use-argname property

[davidcarlisle]

This PR addresses issue w3c/mathml#524 if accepted a small change to the main spec to list use-argname in the abbreviated list of properties would follow.

This adds use-argname to the properties YAML file, with translations of the main core arg names handled by the existing show available languages form at the top of the page, resulting in a display like

The page is online at my fork as

https://davidcarlisle.github.io/mathml-docs/intent-core-properties/#prop-use-argname

[NSoiffer]

I don't think

Used as a self-property on an element with arg, causes the value of the arg attribute to be used as the name to introduce the argument in speech.

captures its intent. The goal for them (at least that was the goal when I created the issue) is to have a name to use when navigating. For example, fractions have numerators and denominators and when moving into the first child of an mfrac, it is better to say "in numerator" rather than "in first child". The description should mention navigation/moving around an expression. Maybe something like:

Used as a self-property on an element with arg, causes the value of the arg attribute to be spoken by AT when the user moves to that child. For core concepts, this value should override argument names that would otherwise be used by AT.

If you want to keep the fraction example, some comment should be made that they show what AT might already say. Perhaps you can think of some other words that might be used and use them (maybe ask a middle school teacher you might know ...).

You mention self-property, but that is not defined. There is self-property-list in the grammar, but no self-property.

I'm not sure the group agreed to this, so I'll ask at the meeting. Because they won't be listed anywhere, they can't be automatically translated. Perhaps my use of "should" for core concept in the description needs some qualifications.

[davidcarlisle]

to use when navigating.

ah yes, I'll adjust before the meeting tonight, it's harder to give examples of navigation in a static page but I'll take some of your wording and see what happens, probably end up with fewer examples...

You mention self-property, but that is not defined. There is self-property-list in the grammar, but no self-property.

yes although I thought it was implicit that self-property was an item of a self-property-list. That was how we clarified what "self-property" meant when it was raised in a different issue the other day. That is, it is a property given in an intent that starts with :

Because they won't be listed anywhere, they can't be automatically translated. Perhaps my use of "should" for core concept in the description needs some qualifications.

I addressed that as much as I could here (although other suggestions welcome) like concept names, if used in an "open" setting so they just appear in the expression you can't translate them, but a small-ish list of known core concepts are translatable. Similarly here a (very) small list of known arg names are listed and translatable (with English French and German examples). We could (perhaps?) also say that (unlike concept names) any argname may be translated if the voice language is not the language of the document, and the argname has a known translation. (perhaps) these names are closer to natural language ("lower limit") so safe to translate unlike concept names where they are typically technical terms that you don't want to automatically translate as natural language.

[davidcarlisle]

@NSoiffer I used your wording for the property description

[NSoiffer]

As per the meeting today, I'm accepting this. However, the "examples" need to become more formal and expanded to include all possible values from the core concept list.