2020 at 1:07 PM Neil Soiffer <soiffer@alum.mit.edu<mailto:soiffer@alum.mit.edu>> wrote:
     >
     >
     > With Cary's permission, I am forwarding this message from the Chemistry CG to the MathML mailing
     list.
     >
     > ---------- Forwarded message ---------
     > From: Supalo, Cary A <csupalo@ets.org<mailto:csupalo@ets.org>>
     > Date: Mon, Sep 7, 2020 at 4:09 PM
     > Subject: RE: Request for phone meeting
     > To: Neil Soiffer <soiffer@alum.mit.edu<mailto:soiffer@alum.mit.edu>>
     >
     >
     >
     >
     >
     >
     >
     >
     > Dear Neil,
     >
     >
     >
     > A small group of chemistry community members met informally last week to discuss your request more
     at length. The following bulleted list of points are what we perceive as a starting point for a
     fruitful discussion on what further disambiguation of chemistry terminology is needed to support this
     collaborative effort between the MathML and the chemistry community. We feel and know that without
     the strong collaboration we have, and without the input from the MathML community, it would be a
     detriment to the print disabled communities. On our initial assessment of your request at our last
     meeting, we have three examples of conventions we would like to see fully implemented:
     >
     >
     >
     >
     >
     > We feel elements and compounds should be read as letters, including the designation of the capital
     letter. Sodium (Na), for example, should be read “cap N a”. Chlorine (Cl) as “Cap C l”. Sodium
     chloride “cap N a cap C l”. Methane (CH4) as “cap C cap H sub 4”.
     >
     >
     >
     > With regards to the question regarding elements and compounds being read with their proper names we
     feel is a pedagogical question that should be left up to the individual user to decide. If they wish
     proper names to be spoken by their screen reader, this can be enabled by means of the custom language
     dictionary that is offered by JAWS and NVDA.
     >
     > We also feel strongly that units should be read as units. We imagine MathML probably already has
     many unit designations already that we can leverage. Chemistry content should be able to benefit from
     the same unit designations.
     >
     >
     >
     > Anything that is not disambiguated/defined in our table should also be read “as is,” as the letters
     they are defined.
     >
     >
     >
     > We hope this clarification is the first step to a lengthier discussion on optimization of
     disambiguation of chemistry content. It was our position all along that the table we provided to the
     MathML group was a first step. We welcome the opportunity to meet with you to discuss logical next
     steps at future meetings. We welcome and highly value this collaboration between the chemistry and
     math ML communities. If you would like to include this type of topic as an agenda item for a future
     Chemistry Community meeting, we are certainly happy to do so.
     >
     >
     >
     > Thank you very much for your continued support of our community where we value your input and
     feedback highly.
     >
     > Cary
     >
     >
     >
     >
     >
     >
     >
     > From: Neil Soiffer <soiffer@alum.mit.edu<mailto:soiffer@alum.mit.edu>>
     > Sent: Sunday, August 30, 2020 12:35 AM
     > To: Supalo, Cary A <csupalo@ets.org<mailto:csupalo@ets.org>>
     > Cc: Barrett, Dan <Dan.Barrett@hmhco.com<mailto:Dan.Barrett@hmhco.com>>
     > Subject: Re: Request for phone meeting
     >
     >
     >
     > The MathML CG talked about chemistry issues a little this week in the context of a larger issue
     that we are trying to resolve. As I mentioned, a subject area sets defaults, so most authors don't
     need to worry about labelling every token. But there are defaults, so every token has a meaning given
     to it by an attr (currently being called "semantics" but likely will change).  For chemistry and for
     that matter, units, the issue came up: how detailed are the attr values? For example, do we have
     semantics="units" or do we have semantics="millimeters", etc. For chemistry, is it
     semantics="element" or semantics="hydrogen", etc. I had thought everyone was on board with "unit" and
     "element", but I was wrong, so this issue needs to be hashed out. Because this needs resolution, it
     probably doesn't make sense to discuss it at the chem CG call this week.
     >
     >
     >
     > There is a question that you or maybe the group can answer... if I have NaCl suitably marked up in
     MathML, should that always be 'spelled out' in speech? If sometimes it should be "sodium chloride" in
     speech and maybe sometimes "salt", who determines that? The author? The reader? The answer to that
     question will inform a discussion about labelling them 'element' (which would allow for various forms
     of speech) vs. specifically labelling them 'hydrogen' or for that matter not labeling them at all so
     that the letters are always just the letters.
     >
     >
     >
     >     Neil
     >
     >