|
|
Subscribe / Log in / New account

What's new in TeX, part 1

What's new in TeX, part 1

Posted Sep 17, 2015 3:30 UTC (Thu) by butlerm (subscriber, #13312)
Parent article: What's new in TeX, part 1

TeX produced publications invariably use TeX fonts, which makes them instantly distinguishable from anything that uses more conventional fonts, and not in a good way.

I don't know what the problem is, but TeX fonts looks they were frozen in time about thirty years ago. Adequate from a pragmatic point of view perhaps, but I wonder why TeX can't or doesn't use more modern fonts to produce first rate output instead of what is arguably and unfortunately second rate compared to what is being used for the majority of professionally published full length math and science texts.

Possibly those books are produced using TeX, but whatever they are doing they are doing something so that it doesn't look like the run-of-the-mill TeX produced documents that are so common and so visibly distinguishable from everything else.


to post comments

What's new in TeX, part 1

Posted Sep 17, 2015 4:40 UTC (Thu) by xanni (subscriber, #361) [Link] (1 responses)

Did you comment before reading the second section of the article, entitled "Fonts and Typography", which demonstrates that TeX can now make use of all OpenType and TrueType fonts?

What's new in TeX, part 1

Posted Sep 17, 2015 5:08 UTC (Thu) by butlerm (subscriber, #13312) [Link]

If that fixes the problem, that is wonderful.

'Computer Modern' ain't

Posted Sep 17, 2015 16:14 UTC (Thu) by pr1268 (guest, #24648) [Link] (4 responses)

I don't know what the problem is, but TeX fonts looks they were frozen in time about thirty years ago.

Yes, the default TeX serif font ("Computer Modern") is anything but. ;-)

But, even without using the specialized TTFs mentioned in the article (and xanni's comment), \usepackage{times} gives your TeX doc a contemporary Times New Roman-like font. I use this by default in my TeX docs.

For the record, I think LaTeX is difficult to learn and master, but the PDFs it creates are absolutely gorgeous! (IMO.) At least there's a plethora of online documentation; if you want to create a fancy TeX document, and you're not too sure how to do it, then be sure to have a good Internet connection! :-)

'Computer Modern' ain't

Posted Sep 17, 2015 17:17 UTC (Thu) by dskoll (subscriber, #1630) [Link] (3 responses)

\usepackage{times} is a good solution. I also like to use \usepackage{mathptmx} so that equations are set in Times-Roman rather than Computer Modern.

'Computer Modern' ain't

Posted Sep 18, 2015 8:15 UTC (Fri) by eru (subscriber, #2753) [Link] (1 responses)

\usepackage{times} is a good solution. I also like to use \usepackage{mathptmx} so that equations are set in Times-Roman rather than Computer Modern.

Using these also results in far smaller PDF:s, because embedded fonts are not needed so much.

'Computer Modern' ain't

Posted Sep 18, 2015 10:56 UTC (Fri) by jnareb (subscriber, #46500) [Link]

\usepackage{times} is a good solution. I also like to use \usepackage{mathptmx} so that equations are set in Times-Roman rather than Computer Modern.
Using these also results in far smaller PDFs, because embedded fonts are not needed so much.
Actually the new way is (supposedly) to use
% New TX (URW Nimbus Roman) - nie ma zainstalowanego
\usepackage{newtxtext} % roman text font provided by a Times clone
\usepackage{newtxmath} % math italic letters from a Times Italic clone
There is also the problem of creating a PDF in such way that copy'n'paste works correctly even in the presence of characters outside US-ASCII set, in UTF-8 encoding, and ligatures. This is not the problem for LuaLaTeX and XeLaTeX; for pdfLaTeX I ended up using:
\usepackage{tgtermes}       % use TeX Gyre Termes, a font family that extends URW Nimbus Roman
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc} % write document using UTF-8 encoding

\input glyphtounicode
\pdfgentounicode=1
Which is admittedly not very newbie-friendly...

Learned something new

Posted Sep 21, 2015 5:22 UTC (Mon) by pr1268 (guest, #24648) [Link]

I also like to use \usepackage{mathptmx} so that equations are set in Times-Roman rather than Computer Modern.

Thanks! I learned a new feature today about LaTeX I didn't know previously.

But now, my \mathcal 'Ƶ' (as in "z ∈ Ƶ") inside equations and embedded math mode looks like a twisted pound sterling character £. Sigh...

As for Computer Modern, in retrospect, it doesn't look all that bad, even if it appears "old". I do like consistency, so mathptmx FTW.

Now to try Lee's suggestions on modern fonts...

What's new in TeX, part 1

Posted Sep 17, 2015 18:40 UTC (Thu) by anselm (subscriber, #2796) [Link] (2 responses)

With LaTeX, it is trivial to use fonts other than Computer Modern, and it has been for nearly the last 25 years or so. If your system comes with the standard PostScript fonts (Times, Helvetica, Palatino and friends), these are usable “out of the box” based on very simple commands. Other fonts may need some work to generate the required TeX support files, but it is quite doable and there are tools that help with this. Modern TeX implementations (such as XeTeX or LuaTeX) can use OpenType fonts directly, again based on very simple commands (which the article illustrates).

Incidentally, Computer Modern is not a bad design as fonts go. Its main problem is that it was designed to look good in phototypeset output (IIRC, Knuth used to use an Alphatype CRS at Stanford, which has a resolution of 5000+ dpi), and the clunky, comparatively low-res, laser and inkjet printers that everyone uses today hardly do the fonts justice. The other problem is that Knuth's font technology was really years ahead of its time, but was passed on the inside track by less capable but more popular – and more accessible, to people who aren't genius computer science professors – approaches (like PostScript).

Frankly, if you believe that all TeX-produced documents look the same, then you must not have looked at LaTeX in a while. It is now quite straightforward to get LaTeX to produce output that is radically different from the standard look, based on popular and well-documented extension packages that modify, e.g., the page layout or the appearance of chapter/section/… titles. I wrote a book on LaTeX that appeared in the O'Reilly “Hacks” series (German only, unfortunately), and getting LaTeX to produce the series layout wasn't that difficult (the details are in the book).

What's new in TeX, part 1

Posted Sep 21, 2015 6:07 UTC (Mon) by pr1268 (guest, #24648) [Link] (1 responses)

Thanks for the info; you've now got me google-ing for the Alphatype CRS. (Not much found; I did see its claimed resolution of 5333 dpi—WOW!).

This interview with David Fuchs (regarding the earliest days of TeX) was interesting, as was this posting on Tex Stackexchange (how did they view their pretty TeX output on a dumb VT100 terminal console? [they didn't.]).

What's new in TeX, part 1

Posted Sep 21, 2015 12:58 UTC (Mon) by anselm (subscriber, #2796) [Link]

In the early days of TeX, the “preview” device of choice was a Xerox Dover (early laser printer). This was mostly good for checking that everything appeared in roughly the correct place on the paper but otherwise the output looked quite terrible, to a point where people would print deliberately enlarged output and reduce it on a normal photocopier to make it look nicer. While very expensive and – as basically a modified 1970s photocopier – bodily huge (think large chest freezer), the Dover did have the advantage of being very, very fast indeed (60 pages per minute or so, off a large roll of paper that it would cut into individual sheets).

The first TeX implementation I had (on the Atari ST) didn't have a proper DVI previewer, either – it came with a program that basically “printed” pages to the screen, so you were able to look at your document one page after the other but there was no going back to earlier pages (you had to restart the “printer”). Eventually I wrote my own, more reasonable, DVI previewer for the Atari, which in the end supported multiple windows, facing pages, hyperlinks, and that sort of thing. After I had started using Linux I wrote another one, based on Tcl/Tk, but at some point I went over to using PDF for LaTeX output exclusively, and even though the project was still interesting there was no longer an itch to scratch.

What's new in TeX, part 1

Posted Sep 21, 2015 11:51 UTC (Mon) by marcel.oliver (subscriber, #5441) [Link]

I only half agree with this comment. I remember that I considered CM "oldfashioned" when I first started to use TeX about 25 years ago and tried various other font options from time to time. Most of these experiments I abandoned rather quickly.

When doing mainly plain text, there are actually a lot of good and more "modern" options, whatever that means. However, when it comes to typsetting math, it is very hard to compete with CM for two reasons: First, CM is incredibly well done for writing mathematics, as you can see in comparisons like

https://tug.org/pracjourn/2006-1/hartke/hartke.pdf
http://maverick.inria.fr/~Nicolas.Holzschuch/texmath.html
I think to some extent it comes down to having an italics font which is visually very distinct from the upright shape so that the visual parsing of mathematics is easy on the eye. Then big symbols like sums, integrals, and big parentheses are sufficiently large without being very black, so they don't stick out from the page but rather blend in well with the entire formula. Finally, the font metrics are done well and produce well-balanced output.

As second issue when writing professional mathematics is the availability of special symbols beyond the core symbol set, where a lot of the available ones are visually compatible with CM and stick out when used with other math fonts. Here other fonts have caught up in recent years, but the set of CM-compatible symbol fonts is still larger and more readily available than anything else.

So I think a lot of the "professionally produced math and science books" will fall into the category "text with math" rather than "mathematical writing". I probably would not use CM for an introductory Calculus textbook (although this could also be done well!), but if you look at the mathematical research literature, there is a lot of very good publishing in CM, and it seems to me the trend is more toward CM than away from it.

A lot of what you may perceive as a difference between professional and amateur publishing is due to poor document classes, I believe. (And of course poor copy-editing.) Especially if you go beyond relatively dense scientific papers, there are few really good document classes. The LaTeX default is relatively good in that it produces easy-to-read documents, but to my eye fails on aesthetic criteria. Then there is a large number of mutually incompatible publisher classes, some of which are quite good, but most have warts and limitations, and also follow their own conventions for frontmatter items and such so that it's hard to switch document class. So in comparison, the situation with fonts is actually not so bad...


Copyright © 2025, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds