Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

MathML in Firefox 3 beta 5 (Mozilla 1.9)

11 views
Skip to first unread message

Justyn

unread,
Apr 21, 2008, 2:02:13 PM4/21/08
to
Hi,

The spacing issue of bug 363240 (https://bugzilla.mozilla.org/
show_bug.cgi?id=363240) seems to be resolved since Firefox 3 beta 4.

But I notice while viewing examples with a sum/product operator in FF3
beta 5 that the sigma/etc symbol is not enlarged to enclose the
corresponding equation. Is this a bug or a feature, does anyone know?

See http://www.mozilla.org/projects/mathml/start.xhtml (particularly
examples nearer the bottom).

Regards, Justyn.

Karl Tomlinson

unread,
Apr 22, 2008, 6:42:05 PM4/22/08
to
Justyn writes:

> But I notice while viewing examples with a sum/product operator in FF3
> beta 5 that the sigma/etc symbol is not enlarged to enclose the
> corresponding equation. Is this a bug or a feature, does anyone know?

That would be most easily fixed by

https://bugzilla.mozilla.org/show_bug.cgi?id=407101
"Complete entries in mathfontSTIXSize1.properties"

and more completely fixed by

https://bugzilla.mozilla.org/show_bug.cgi?id=414277
"scale stretchy operators when there is no glyph of suitable size"

Alvaro

unread,
Apr 28, 2008, 5:44:03 PM4/28/08
to
Justyn escribió:

Sum, Product and Integral should not stretch arbitrarily, but should be
larger than they display now (Integral shows small in some cases). Sum
looks like a normal uppercase Sigma, when it shoud be around twice as big.

And parenthesis and braces are usually too small: these should stretch
to their content. Even with stretchy="true" they don't.

Alvaro

unread,
Apr 28, 2008, 5:52:55 PM4/28/08
to
Justyn escribió:

Hi,

Apart from the size of sum/product operators, and parenthesis/braces
stretching (not too bad), the other annoying problem are *root symbols*:

See http://www.mozilla.org/projects/mathml/demo/texvsmml.xhtml

They don't scale vertically to enclose their operand.

BTW, integrals are OK. They look small with stretchy="false" (well,
maybe too small).

Ray Kiddy

unread,
Apr 29, 2008, 12:41:44 PM4/29/08
to

Does anyone have reftests of mathml functionality working on FF3? On
what platforms?

I am on Mac OS X and, for quite a while, I have been reading what people
say about what they see and I have no idea what I am supposed to be
seeing. I look at the W3C MathML tests and I see vast differences
between what is rendered and what is supposed to be there. But I also
have questions about when I should have to install fonts and where and
how they get installed. I want to put fonts in some place, and not run
an application which does I know not what.

I have tried to figure out ways for people to share pictures of what
they see. However, swapping screenshots is a pain. But reftest would
work perfectly.

I think one problem may be that the reftest engine may have to be
extended to have OS-specific images instead of having "skip-if",
"random-if" and "fails-if" in the manifest files. Perhaps this is why
the reftests of mathml functionality do not seem to be getting run as a
part of any automatic process.

If anyone has questions about reftests, please see
http://developer.mozilla.org/en/docs/Creating_reftest-based_unit_tests
or ask questions. If anyone has reftests, please speak up. If anyone
knows of a better way to share info about how these things are supposed
to look, please share any and all ideas.

cheers - ray

Karl Tomlinson

unread,
May 1, 2008, 7:41:33 PM5/1/08
to
On Mon, 28 Apr 2008 23:44:03 +0200, Alvaro wrote:

> Sum, Product and Integral should not stretch arbitrarily,

I'd be interested to hear why you say that?

I had thought that Sum and Product looked better without
stretching to enclose their contents but that maybe due to looking
at too much LaTeX output.

The MathML spec explicitly says: "Also, operators such as ∑,
∫, /, and vertical arrows stretch vertically by default."

http://www.w3.org/TR/2003/REC-MathML2-20031021/chapter3.html#id.3.2.5.8.2

UTR #25 also indicates that these sometimes stretch.
"Large Operators include n-ary operators like summation and
integration. They may expand in size to fit their associated
expressions."

http://www.unicode.org/reports/tr25/tr25-9.html#_Toc323

Karl Tomlinson

unread,
May 1, 2008, 7:54:27 PM5/1/08
to
On Mon, 28 Apr 2008 23:44:03 +0200, Alvaro wrote:

> And parenthesis and braces are usually too small: these should
> stretch to their content. Even with stretchy="true" they don't.

Sounds like you may be using Windows XP without fonts for Unicode
math support.

Nightly builds now handled many of these situations specially
using SymbolMT.

https://bugzilla.mozilla.org/show_bug.cgi?id=425367

On Mon, 28 Apr 2008 23:52:55 +0200, Alvaro wrote:

> ... the other annoying problem are *root symbols*:

> They don't scale vertically to enclose their operand.

Currently you need STIX fonts to get stretchy radicals

http://www.mozilla.org/projects/mathml/fonts/stix/beta-license.txt
http://www.mozilla.org/projects/mathml/fonts/stix/STIXBeta.zip

It would be possible to do something with Cambria Math and
OpenSymbol (from OpenOffice) as described here:

https://bugzilla.mozilla.org/show_bug.cgi?id=372351#c6

> BTW, integrals are OK. They look small with stretchy="false"
> (well, maybe too small).

They should be larger. That could be easily corrected when STIX
give their various integral fonts unique family names as indicated
here:

http://www.stixfonts.org/STIXfaq.html

David Carlisle

unread,
May 2, 2008, 4:14:47 AM5/2/08
to moz...@karlt.net, dev-tec...@lists.mozilla.org

> I had thought that Sum and Product looked better without
> stretching to enclose their contents but that maybe due to looking
> at too much LaTeX output.

LaTeX (with default font setup) doesn't stretch these symbols, it just
has a larger size for display mode.(Integral sometimes looks better if
stretched, but sum usually not, I think) Integral you can stretch by
moving to an upright shape and inserting straight bits, as you'd stretch
a bracket, but summation you'd just presumably have to optically scale
the whole glyph which will make it look fairly horrible if the
expression gets large.

David

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs.
________________________________________________________________________

Alvaro

unread,
May 2, 2008, 6:47:22 AM5/2/08
to
Karl Tomlinson escribió:

> On Mon, 28 Apr 2008 23:44:03 +0200, Alvaro wrote:
>
>> Sum, Product and Integral should not stretch arbitrarily,
>
> I'd be interested to hear why you say that?

Open any good math or physics text book and look at its equations. Or
look at the output by LaTeX. These two are better sources of what math
should look like than a standard, IMHO.

The symbol for Sum, Product, Union, Intersection, etc. is a big Sigma,
Pi, U, ... but of a fixed size

> I had thought that Sum and Product looked better without
> stretching to enclose their contents but that maybe due to looking
> at too much LaTeX output.
>
> The MathML spec explicitly says: "Also, operators such as ∑,
> ∫, /, and vertical arrows stretch vertically by default."
>
> http://www.w3.org/TR/2003/REC-MathML2-20031021/chapter3.html#id.3.2.5.8.2

I did read that line from the spec I think it is wrong. I prefer proper
looking *math* rendering than following blindy a spec. :-)

0 new messages