STIX Project Releases v1.0 of Its Scientific Fonts Set

starseeker writes "The Scientific and Technical Information Exchange (STIX) font creation project has released version 1.0 of its font set. This release is the product of almost 15 years of work, with the goal of creating a comprehensive set of fonts for scientific and engineering manuscript creation. The fonts have been released under the SIL Open Font License, and can be downloaded here. Among the many potential applications is proper universal support for MathML in web browsers." If you want a peek, here's "a page for viewing the thousands of glyphs (as a first approximation, think of a glyph as an individual character)."

Comments

Re:Why I am supposed to care?

[Anonymous Coward]

Because Arial doesn't have the -right- oddball glyphs.

"The largest component of the fonts is devoted to the thousands of mathematical operators and technical symbols necessary to report research."

Fonts

[Arker]

The biggest problem with 'modern' fonts I can see is that so few have proper differentiation between O and 0. It's an ugly thing, particularly when it's a problem we solved decades ago and should have stayed solved. Yet somehow it doesnt.

Is downloading this package going to help with that problem? MathML is nice but I dont actually need it. 0s that actually look like 0s would make me very happy though.

Re:Fonts

[Arker]

Using a zero for an empty set is perfectly understandable, both may be voiced as 'null.' It is difficult to think of a situation where one usage of it would be confused with another - i.e. where context would not make it clear and obvious which is meant. Also there are other ways to symbolise a null set, {} coming to mind immediately.

Slashed zeros look nothing like a theta. The only thing other alphanumeric character they resemble is the letter Ø used in Norwegian and Danish. The chances for confusion there are vanishingly small, even for those of us in the tiny minority that use that character.

Slashed zeros look nothing at all like thetas, you must have meant the dotted zero. Yet the dot and the crossbar are significantly different, and again, only a tiny minority of us use thetas to begin with, and in context the chance of confusion is miniscule.

On the other hand distinguishing between the capital 'o' and the numeric zero is a much more common case, occurring in every language rather than a tiny fraction of languages, and it is fairly frequently not immediately obvious from context which is intended, so the rational thing is to differentiate those characters clearly, even if it creates ambiguity in a handful of far less common cases.

What a peek

[imsabbel]

call me spoiled, but if you announce a link for a "peek", i expect something other than a website that prompts me to install the fonts i wanted a peek at.

How about something useful, like comparisions with existing fonts to show what the big deal about these new ones is. Preferably in a way that doesnt require having them installed.

Hell, how about making that stupid 100 screen long page a PDF with the font embeded?

Re:What a peek

[Phiz]

call me spoiled, but if you announce a link for a "peek", i expect something other than a website that prompts me to install the fonts i wanted a peek at.

I agree. However, if you download the font, in the archive you'll find a directory "Glyphs" that has a bunch of PDFs with tables of the characters shown. Therefore you don't actually have to install the font to take a look at it.

They should have made it easier to find these PDFs....

Re:What a peek

[Dynedain]

Seriously.... this has got to be the only website out there that is pushing a font without actually providing any preview of what it looks like.

Slow news day...

[fishthegeek]

15 years? Sounds like Stix has too much time on their hands.

Re:Slow news day...

[starseeker]

Despite the (simultaneously amusing and frustrating) schedule slippage that the STIX project became infamous for, they deserve tremendous praise for the work they have done here. They have created a monumental work, stuck with the project for well over a decade despite all the setbacks, and released the results free to the world as an open source font. To understand the magnitude of this task, consider how long it would take to design a good quality font for just the standard character set (which isn't easy, as witnessed by the large number of bad fontsets floating around...). Now, scale that up to 8,000+ characters. Not only that, but many of these characters are obscure to any outside of specific scientific fields, and hence the font designers won't immediately know how the characters are supposed to look - the background research must be done, the results organized into some coherent framework, and a LOT of characters have to be created more or less from scratch. This work was being funded by scientific publishing houses who wanted a font for high quality consistent output, so the goal wasn't met by "partial" success - it had to be judged a finished work before any benefit could be realized. They couldn't use the TeX approach of allowing the user to custom-roll their own solution to strange characters - everything had to be handled "up-front" and built into the font.

That is an ENORMOUS task, and the result is a VERY significant contribution to the open source world. I have nothing but admiration and gratitude for the people behind this who persevered and succeeded, not just technically and organizationally but also in working through the legal questions raised by the open source community when selecting an open font license. The publishing houses could have decided that "done and usable in journal paper printing" was "good enough", but the project elected to listen to and work with the community to arrive at the OFL, which is a little odd but apparently workable both for the companies involved and the open source community. So, to those who worked on this project: Thanks!

Re:Why I am supposed to care?

[Anonymous Coward]

None. But then, if you are writing real math papers, you already would have learned to stay away from MS Word & Arial Unicode.

Since Arial is not open, it can't be (for example) shipped with a TeX distribution.

Why did you even comment on this story, anyway?

How to use them properly?

[tenco]

Finally. But how to use them properly with webbrowsers? For Firefox I had to use

math { font-family: STIXGeneral, STIXVariants, STIXIntegralsSm, STIXIntegralsD, STIXIntegralsUp, STIXIntegralsUpSm, STIXIntegralsUpD, STIXSizeOneSym, STIXSizeTwoSym, STIXSizeThreeSym, STIXSizeFourSym, STIXSizeFiveSym, STIXNonUnicode; }

in my userContent.css to get partial math glyphs from installed STIXfonts. To get every MathML char rendered with STIXfonts I had to disable custom fonts for webpages and set all fonts to STIXGeneral. Unfortunately STIXGeneral seems to be only available with serif. Which isn't all that good for reading text from displays.

Re:Why I am supposed to care?

[VGR]

Because of what they've encoded in the Private Use Area block at code point U+E0F2. Check STIXv1.0.0/Glyphs/STIXNonUnit.otf.pdf in the zip file to see it, and check the last link in the summary for the character's name. I hope that gets folded back into the Unicode Standard someday....

TeX?

[Anonymous Coward]

Why can't we just make basic latex commands part of html? There are some pretty lightweight tex compilers out there which, in a perfect world, should have been shipping standard on new computers for ages now.

Re:TeX?

[bcrowell]

If what you want is to embed tex in html, you can do it: http://asciimathml.sourceforge.net/ But that doesn't do anything about solving the font problem. You need the fonts, otherwise it can't be rendered. Knuth's Computer Modern fonts were an impressive achievement, but they date back to 1992, and the technology they're built on is obsolete and doesn't fit well with the modern operating systems, or with modern encodings such as unicode. Knuth invented scalable fonts before Adobe reimplemented and commercialized them.

Re:TeX?

[SEE]

TeX4ht renders TeX as HTML.

MathTeX is a CGI that renders TeX embedded in a webpage as an image.

The Techexplorer plugin, originally by IBM, will directly render TeX embedded in HTML.

LaTeX package?

[BitterOak]

I've looked at these fonts, and they look very good. Is there a LaTeX package that will let me use these fonts with LaTeX yet? I think they look much more attractive than the Computer Modern fonts that ship with LaTeX.

Re:LaTeX package?

[SEE]

Not yet. Per the website, this set is 1.0; modifications to work with MS Ofice are targeted for 1.1 (expected release this year), and LaTeX with 1.2 (expected release next year).

Re:LaTeX package?

[iris-n]

No but yes. What you need to use them now is XeTeX, a TeX engine that lets you use OpenType fonts in your TeX documents.

LaTeX per se uses only Type 1 (actually a few more) fonts, and these aren't ready yet.

There is a much bigger problem than "0" and "O"!

[Bruce Perens]

Uh-oh. Nobody told those publishers that the SIL license was not written by a lawyer and never had any legal review. OSI, unfortunately, will approve a license that hasn't been crafted by a lawyer, and this can be a big problem for the users of that license, when the license acts radically differently in court than they expect it to.

The problem in this case is that the license allows conversion of the font to any other license or public domain once it is embedded in a document. The license explicitly says that it no longer applies once the font's embedded. And the authors didn't realize that if you extract the font from the document, the license doesn't come back!

I left a note on their web submission form.

Bruce

Downloadable font demo of these fonts

[Animats]

To see what it was like to use these fonts on the web, I created a test page. This uses dynamically downloaded fonts, which work in most current web browsers. (Firefox users need Firefox 3.6 or later.)

This sample is sized at 16 point. Smaller than that, many of the symbols are unreadable. That's something to be careful about. When you have a huge symbol set, the symbols need to be bigger. However, some of the symbols don't scale up well. If you scale up that page, the integral symbols look great, but the arrows become pixilated. Some of the symbols seem to have been were badly encoded.

This is just a raw demonstration of the font; for formulas, you'd use MathML. I'm not sure if MathML, the W3C names for math characters in HTML, and the STIX fonts are all synchronized yet. But at least you don't have to tell people "to display this page, install all these fonts first."

Re:There is a much bigger problem than "0" and "O"

[Bruce Perens]

I agree that font outlines are not copyrightable (although font programs are) and that fonts can be traced without infringement. It is easy for the font to leave the control of the copyright holder, although perhaps without any hinting that is embedded in the font.

But in this case, the entire intent of the SIL license is that the fonts do not appear in one of those "10,000 Fonts!" packages sold by folks who only aggregate them and do none of the work. And that intent will be thwarted.