w3c test minus9

[christianp]

http://www.w3.org/Math/testsuite/build/main/Content/ArithmeticAlgebraLogic/minus/minus9-form.xhtml

Choice - currently we collapse the + - into just a minus, but the sample rendering keeps it and puts brackets around the negative term.

[davidcarlisle]

On 11 April 2014 11:14, Christian Perfect [email protected] wrote:

http://www.w3.org/Math/testsuite/build/main/Content/ArithmeticAlgebraLogic/minus/minus9-form.xhtml

Choice - currently we collapse the + - into just a minus, but the sample
rendering keeps it and puts brackets around the negative term.

There will always be someone teaching negative numbers who wants to write 1

• (-1) = 0

but generally I to write x -y + z I think the most natural way is

(plus x (minus y) z)

and rely on the render to use the infix -.

So I'd collapse the -+ The user could always "hide" the negative term (in
for example to avoid collapsing in cases where it is not
needed..

[pkra]

I wonder where to expect the construction from the MML test. I can't think of a situation where someone/something would create that Content MML markup without expecting it to be rendered 7 + (-1). If it's e.g. part of a derivation, then I would expect that derivation to continue on to 7-1 and wouldn't want the renderer to skip that step. I'm sure I'm missing a use case here.

[davidcarlisle]

On 17 April 2014 22:37, Peter Krautzberger [email protected] wrote:

I wonder where to expect the construction from the MML test. I can't think
of a situation where someone/something would create that Content MML markup
without expecting it to be rendered 7 + (-1). If it's e.g. part of a
derivation, then I would expect that derivation to continue onto 7-1 and
wouldn't want the renderer to skip that step. I'm sure I'm missing a use
case here.

—

It's a test Neil set me right at the beginning:-) to make sure that infix
minus got used. Basically the use case is polynomials you want to express a
polynomial of to a list of monomials and that ends up

x^3 -3x +2 as being

(plus (power x 3) (times -3 x) 2)

and so ctop goes to some lengths not to render that as

x^3 + -3x + 2

[pkra]

Thanks for the use case! I can see why one might prefer that. Personally, I'd still prefer to leave + (-x) (for me it corresponds better with the markup) but it's no big deal (and since ctop has been doing this forever, we should probably keep it).

[christianp]

If you're concerned about getting a particular rendering for something that has the same meaning, you should be using Presentation MML anyway, right?

[pkra]

Well, yes, but I think it's an example for #30 -- in a larger body of content, the author might want to change this behavior globally without having to generate a suitable Presentation for each equation. But I'm ok with leaving the default as it has been.

[christianp]

I've added a setting CToP.settings.collapsePlusMinus which dictates whether the collapsing happens. The default is for it to happen.

[davidcarlisle]

On 2 May 2014 14:55, Christian Perfect [email protected] wrote:

I've added a setting CToP.settings.collapsePlusMinus which dictates
whether the collapsing happens. The default is for it to happen.

—

Excellent!

I always had in mind that a content mathml rendering system should be
configurable but just telling people "there is the xsl, edit it" doesn't
really count as a configuration option in a web-connected world.

David