?

Log in

No account? Create an account

Previous Entry | Next Entry

So feorlen tells me the Apple Design Award winner int eh Open Source category is a front-end to TeX.

Oh, for God's sake.

C'mon people, grow the fuck up. This is the new millennium. There have been superior typesetting tools to TeX since...oh, I don't know, about 1990 or so.

The Unix world is amazing. Instead of TTY terminals, Unix users now run a GUI so they can open a dozen TTY terminal windows. Instead of WYSIWYG editors, Unix people still cling to antiquated systems like TeX and Scribe. Until Apple came along, nobody had ever written a GUI for *nix that was worth a goddamn anyway.

ATTENTION, LINUX AND UNIX USERS:

Linux will never beat Windows ont eh desktop until my father can install it and make it work. Got that?

Yes, Linux is technically superior to Windows. Yes, it's more stable. Yes, it's more efficient. But guess what? XWindows sucks. The Red Hat installer sucks. The productivity software, for the most part, sucks.

Clue-by-four time: Desktop users do not LIKE to have to make kernel mods. They don't WANT to drop down to a command line to find out what's up with the cable modem not responding. They want to turn the computer on and make it go.

I've been using computers in general since 1976. I've owned two PDP systems (as in, in my house). I've been using Unix since before most Linux hackers were born. I can code in assembly language as fast as I can type. And you know what? It takes me 20 minutes to set up a MacOS or 'Doze system and all friggin' WEEKEND to set up a Linux system. Not because I don't know Linux, but because every piece of hardware that's at all unusual means a trip to Google and an hour of configuration hell--and God help you if you're installing on a laptop.

Writing installers and GUI middleware isn't sexy and wins you no "cool points" with the open source community the way a nifty kernel hack does, but until someone gets it working easily enough that my barely-computer-literate dad can make it work, Linux is going to continue to have its clock cleaned on the desktop. Got it?

We now return you to your regularly scheduled LiveJournal activity.


Comments

( 5 comments — Leave a comment )
progress
May. 12th, 2002 06:11 pm (UTC)
dude-- I love you for saying every single word of that.

I'm all for a stable system-- but not at the sacrifice of functionality and compatibility.

-nik
chipotle
May. 13th, 2002 07:34 am (UTC)
A dissenting opinion on TeX
For the kind of technical, graphics-light and math-heavy documents TeX is best at handling, the only graphic layout program that seriously competes with it is FrameMaker. I've had a lot of long, dreary experience recently in seeing what's being used in the technical publication industry, and while Frame comes up a lot more often than TeX as a required skill, Quark doesn't come up much more often (and your much-despised PageMaker is likely to come up at least as often, because out of the box PageMaker has indexing and list generation features that are expensive add-ons in Quark). SGML is perhaps more prevalent than either TeX or Quark, and if you ain't usin' FrameMaker on your SGML you're probably usin' Arbortext... or Emacs.

I do a quarterly APA submission that's all text and no graphics, and have ended up coming back to LaTeX after trying several other ways to produce it, including InDesign and PageMaker. It's just easier to work with LaTeX's "rational defaults unless overridden" mindset for this kind of document, and it produces perfectly nice--one could defensibly say "beautiful," I think--output.
tacit
May. 20th, 2002 01:33 pm (UTC)
Re: A dissenting opinion on TeX
"For the kind of technical, graphics-light and math-heavy documents TeX is best at handling, the only graphic layout program that seriously competes with it is FrameMaker."

Which also sucks. i teach Frame, and at every turn I am nothing short of astounded at the sheer idiocy of its user interface design. Basic, basic stuff, stuff that's been in the Apple Human Interface guidelines since the 1980's, is completely disregarded.

In any event, i've always thought that the "But they are better for text documents with no graphics!" argument in favor of TeX and Scribe and Frame to be slightly disingenuous.

It's like the old charges levelled against the Amiga--"Only good for games!"--in reverse. Games are very taxing on a computer, and push its resources to their limits; a system good at games is also good at the more mundane number-crunching jobs of, say, spreadsheets.

Likewise, a page layout program good for graphics is also good for text. By saying "TeX is good for straight text," you're adding an unspoken "...because it's so primitive it sucks for anything else." Likewise, you'll always hear teh Frame camp tirelessly saying "Frame is good for long documents," completely ignoring that the modern versions of Quark handle long multi-document books much better than the most recent versions of Frame.

"I've had a lot of long, dreary experience recently in seeing what's being used in the technical publication industry, and while Frame comes up a lot more often than TeX as a required skill, Quark doesn't come up much more often (and your much-despised PageMaker is likely to come up at least as often, because out of the box PageMaker has indexing and list generation features that are expensive add-ons in Quark)."

The problem with the technical publication industry is that it's still trapped in a 1980 time warp. Once upon a time, index generation was not a standard part of page-layout apps like Quark. Of course, once upon a time, we all used DOS 3.3, too.

Index generation, TOC generation, multi-document book handling, rule-based text flow, and automatic book generation are standard parts of QuarkXPress, right out of the box, and have been for years now.

The industry stands still while the state of page generation keeps moving. Which is a flaw not unique to technical writing, to be sure; nevertheless, the old arguments are getting thinner each day.

"It's just easier to work with LaTeX's "rational defaults unless overridden" mindset for this kind of document, ..."

This sort of thing can be done in any modern page-layout app, if you desire. And for people who prefer a workflow in which they type in a word processor with embedded tags for formatting before running the word processor's file through a layout engine a la teX or Scribe, there is an answer: Quark's XPress tags, which work in a similar way, but offer the (rather significant) advantage that the processed file is a Quark document and can be edited WYSIWYG.

"...and it produces perfectly nice--one could defensibly say "beautiful," I think--output."

It's not about the output; hell, the output from an old-fashioned phototypesetter is perfectly nice. (And if it's beautiful typography you're after, you will not be better served by teX than by InDesign anyway; InDesign's type engine is a work of art, producing typography that's nothing short of breathtaking even if you do not do any manual tweaking at all.)

No, it's not the output; it's the process. People no longer use phototypesetters because there are ways that offer better results and improved productivity with less investment. Likewise, archaic systems like TeX require a greater investment of time and energy, both in workflow and in learning curve, for results that are inferior and more dificult to edit. I almost hesitate to think about the cost in terms of man-hours and wasted productivity that the technical writing industry loses every year by insisting on clinging to outmoded workflows that produce results inferior to more sophisticated workflows, and at greater cost.

It's not about output. it's about inertia, it's about an industry that's painfully conservative and resistant to change, and it's even--dare I say it?--about elitism.

[continued]
tacit
May. 20th, 2002 01:33 pm (UTC)
Re: A dissenting opinion on TeX
[part 2]

In a perverse way, many of the people I know who are fans of TeX like it precisely because it's nonintuitive, it's non-WYSIWYG, and it's difficult to master. It offers them a psychological benefit--a feeling of "I have mastered this arcane thing and you have not." The Dilbert "Unix Guru" character ("Here's a nickel, kid, go buy yourself a REAL computer") is not too far off the mark for many TeX users.

After all, anybody can do WYSIWYG page layout, but TeX?...

The emperor has no clothes.

There. I said it. The emperor has no clothes.

Systems like TeX do not offer any real advantages. They do not produce superior output. They do not offer superior workflow. They are not "better for long text." What they are are the relics of a conservative industry unwilling to face change.
chipotle
May. 20th, 2002 04:32 pm (UTC)
Re: A dissenting opinion on TeX
It's a very eloquent, well-reasoned response--but it doesn't convince me that the technical publication industry uses different tools than the prepress industry because everyone in the technical publication industry is elitist or resistant to change, which is essentially what you're arguing.

I've gone through this before with another friend in the graphics/prepress business. It's obvious that doing technical publication work in Xpress would be much simpler than doing intensive graphics work in LaTeX. But you're leaping from that to the conclusion that doing technical publication work in Xpress would also be simpler than doing it in LaTeX--or as a perhaps better non-GUI example, DocBook--rather than allowing for the possibility that the two approaches have different strengths and weaknesses.

This is essentially a prepress variation of the argument for what word processor to use. Since Word can do everything WordPerfect can do and things WordPerfect can't do, obviously, everyone should be using Word. Except that there are some things WordPerfect does more easily than Word does, and those things happen to be very important to law offices. Likewise, there's a lot of things Word can do that Nota Bene (another word processor) can't, and NB is ugly and klunky by comparison. But if you're doing academic papers, NB's bibliography manager, text retrieval system, and framework-based style sheet system will blow Word out of the water.

Technical publishers use Frame, and SGML/XML-based systems because those systems do what they want and Quark and InDesign don't. I did check on indexing and table of contents--yes, Quark 5 includes it. For Quark 4, those abilities required an extension. While I won't conclusively say there's no GUI that typesets math as completely and quickly as LaTeX, I haven't seen one yet--the closest I've seen was Word's inline equation markup, as opposed to the GUI equation editor. And, for technical document maintenance (which is really not in TeX's purvey), you're getting into the realm of SGML/XML.

You may say, "Ah-ha, Quark can do that!" but, well, only partially, and like indexing, it's new. Those potential customers are going to keep maintaining them in ArborText and running those DocBook-to-whatever translators for output unless a proposed replacement system gives them what they want. (And, yep, the "whatever" might well be TeX when they're preparing printed output.)

The other observation I'd make about LaTeX--as distinct from "plain" TeX--is that it's no more complex to write than HTML, and if a user doesn't screw with its defaults he'll get fine (if unimaginative) results without having the faintest idea how to do typesetting. Before you argue that the same's true for Xpress or InDesign, honestly think about documents you've seen made by people who've only been using those programs for a few days. I've seen LaTeX documents made by people who just spent a couple hours with a tutorial document--like I said, unimaginative, but perfectly nice.

And, as I mentioned, InDesign's work of art type engine is TeX's. :)
( 5 comments — Leave a comment )