Michael Kohlhase wrote on 14/07/16 at 11:06:
[...]
We see three things we need to do:
[...]
2. think/discuss about extensions of OpenMath, e.g. (from memory)
1. recognize content MathML as an official OpenMath encoding
(legitimized by MathML3)
That sounds good because of a detail in OM that causes lots of
trouble in practice: that IEEE Floating Point numbers have a prominent
and convenient encoding <OMF dec="..."/> but the closest to real numbers
/ bignums is the bigfloat symbol:
http://www.openmath.org/cd/bigfloat1.xhtml#bigfloat
The result was that people kept encoding real numbers with OMF which
is incorrect.
That problem often wasn't obvious because the applications decoding
those numbers made the same mistake so it appeared to cancel out.
CMML3 in constrast has a sane encoding of numbers that doesn't
penalize correct usage:
https://www.w3.org/TR/MathML3/chapter4.html#contm.cn
Is your number a real? Just use type="real" instead of type="double"
and everything is fine! No extra writing/programming effort.
[..]
3. how about a notation system
This is another important aspect in practice.
The first priority of someone not already interested in semantic
encoding is that their formulas look good, which is not a frivolous
concern but an important component of readability.
A solid built-in and supported (i.e. tools) presentation mechanism
would make it less painful to use OM. As it is, anyone that wants to
test drive OM has to first find a way of encoding their formulas
semantically and then a way of displaying them before they can even be
back at the point there started at: showing their formulas to other
people. That's two hard tasks with no immediate benefit, and I say that
as an enthusiastic long time OM supporter that has written several OM
tools over the years: using OM for anything is a pain.
Cheers,
--
Alberto González Palomo
Toledo, España / Saarbrücken, Deutschland
http://www.matracas.org
_______________________________________________
Om mailing list
[email protected]
http://openmath.org/mailman/listinfo/om
