Re: State of the N-ary

On 2 Jul 2008, at 14:04, Ivan Herman wrote:

> Bijan Parsia wrote:
>> On 2 Jul 2008, at 13:47, Ivan Herman wrote:
>>> Bijan Parsia wrote:
>>> [snip]
>>>> Fourth, syntax:
>>>>     The XML/functional syntax is easy, though we could add a bit  
>>>> of sugar to make writing equations nicer. I don't see any reason  
>>>> not to use MathML.
>>>>     For RDF, I thought equations could use MathML too (as a  
>>>> literal or data uri) for inline equations. We should also allow   
>>>> naming predicates.
>>>
>>> Just for my understanding (and to be a bit more precise)...
>>>
>>> MathML is actually a strange beast, because it is two different  
>>> markups in one specification. They have a Presentation Markup[1]  
>>> and a Content Markup[2]. (Roughly speaking the presentation  
>>> markup is, well, for the presentation of mathematical equations  
>>> and formulae, whereas the content markup describes the the  
>>> abstract mathematical notions. In some cases they can be mixed.).
>>>
>>> I would expect that we would restrict to the content markup in  
>>> this case. Am I right?
>> That was my intent and how I sketched it out. If presentational  
>> markup proved so much nicer, we could use that with an  
>> understanding of exactly what it represented.
>
> Thanks. And I agree with your choice. I think the presentational  
> markup is not really appropriate here, actually. You want to  
> express the intent of the mathematical formulae

That's what motivated me.

> and not, say, whether the plus sign is infix or not (and the  
> presentation markup forces you to make this choice)...

Right, but for authoring that makes sense. So, say, in the xml  
syntax, you might support a "nice" presentation syntax and define a  
mapping to the content. <shrug> That also could just happen at the  
tool level, but in my current experience, having tools fiddle with my  
stuff is nasty. Protege4 strips parenthesis I put it to help mark  
precedence . The result is pretty printed and it's really frustrating :)

But in the abstract model, it should be content all the way.

Cheers,
Bijan.

Received on Wednesday, 2 July 2008 13:18:42 UTC