New CTAN package: The STIX fonts
CTAN Announcements
> The following information was provided by the package's contributor.
>
> Name of contribution: The STIX fonts
> Author's name: Will Robertson
> Location on CTAN: fonts/stix
> Summary description: Unicode mathematics fonts, with a complete repertoire of glyphs
> License type: ofl (SIL open font licence)
>
> Announcement text given by the package's contributor:
> ----------------------------------------------------------------------
> The STIX fonts are a suite of unicode OpenType fonts
> containing a complete set of mathematical glyphs.
> They may be obtained from <http://www.stixfonts.org/>.
>
> This repository is a mirror of their official release,
> organised in the form of the `TeX Directory Structure'
> for inclusion in TeX Live and MiKTeX.
>
> The STIX fonts are released under the SIL Open Font
> License, Version 1.1. They have copyright (c) 2001-2010
> by the STI Pub Companies. Please see the official
> documentation for further information.
>
> ----------------------------------------------------------------------
>
> Users may view the package catalogue entry at
> http://www.tex.ac.uk/tex-archive/help/Catalogue/stix.html
> or they may browse the package directory at
> http://www.tex.ac.uk/tex-archive/fonts/stix
the catalogue entry won't appear until some time tonight.
Robin Fairbairns
For the CTAN team
Tariq
I am not sure if very many folks have already given a try to these
fonts under XeLaTeX or not but this is what I have noticed after my
first compilation.
1. The kerning is awful; (for example, it is very striking between
pairs like VA or Te) and this is after all the weeks and months they
spent on setting up the kerning tables. The output looks more like
something from M$ Word.
2. There are no smallcaps available.
3. There are no text (non-lining ) digits available.
Overall, the output looks like some pirated version of Times Roman or
something. This sort of stuff might be good for online viewing on a
web page. I would hesitate to put it in my pdf documents that I print
for my students. My comments seem a bit harsh (even to me) but that is
probably after all the hype that had been built for all those years
around these fonts. All in all, the kerning should at least be set
right even if other issues cannot be dealt-with right away.
Tariq
Joris
That is pretty disappointing.......
Will Robertson
> 1. The kerning is awful; (for example, it is very striking between
> pairs like VA or Te) and this is after all the weeks and months they
> spent on setting up the kerning tables. The output looks more like
> something from M$ Word.
>
> 2. There are no smallcaps available.
>
> 3. There are no text (non-lining ) digits available.
Remember that the primary service of the STIX fonts is for mathematical
typesetting. (Although I'd welcome a well-designed set of small caps
and OSF, as no other free Times font has them either. Do any commercial
ones?)
Look in the glyph tables for the thousands and thousands of
mathematical symbols that have been drawn specially for these fonts.
W
Andrew
Times from Adobe or Linotype has really good OSFs (and SCs), but that
doesn't really help. The txfonts package is another place to look, but
txfonts's OSFs aren't all that flash. STIX does have OSF, but they're
quite bad, I think.
Andrew
Grant Lotter
The TeX Gyre Termes font -- a Times lookalike -- has both SCs and
OSFs. Not sure how good they are.
Also, maybe we should wait till the official STIX LaTeX support files
are released before passing judgement? Based on the project\s past
record that should be in about June 2015.
Tariq
Will's point is well taken. However if simply pure math support was
the only thing to worry about, Knuth might never have bothered to mess
with the idea of "typesetting beautiful books" to the extent that he
did in the first place. I tend to think that good typography should
encompass both text and math. Sure enough, the number of glyphs in
STIX is very impressive. I just wish they had opted for some original
design for the typeface rather than settling for a bland (boring?)
"Times."
Grant, thanks for pointing out the TeXGyre project; they have really
well-designed Times substitute called Termes and it might be
worthwhile to try combining the Termes otf with the math support from
STIX until some decent (La)TeX support is provided for STIX.
Regards,
Tariq
Robin Fairbairns
> [...] Sure enough, the number of glyphs in
> STIX is very impressive. I just wish they had opted for some original
> design for the typeface rather than settling for a bland (boring?)
> "Times."
the whole project was funded by a group of mathematical/scientific
publishers. how many journals do _you_ read in fonts significantly
different from times?
(for me, it's one: tugboat.)
--
Robin Fairbairns, Cambridge
William F. Adams
> > design for the typeface rather than settling for a bland (boring?)
> > "Times."
>
> the whole project was funded by a group of mathematical/scientific
> publishers. how many journals do _you_ read in fonts significantly
> different from times?
I do journal production for an international publisher of STM journals
and off the top of my head we use:
- Bembo
- ITC Berkeley
- Bliss Pro
- Adobe Garamond
- Palatino
- ITC Slimbach
Not much math in the ones we do though.
William
Tariq
>
> > the whole project was funded by a group of mathematical/scientific
> > publishers. how many journals do _you_ read in fonts significantly
> > different from times?
>
In the field of molecular & cellular biology/microbiology/
biochemistry, there might be only a handful of journals that still use
Times. There was a time (way back) when some of the journals had Times
as text font. Almost all the older journals have already switched to
aesthetically more appealing typefaces and the newer ones probably
always had access to other digital fonts when they came into
existence.
I must add one more observation: There are some very prominent
journals in these areas that (for some mystical reason) have been
using sans serif fonts (like Helvetica or its clone) and their
typographic taste is atrocious, to say the least, on many accounts
besides the choice of typeface.
Tariq
Christophe Caignaert
>
> 1. The kerning is awful; (for example, it is very striking between
> pairs like VA or Te) and this is after all the weeks and months they
> spent on setting up the kerning tables.
Don't forget that the kerning is very hard to finish.
If you are working on screen, you use scaled letters and often the
kerning is too important.
Good kerning needs to print test files, corrects, print again...
Christophe
Joris
wrote:
I'm sure it's hard, but that doesn't make it any less important or
less surprising that they didn't do it.
Rogério Brito
On 06/01/2010 07:50 AM, Robin Fairbairns wrote:
> the whole project was funded by a group of mathematical/scientific
> publishers. how many journals do _you_ read in fonts significantly
> different from times?
I am not exactly sure what "significantly" means, but I happen to have some
papers here:
* SIAM J. Appl. Math., 1992: it uses Computer Modern
* Theoretical Computer Science, 1997: it uses Times
* Mathemastische Annalen, 1982: it uses Times
* Bulletin of Mathematical Biology, 1993: it uses Times
* Proceedings of the IEEE, 1989: a sans-serif font that I don't know
* Revista do Professor de Matem�tica, 2010: it uses Utopia
* National Mathematics Magazine, ca. 1940: a serif font that looks like Old Standard
* Journal of Computer and System Sciences, 1980: perhaps Times, but it looks a
little bit more spaced between the letters
* MAA FOCUS, 2009: Optima.
The article by Fernando Quadros Gouv�a, editor of the MAA FOCUS states, on page
27 states some of the guidelines that they take:
http://www.maa.org/pubs/jan09web.pdf
But I would guess that Times is all will remain the pervasive font of use. I
just wish that people would use something like Linux Libertine, instead, but
many journals would interpret that as "ruining their identities", I think.
Whatever. I hope that, at least, I can make some glyphs of a derived version of
URW Garamond No. 8 not suck. :-) Interested parties are invited to take a look
at http://github.com/rbrito/urw-garamond
Regards,
--
Rog�rio Brito : rbrito@{ime.usp.br,gmail.com} : GPG key 1024D/7C2CAEB8
http://rb.doesntexist.org : Packages for LaTeX : algorithms.berlios.de
DebianQA: http://qa.debian.org/developer.php?login=rbrito%40ime.usp.br
Christophe Caignaert
Sure, you're right !
But I think it's not so hard to correct, if you make some kerning at
huge size, on screen:
- reduce great kernings (VA, To...) down to 50% of its value
- reduce light kernings (co, ca, te...) down to 70% of its value
produce usually not too bad start point.
Christophe