I don’t mean blog posts or emails. For blog posts I use souped up gedit, and for emails I use pico. (There was a time when I thought I’d be using emacs for both of those, but emacs suffers from a deep philosophical dysfunction.) I’m talking about longer documents that have sections with headings, bibliographies, etc. I may well make this transition with the never-ending paper I’m writing with Lizzie.


It is hard to describe the difference between what are called markup systems and, say Microsoft Word, OpenOffice.org Writer, or AbiWord to people, especially to some of the newer people who were not initially weaned on ed. And, it is especially hard to explain LyX. But I’ll give it a go.

Word processors such as those just mentioned have you putting letters and white space on a virtual sheet of paper, formatting (type setting, really) these words and white spaces as you wish. The page you see on the screen is like the page that will be printed out on the default printer for your computer (though it will look different on any other printer or computer, likely) or if made into a PDF file (for software with that capability).

Word processors ALSO have features such as cross referencing, making a table of contents, etc. You can insert graphics, and make tables. Etc. That does not make something a word processor in modern parlance. A word processor can not have those features, and something with those features is not necessarily a word processor.

There is a different category of “processing words” (as it were) known as “markup.” Early word processors were markup systems, but today when we use the word “word processor” we always mean the “what you see is what you get” (WYSIWYG) software. Anybody ever use WordStar? Where .pa at the beginning of a line meant “page break” for instance? In WordStar you could see the text you wrote, then you could see the formatting commands that go along with the text, which changed the way the text was rendered.

You may also remember WordPerfect, which had an on screen version of the text that showed italics and bold and stuff (if you had the right graphic card) but also had a “reveal codes” features that let you see the markup codes.

Back then, in pre Windows times, was also xywrite. This was a markup system capable of handling documents of arbitrary size (in computer speak “arbitrary” means “any”) and was very stable and very fast. xywrite was the “typesetting” software used by many major publishers.

Each of these early word processors eventually evolved a “what you see is what you get” version that would run on Windows, just before being killed off by Microsoft Word. Word dominated the market and took over for two reasons. 1) It never looked like a markup system. It always only showed you WYSIWYG versions of your document. This made some people happy, other people nervous, but as usual, the happy people won. Damn them; and 2) Microsoft cheated, lied, cajoled, payed off, threatened, and bribed as needed to makes sure their product took over even though it was not nearly as good as WordPerfect at any moment that the two products existed side by side. I assume.

So that was the end of markup for the average person.

But, back in those days, there were other markup systems that never made the transition to WYSIWYG implementation. Nobody wanted them to make this transition and there was no need for this transition. One of these systems was Tex, run on Unix, and it’s variant LaTex (the details of this history will be glossed here, but if you search for TeX and LaTex on the internet you will find more than you will want to know about that).

The way these markup systems work in their most basic form is that you produce a text file with markup in it, which will render a certain layout of words and whitespace when printed. Usually, your first printings are into a format that is not on paper but rather electronic and on your hard drive, and you can look at this version to see if it is what you want before sending it to your printer, usually as a PDF file.

LyX is a bit of software (they call it a “document processor”) that is a little bit in between a text file with markup and a WYSIWYG word processor. It allows the user to enter text, spell check it, format it, make cross references and stuff, and so on, but the on screen version of the document you are working on is not what is going to show up when printed or converted to a PDF or whatever. To see that you press a button and the document is re-rendered into a visible format exactly representing the final form. When you are done with the document, you can turn it into something on paper, a PDF file, and HTML document, whatever whatever.

In other words LyX in a fairly easy to use graphical interface for Tex/LaTex documents.

One of the things that makes people go gaga over LyX and Tex and Latex is that it works great for mathematical formulas. This is a problem for me because I hardly ever need to work with mathematical formulas. But, any manual, any on line information, etc. about LyX, TeX, or LaTex is about one third or more about mathematical formulas.

Why would you NOT use a WYSIWYG word processor for a document that has several sections, several graphics and tables, and is a hundred pages long or so with a table of contents and references?

Have you tried to do that? And you still don’t know why you would want to avoid, say Microsoft Word or even OpenOffice.org Writer for such a thing?

Here, have some more of this:
i-25c96bc97af43f42118b700f5de909a8-drink_the_kool_aid_baby.jpg

Comments

  1. #1 Erp
    November 7, 2009

    I confess I still use straight LaTeX (and emacs) with sometimes dips into TeX for macro writing. For anything substantial such as a PhD thesis it is the way to go (not that I’ve done a PhD thesis but I’ve helped a fair number of students format theirs over the years).

  2. #2 Greg Laden
    November 7, 2009

    Erp, are you saying that LaTex is the way to go for something big instead of Lyx, or instead of, say Word?

  3. #3 NewEnglandBob
    November 7, 2009

    Ah, Wordstar. I used to love it, back in the day. I think I still have the manual around here somewhere.

  4. #4 Eric Lund
    November 7, 2009

    I’m sure it’s possible to write a Ph.D. thesis or similar in Word and have it be convertible without too much trouble to some other format (book or set of journal papers) that you might publish alongside your thesis. But not more than one in a thousand Word users, and probably closer to one in a million, has both the know-how and the self-discipline to pull it off. If you’re one of the other 999(,999) Word users, it’s going to be a huge mess. Likewise, it’s possible to write an impossible-to-adapt long document with LaTeX. However, you have to actually *try* to write such a disastrous document, whereas in Word that’s the default. LaTeX is definitely the way to go.

    For instance, you might discover at the last minute that your document formats are not quite what the library or publisher wants. Tweaking it in LaTeX is relatively easy. Making those changes in Word (unless you are one of the aforementioned few who managed to get it right) will be painful.

  5. #5 Linux in Exile
    November 7, 2009

    Interesting that you mention “what-you-see-is-what-you-get” (“WYSIWYG”) and “Word” together, since Word’s WYSIWYG isn’t “what they get”.

  6. #6 Ian Tindale
    November 7, 2009

    Good post. I too recently blogged about Where word processors went wrong.

    Incidentally, I used to be very highly against any TeX based way fo doing things, for their violation of the idea of keeping content and presentation separate – they simply don’t. These days I recognise that people writing don’t necessarily have editors, subeditors, designers and a complete publishing house behind them, so yes there has to be a way of incorporating presentational aspects (although I’d still prefer it if it were separated and ‘applied’ according to media choice given in the request for content).

    Incidentally II, each time I’ve ever tried learning emacs, I gave up. I now realise what that point was. It was the point where emacs began using me, not the other way round (and that point is reached quite early int the curve).

    Incidentally III, what on earth is that colourful and in-your-face graphic of some cartoon character offering me a glass of unseasonably cold-looking red liquid? Well, not so much what is it, but what is the relevance? I don’t get it.

  7. #7 James
    November 7, 2009

    I remember using LaTeX to write documents in college. I experimented with a lot of technology back then, even though I was a physics major. I used LaTeX to write several physics lab reports, and more than one paper for English class.

    I liked that I didn’t need a GUI to use LaTeX, so could drop onto any of the VT220 text terminals we had on campus (there was usually at least one in every building, for student’s to access the VAX system) and write my paper. That saved a trip to the main computer lab (or back to the dorm) just to add to my class paper. The ‘aspell’ program could do spell checking, and ignore the LaTeX codes – so I never had a typo.

    When I was done, I’d generate a DVI document and print it (to the HP laser printer if I was in the computer lab, or to an Epson dot matrix if I was in my dorm room.)

    LaTeX also let me focus on the content of the paper, rather than how it looked. For a student, I’d argue that was very important to learn. Even now when GUI is taken for granted, I sometimes wish for a simple system like LaTeX to put together a document.

    Once you’d used LaTeX for a little while, your mind entered this weird space where you no longer saw the codes anymore. You saw the formatting they generated. I’d type {\em Book Title} but my mind would see: Book Title

    For a while, I also experimented with roff. That was okay, and reminded me a lot of ImagePrint on MS-DOS, but the dot-commands became too distracting. The slash-commands in LaTeX were a lot easier to use (blank line for new paragraph, etc.)

  8. #8 James
    November 7, 2009

    Ian: Greg is hinting at “drinking the Kool Aid”. Hence the reference to the Kool Aid man.

    For others: thought I’d share some screenshots of LyX in action. They also have some video.

  9. #9 Stephanie Z
    November 7, 2009

    Ian, to take it further back to the source, it has to do with the fact that no one knows what Flavor Aid is.

  10. #10 Greg Laden
    November 7, 2009

    James, yeah, my ref to Word in the post is actually to that discussio.

    Ian: “I now realise what that point was. It was the point where emacs began using me, not the other way round (and that point is reached quite early int the curve).” … well put. I love emacs in principle. I just want it to act like this is the twenty-freakin-first century and not 1973.

    The kool aid actually refers to the equivilance of Microsoft and Jonetown. (http://en.wikipedia.org/wiki/Jonestown)

  11. #11 Greg Laden
    November 7, 2009

    I would like to add, BTW, the following:

    1) I did my thesis is DOS Word Perfect then at the last minute fed it into Word for Windows so that I could do some automatic stuff that DOS WP did not do. On Windows 3.1. That was funny. In a “funny strange” kinda way. Didn’t go too well. (Some may find this hard to believe but there was a time when I was a total windows fan boy.)

    2) I was also a Moony for three years

    3) There are excellent LaTex tools that convert Latex into HTML, and I’m pretty sure HTML to Word or Openoffice’s format is pretty good, tough I’ve not messed around with this, and I don’t really recommend it for important project.

    4) I will attempt to use Lxr as a blogging tool.

    5) Regarding #2 above, I was totally kidding. I was never a mooney or even close, tough I must say that as a 14 year old boy (when they started to come around) it was a small thrill to be asked to dinner by pretty young 22 year old girls on every street corner who would start the conversation by giving you their phone number. Then they started selling flowers and totally ruined it.

  12. #12 Mike Haubrich, FCD
    November 7, 2009

    Re: 5) too bad. It would have made for an interesting series of blog posts.

  13. #13 Bill James
    November 7, 2009

    4) I will attempt to use Lxr as a blogging tool.

    nail.meet.hammer

  14. #14 Alex
    November 7, 2009

    I started using LaTeX in college to write lab reports, but quickly moved on to Lyx as it was a lot quicker to use and it was nice to see images in-line. I really loved how LaTeX made text look so nice and professional (like adjusting spacing so text lines up on the right margin, rather than the ragged edge of your typical word document), and equations were so easy to input and pretty. Combined with beautifully vector-scaled eps plots produced with gnuplot, I definitely had the nicest looking lab reports in class.

  15. #15 mark
    November 7, 2009

    The way some of our reports have been processed in the past few years, I guess I might as well do my next one using my Leroy Lettering set.

    I am always prepared to be astonished at what emerges from the printer.

  16. #16 Dan J
    November 7, 2009

    I was introduced to WordPerfect 5.1 in a corporate environment where it ran on our VAX/VMS system. “Reveal Codes” was like turning on a light in my head.

    In a way, the method I use to create web documents is similar. When I write the HTML, I specify what kind of content is enclosed in the tags. When I create the CSS, I specify how each type of content is supposed to be rendered. Well structured content at the beginning makes rendering options much easier later on.

  17. #17 edivimo
    November 7, 2009

    I’m too young and too microsoft-dependent to understand half of what are you talking about… but I’m writing my little thesis of mere forty pages in Word 2003 making cross-references in my discussion to statistical tables that I construct in the middle and in the end of the document… Oh ,delicious kool aid!!

  18. #18 joemac53
    November 7, 2009

    I hate to pull the old man role on you young people, but there was a scientific word processor for the Mac twenty years ago called Mathwriter. I bought a version that arrived in the mail when I had no electricity after Hurricane Bob.

    I still use it. I begged for an old Imac in school so I could run system 9.2 with an old Epson printer I brought in from home.

    I could keep track of versions, have different page layouts all over, footnotes, math stuff without code, nudging things to look better, import pictures and graphs from all over and have a great time.

    They stopped making it. It was too easy, so the tough guys made fun of it. When I retire from teaching high school math and physics this year, and I have to communicate with the rest of the world, my Mathwriter days are over.

  19. #19 Steve
    November 8, 2009

    How does DocBook or XML/HTML relate to Lyx. I have seen DocBook used for a fairly long book on subversion.
    http://code.google.com/p/svnbook/source/browse/trunk/src/en/book/book.xml

  20. #20 Bob
    November 8, 2009

    Somewhere I have a text copy of my thesis, written in LaTeX. My guess is that 18 years after the fact it’ll still render a reasonable facsimile of the original. Try that with a proprietary word processor format.

    When I started using Real Computers(tm) emacs was not installed by default, so I had to learn vi. There was much wailing and gnashing of teeth, but I eventually sorted it out. I’ve tried learning emacs multiple times but every time I try I end up wanting to pull off an arm and beat myself to death with it. Now I use either SciTE/Notepad++ or (g)vim – one of those is guaranteed to work on whatever platform I’m using at the moment.

    In the past year I’ve switched from sysadmin work to hard engineering work so I’m again in an environment where I regularly deal with complex equations. Sadly, the workplace has ‘standardized’ on Word, though half have the abomination of Word 2007 and the rest have Word 2003 and (of course) 2007 saves to .docx by default and 2003 can’t read .docx. So much for standards. Most of our work is formatted in a very specific way (memos, technical reports) and is transmitted via PDF which would make TeX/LaTeX/LyX a shoe-in if only engineers would pull their heads out of their asses and learn something new (I swear, engineers hate new technology as much as educators hate learning new things.)

    As for ‘reveal codes’, I’ve accidentally slagged Word formatting more times than I can remember by unknowingly backspacing over some invisible landmine which shitted up the entire document. People look at me like I’m a luddite for preferring plain text but I have never had to start writing from scratch because I accidentally added or removed a code in a text file. And given how Word 2007’s UI was gratuitously and arbitrarily changed, you’re better off spending the effort learning LaTeX than the latest reorganization of the same old Word functions.

    I don’t hate Word more or less than other word processors, Rather I believe word processors are the wrong tool for almost every application they’re applied to, kind of like duct tape. Duct tape is suitable only for taping ducts; for everything else you need to fix the underlying problem, or in an emergency, use gaffers tape. Cleaning up after a duct tape ‘hack’ is suboptimal.

  21. #21 Dmitri
    November 8, 2009

    would be nice to compare lyx to texmaker
    http://www.xm1math.net/texmaker/

  22. #22 Ian Tindale
    November 8, 2009

    That’s a very good point – DocBook is well worth using. It’s designed for a specific use, but it’s general enough to be shoehorned into other uses with ease. The only down side is that you end up marking in XML as you go along, and as you progress up the ramp of learning DocBook XML. But that’s not wasted learning at all. For me, it’s not a complete fit to my usage, I’d say it’s about a 50% overlap (rough estimate) or less.

    Personally, this exact use case is becoming the reason why I’m learning HTML5. It’s ideal for what I author (and what I assume we are all aiming for). It’s simple enough (which can’t be said of DocBook) and it’s designed very clearly for documents and literature (as well as web apps, but you can ignore that side of it). My suggestion is to look at HTML5 and see if it offers anything you value.

    The critical thing in any of these endeavours is to get the text ‘tagged up’ as close to the point of authoring as possible, in as meaningful a manner possible. How exactly you mark it up is not quite so important, as you can always design and run a transformation to make it into another format (eg, using XSLT), but if it’s tagged haphazardly or meaninglessly, you don’t have a structured document in the first place, and can’t easily turn it into a different representation. This is partly why it’s deceptive to use a technology that doesn’t separate content from style (or presentation).

    PDF appears to fail in this respect: a list item is not in any way related to it’s neighbouring list item, other than coincidentally they share the same typographical attributes, but on the other hand they’ll have differing positional coordinates – it doesn’t ‘see’ the elements as related. A headline is not distinguished as a headline, but merely as some text featuring certain attributes. Another headline elsewhere is unrelated. You can’t derive the structure backwards from a PDF very easily (and hence, you couldn’t run a query on “just the headlines”). But then, PDF is purely an output format. It’s the equivalent of the paper, and not intended to be the editable production substrate. As I point out in my blog post, linked above. Formats that are structured can easily be transformed or rendered into PDF, and this is a good demarcation – a managerially useful (to us) demarcation.

    So, yes. I’m recommending HTML5 as an ideal ‘source format’.

    http://diveintohtml5.org/
    http://dev.w3.org/html5/html4-differences/#new-elements

  23. #23 Paul
    November 8, 2009

    TeXmacs http://www.texmacs.org/ is another scientific document editor. LyX is good but I prefer pure TeX/LaTeX and Emacs + AUCTeX for editing it.

  24. #24 Michael Spencer
    November 8, 2009

    You kids are SO funny.

    I did my thesis on a typewriter. Ever heard of it?

  25. #25 Greg Laden
    November 8, 2009

    Mark [15]: “Leroy Lettering set.”

    (shudder) … memories. I’ll never forget the day I sold my Leroy Lettering set to a bright eyed architect student.

  26. #26 Greg Laden
    November 8, 2009

    edivimo [17]: That could work. Or not.

    But seriously, check your printer now. Is the printer you are going to print this on the one set as your default printer on the computer you are using and/or the Word install you are using? If not, make it so. Is the printer set up with the most current driver (so you won’t be changing or updating drivers later)? If not, make it so. Are the fonts you want to use installed? Good.

    Now, just make sure you type really slow and don’t make any loud noises while the document is open. Don’t take your eyes off the screen while the document is open. Don’t use any other software while the document is open. You might be OK.

  27. #27 greg laden
    November 8, 2009

    Ian: [22]How exactly you mark it up is not quite so important, as you can always design and run a transformation to make it into another format

    Exactly. Marked up/down plain text is always there underneath all of these approaches, and this is why god invented sed.

  28. #28 Greg Laden
    November 8, 2009

    …PDF appears to fail in this respect: a list item is not in any way related to it’s neighbouring list item, other than coincidentally they share the same typographical attributes…

    To me, PDF stands for “Printer Done Fuckingitup” … or “Printer Digitally Foryoureyesonly” … It seems to do a great job of matching what is on the screen/file to what comes out on paper. Makinga PDF file is a one way street from your document stoftware outwards, just like putting it on paper but you don’t need to load the paper, and can email it.

  29. #29 edivimo
    November 8, 2009

    Jeebus Greg! I was more concerned trying make it look right. Let’s see: I learned the relationship between the default printer and word when word always failed to close properly… because the faulty printer driver.
    Where I’m gonna print my thesis? Ummm…
    Fonts? Eeeh…
    Type really slow? Loud noises?
    Wait a minute, you know, in the last months my kool aid have tasted funny…

    [edivimo stop drinking kool aid, open a new tab and search "lyx" in google]

  30. #30 Sam C
    November 8, 2009

    Old fartery is fine, but this aversion to WYSIWYG has the smell of people who think vi is a decent editor and can’t understand why it’s a horrible heap of poo (and why any computer science who doesn’t understand that should get a fail grade!).

    I used roff/troff thirty years ago, and found it useful. But I don’t yearn for it now. Now I use Word because that’s what my employers use. I scream and shout at later versions of Word as much as anybody, but I know several things about it.

    Firstly, what you say about default printers is rubbish; our documents print on many different printers from different manufacturers perfectly OK. In a system that’s not irredeemably badly set up, Windows handles fonts fine.

    Document size in Word is really only an issue when it’s got megabytes of graphics or embedded stuff; you can have lots and lots of plainish text without problems, many many pages. On big, big, complex documents it’s a pain though, and many have come unstuck when seduced by master and sub-documents!

    To mimic text processing’s plus points in Word, the key is ruthless use of styles, carefully set up, preferably cascading sensibly. Then adjustments to styles should propagate through quite neatly. But this approach fails if styles are overridden, other than to shift a bit of text into italics or things like that.

    Cross-referencing in Word works; the default way is a bit messy, but if one wants to be anal, use of bookmarks and reference fields keeps it neat.

    Very few Word users have the discipline to create neat, reliable documents that can bear major hacking without some sort of tidying up (and these folk will often trash your pretty document if you let them near it!).

    So what’s the gripe of the commenters: is it that some like text processing software, therefore word processors must be A Bad Thing because WP treads on TP’s toes?

    There is room for both! In fact, many of our Word users don’t really care what the final document looks like as long as it’s moderately tidy and readable – our reports have a limited technical readership who care about the content and its implications, not about what inter-line spacing has been used or the figure numbering convention.

    Of course, WP can’t mimic some anti-WYSIWYG (because you don’t see until later) features of TP, such as automatic decisions about where to insert tables and figures.

    Your snarky comments about PDF are rather silly; the whole point of PDF is that it’s a fairly dumb one-way format. Most of us just use it as electronic paper, with or without jumpable embedded links. As you say, it’s a great way of sending, receiving and storing finalised documents. And that’s it.

  31. #31 Stephanie Z
    November 8, 2009

    Sam, WYSIWYG is just fine. No problems with WYSIWYG. The problem with Word is that there is no good way to reveal most of the underlying formatting when I need it. That means that I can’t effectively troubleshoot a document created or edited by one of those people who doesn’t care what something looks like. That means I can’t troubleshoot a document that’s suddenly tying up printing on my machine. That means I take a blind risk every time I combine pieces from multiple documents, particularly if they have multiple authors. That means I spend a lot of time reformatting after taking things like table text back to its basic styles because I can’t see how much crud (i.e., unnecessary complexity leading to higher likelihood of the document becoming corrupted) other users have introduced by formatting the thing one cell at a time.

    So you don’t have to do that. So most of the people you work with don’t care about being able to do that. So what? I do. Never having used vi, or any text-markup program since Paperclip, my problems with Word have nothing to do with fuddyness. They have to do with the fact that Word is far more program than the vast majority of users should be faced with using–or waiting to launch–while not having the options that advanced users require to clean up the messes left by everyone else poking at things. It’s neither fish nor fowl.

    And consider yourself very lucky that you haven’t run into the printer repagination issue. I suspect the fact that you haven’t is due to HP’s business practices (including monopoly) more than Microsoft’s.

  32. #32 george.w
    November 8, 2009

    I have a related problem in which the academic people I support just can’t seem to make an entry on a web page or blog without using (shudder!) Adobe Contribute. And most often they compound the problem by typing in Word, then copying and pasting directly into Contribute or Expression Web or some other damn thing.

    These are people with PhD’s… they should be able to learn rudimentary HTML in, like, an hour. And file management too, because they have only a dim idea of file paths. If it doesn’t have a “drive letter” they don’t know where it is.

    Some of the CMS systems we’re looking into now have a “Paset from Word” button that strips off superfluous formatting. I’m looking forward to the challenge of getting them to use that button.

    /rant

  33. #33 george.w
    November 8, 2009

    “Paset from Word”
    Doh! Muphrey’s law strikes. Should be “Paste from Word”.

  34. #34 Paul
    November 8, 2009

    “Old fartery is fine, but this aversion to WYSIWYG…” –Sam C

    Nothing to do with old fartery, Sam. At least not as far as I’m concerned. Pure text processing has its advantages anyway but TeX is a typesetting program, not a markup format. It is a good choice for all purposes if appearance does actually matter to you and you are not an expert typographer. The WP versus TeX debate has been done to death on the interwebs (see e.g. http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/latex_advocacy.html ) and there is a place for both, as you say. However, for some purposes (maths/physics/…) there isn’t really: (La)Tex is a no-brainer.

  35. #35 Enoch
    November 8, 2009

    Old fartery is fine, but this aversion to WYSIWYG has the smell of people who think vi is a decent editor and can’t understand why it’s a horrible heap of poo

    Actually, but if you read the post itself you will see that this is not what is being said here.

    I scream and shout at later versions of Word as much as anybody

    But nobody screams and shouts at and of the versions of LaTex or LyX. ‘

    Firstly, what you say about default printers is rubbish; our documents print on many different printers from different manufacturers perfectly OK.

    It is not the truth that every different printer and driver = a different result, but that is not what is being said by anyone. That is it unpredictable is true, not rubbish at all.

    In a system that’s not irredeemably badly set up, Windows handles fonts fine.

    This what we call in my business the blaming of the victim.

    Document size in Word is really only an issue when it’s got megabytes of graphics or embedded stuff; you can have lots and lots of plainish text without problems, many many pages.

    Why would you chose a software with this limitation?

    On big, big, complex documents it’s a pain though, and many have come unstuck when seduced by master and sub-documents!

    This is what we call in my business a “kludge”

    To mimic text processing’s plus points in Word, the key is ruthless use of styles, carefully set up, preferably cascading sensibly. Then adjustments to styles should propagate through quite neatly. But this approach fails if styles are overridden, other than to shift a bit of text into italics or things like that.

    Yes, and here you are describing to us both the actual use and the enforced use of XML, docbooks, or LaTeX Markup, imperfectly and unpredictably imagined by Microsoft Word.

    Cross-referencing in Word works; the default way is a bit messy, but

    but but but but but

    Very few Word users have the discipline to create neat, reliable documents that can bear major hacking

    This is what in my business we call blaming the victim

    So what’s the gripe of the commenters: is it that some like text processing software, therefore word processors must be A Bad Thing because WP treads on TP’s toes?

    I don’t think so. I think people mostly think that a toy like Microsoft Word has the use of making one pagers and memos, but for longer or more complex things you take chances, and the longer and the more complex the worst the chances. So why bother? There should be a two page limitation on Microsoft Word documents in all businesses or agencies which do important things.

    There is room for both!

    Yes, with the two page limit, yes!!!

    Your snarky comments about PDF are rather silly; the whole point of PDF is that it’s a fairly dumb one-way format.

    As I read the above that is just what was said.

  36. #36 Amy
    November 8, 2009

    I love that “LaTex is a ‘No Brainer'” for rocket scientists!

  37. #37 W. Anderson
    November 8, 2009

    The article and comments are very informative about the attitudes and practices of all using various document or work processing techniques and software.

    I have no expertise in this area, although having used most, if not all of the software, and LyX only of late.

    It is unfortunate that the few who postulated about Microsoft Word really could not respond intelligently against the technical and practical benefits and advantages put forth for TeX/LaTex,or LyX.
    Instead all they do is introduce many “it works, but…” or “one can get around this limitation if….” comments which makes the discussion superfluous.

    W. Anderson
    wanderson@kimalcorp.org

  38. #38 abarbarian
    November 9, 2009

    http://antrix.net/stuff/writr/

    In keeping with the KIS philosophy of GNU/Linux why not give “writr” a try. Easy peasy to use and you see what you have produced in one click. :-)

  39. #39 Greg Laden
    November 9, 2009

    writr looks like a very interesting and useful project in its early stages.

    I would like gedit with the ability to permanently (as it were) specify language, autospell, and markup “view” setting so it opens all the time as configured, and a plug-in that does a preview in a dvi, pdf, or html viewer.

    make writr act like that and I’ll get two!

  40. #40 sgtrock
    November 9, 2009

    I’ve toyed with LyX on several occasions over the past 10 years or so. Honestly, I love the way that the workflow of writing works with it. However, I keep tripping over three issues that I’ve never been able to successfully resolve to my satisfaction.

    The first is the need to drop into what is known fondly as ERT (Evil Red Text) to get some formatting options to work correctly. ERT is just TeX and/or LaTeX markup, so those who are comfortable with it can still just keep writing. Unfortunately, those of us who don’t spend enough time writing to learn LaTeX well end up struggling.

    Browsing the forums shows that this particular problem has largely evaporated for most needs as style files have matured over time. That is really good news!

    However, there is another problem that I think inherently limits what can be done with LyX. Rather than use a pre-existing markup language like XML or SGML, the developers of LyX chose to create their own. This obviously limits the portability of any document created with LyX.

    Worse from my point of view, it means that you must export to target platform(s) for final output. Not an issue for a single author. However, collaboration now requires that all members of a team have access to LyX, and preferably a very current version to avoid the aforementioned issues with ERT. This is a big deal for me as much of the content creation that I do at work is in conjunction with any number of people from a variety of teams. Combining their efforts with mine is frustrating enough that I’ve pretty much given up the effort. If I’m writing a standalone document where I know the final target output format(s), sometimes I’ll try LyX. Unfortunately, that’s a pretty rare use case here.

    The final issue that I have with LyX is the lack of separation of presentation from content that both Greg Laden and Ian Tindale (#22) noted. Again, this limits where I can use LyX effectively. Writing content that then becomes an element of multiple compound documents is a challenge at best and a nightmare at worst.

    Reading through what I’ve written so far, I can see how people would think that I don’t like LyX. Nothing could be further from the truth. I wouldn’t keep coming back to it if I didn’t like it a lot. :) The difficulty that I have is that the kind of writing that I do for a living doesn’t fit well within the kind of use case that LyX is designed to address. I still use LyX for some of my personal, hobby oriented tasks.

  41. #41 Jan
    November 9, 2009

    I wrote my masters thesis with LyX. After having used MS Word, OpenOffice and plain LaTeX for my university documents I gave LyX a try and it worked perfectly for my masters thesis.

    A friend of mine used Word 2007 for her thesis and it was a mess. The document looked ok, but she had real trouble getting the document to look like she wanted/have to. I only had to make the margins and the fonts I wanted to use, link my BibTeX-file and I didn’t have to do anything in the document for the whole six months,

  42. #42 mario
    November 9, 2009

    I take offense that OpenOffice is compared to fucking Word here. While OOo is still just a WYSIWYG editor, it has a few more capabilities than MSWord. It’s actually feasible and easy to structure a document in OpenOffice; there are real style classes. And that’s probably why there is a LaTeX-converter for it.

    For anything with more than 100 pages, real books and scientific stuff; it however sounds more senseful to start with TeX right away.

  43. #43 Greg Laden
    November 9, 2009

    Good point about OpenOffice

  44. #44 Nevyn
    November 9, 2009

    LyX has been something I’ve been pushing for a while – before my big anti-spreadsheet sentiments.

    And the way I sell it is this:

    Your computer is smart enough to work out formatting itself. It’ll do a much better job of making things like consistent and lining up tabs etc. So why would you restrict yourself to doing things in a way that hinders the computer’s ability to do this?

    As for spreadsheets:
    There’s only 3 reasons to use a spreadsheet:-
    1. One off’s
    2. Prototyping
    3. Doing a half arse job
    What we need is a decent framework on top of a database that allows people to make up interfaces themselves or to offer the service ourselves….

    There’s surely a better way…

  45. #45 Vincent Manis
    November 9, 2009

    I’m a TeX old-timer, having learned it in 1982 from a listing of the program (before the TeXbook was published). I still sometimes DO swear at it: Knuth’s macro language is so perverse that sometimes one finds oneself doing heroic things just to make it do something trivial in a sensible programming language. I never swore at it for the reasons coworkers swore at Word or FrameMaker (the 2 systems my colleagues seemed to like); when they couldn’t make a simple typographic effect work. I normally swear at it when my notion of what’s clever doesn’t match Don Knuth’s. :)

    I don’t use LyX myself (I like and understand emacs), but I always tell newcomers to LaTeX to use LyX, as it gets them almost all the way there.

    As for LaTeX not supporting content/presentation, I have no idea what that can mean. Most of what I used LaTeX for was producing overhead transparencies, and I evolved a document class file that had everything in it. There was basically NO presentation-level content in any of my documents; I left it all to the document class file (and if you read Lamport’s original LaTeX book, you’ll see that’s exactly what he intended, though he never told you how to write document class files).

    In fact, at one point, I had a 600-slide course that we wanted to turn into a set of lecture notes; remove all semblance of slides, turn the whole thing into sections and subsections, bulleted lists, etc (this saved a lot of paper over 4up reproduction of the slides, but kept all the content). Total time to do this was about an hour.

    It does take some self-discipline to work this way, but it is more than possible to separate content from presentation in LaTeX (and therefore in LyX).

    LaTeX->HTML is unfortunately still not a completely-solved problem (except in some commercial TeX systems), but I’ve always found `good enough’ solutions (generally using custom CSS files) for my needs.

    I am quite prepared to throw LaTeX away and replace it with something better; I’m just not prepared to replace it with something worse.

  46. #46 Nevyn
    November 9, 2009

    Addressing OOwriter being compared with Word:

    Let’s face it. Open Office is aimed at being a drop in replacement to MS Office. Sure, it has some more capabilities. Hell, even MS Word allows structured documents. My main problem though is that each of them still present you with a way of manipulating individual font’s etc. more easily than using styles. It doesn’t do anything to save you time.

    LyX however, ONLY gives you the option to do it right. Because of this you aren’t fussing about with formatting – at all.

    If only Open Office could present you with a view which only showed you styles rather than the myriad of options that do nothing but stuff up your document. I’ve a friend who doesn’t show the formatting toolbar at all but then he has to have the style picker open all of the time. If you could remove the formatting toolbar and have the styles show up in a toolbar like fashion, Open Office might be a viable alternative to LyX.

    At the moment, it’s not.

  47. #47 Richard
    November 9, 2009

    I’ve used word, lyx and oo writer.
    Most of the articles I write are for academic publication in economics/finance journals. They have tables, some formulas and run about 30-40 pages.

    A few observations.
    1. The layout and formatting is largely irrelevant. When I submit to a journal, they will format it according to their requirements. Therefore I tend to keep formatting as simple as possible. I use word XP.

    2. Lyx is nice that you don’t have to worry at all about formatting, but the ERT is annoying. I wasted a huge amount of time just trying to get my front page set up with the authors names and affiliations in line. Also I don’t want numbers on the first page. I know lyx can do this – but do I need to spend hours trying to figure out how. Back to word XP.

    3. My co-authors use word. We put the doc up in dropbox
    and use track changes. It is easy. Word XP.

    4. The default for journals is either word or pdf. Some want latex, but they are rarer.

    At the end of the day, I appreciate what Lyx does, but I can reliably get word to format as I need and then I dump it out to a pdf to share. So I know what others are seeing.

    I’m not a MSFT fan boy either. I hate office 2007 with a passion. I exclusively run ubuntu on my work machine, but I use wine to run word xp.

  48. #48 Jason Waskiewicz
    November 10, 2009

    It’s interesting how emotional this argument gets!

    In my career I’ve worked (in order) with TI-Writer, MS Works, Lotus WordPro, Appleworks, Word, OpenOffice Writer, LyX, and LaTeX. They all have their strengths and perhaps the goal should be to match the tool to the job.

    Most people do not use their word processor correctly. A few years before I stumbled upon LyX, I started using styles with OpenOffice. They were great for consistency. LyX merely takes away the option of “finger painting” one’s styles on the fly. This is not a strength of LyX.

    I hate Word’s equation editor. It’s slow to mouse-click a complex equation. I prefer to type it in, even if I must learn some coding. OpenOffice did this. So does LyX. Word probably could without much effort.

    What truly sets LyX and LaTeX apart is that the styles and typesetting were designed by professionals. My OpenOffice stuff made with styles looked good compared to my old finger-painted stuff. Everything I make from LyX or LaTeX looks professional.

    Maybe I could learn to change my OpenOffice styles to emulate what is already in LyX or LaTeX. However, experts have already done the work for me. I use LyX and LaTeX primarily because the output just looks good.

    Some secondary benefits are the small file sizes, the options available (tests which are easily randomized), margin notes, and the speed of the program. My computer is getting elderly and cranking up the newest version of OpenOffice is an exercise in patience. It takes forever to load. Typing takes forever (the lag) as well. LyX and LaTeX spring to life right away, and move easily between the 3 operating systems I work with daily. It may be this last feature that most locks me into LyX and LaTeX.

  49. #49 Paul
    November 10, 2009

    @Richard

    “Also I don’t want numbers on the first page. I know lyx can do this – but do I need to spend hours trying to figure out how. Back to word XP.”

    To which a LyX + (La)TeX user might respond, “No! Back to (La)TeX – 10 seconds to find out how!”

    But of course – as is clear from your observations – you neither need nor want (La)TeX and you certainly don’t want to be fighting with something deliberately designed to deter you from doing the kind of tweaking you want to do. A word processor is exactly right for your purposes, LyX and (La)TeX are not.

    That’s one of the pitfalls of a GUI front end to (La)TeX such as LyX: it looks quite like a word processor and you feel you ought to be able to do WP-ish things with it, but it isn’t and you can’t (at least not easily).

  50. #50 Richard
    November 10, 2009

    @Paul – I agree with your observations.

    I would love to use all open source software, and I think Lyx is an excellent program, but the gains from having absolute control over the document (as in LYx) vs. compatability with other users: coauthors who want to edit a paper, editors at journals who want submissions in word, and students who want class notes in a form that they can edit electronically, make word my only choice. I wish it wasn’t so.

    Anyhow, I will continue to use LYx from time to time, in part because I really find the process more satisfying. You just type what you want and then do the formatting later.

  51. #51 Paul
    November 10, 2009

    @Richard,

    Well (La)TeX is the electronic lingua franca in mathematics and physics so perhaps you could move into econophysics. ;-) Your position would then be completely reversed w.r.t. (La)TeX and journals etc. and you’d soon find yourself becoming utterly exasperated at students sending you barely readable and uneditable equation editor-made .docs instead of (La)TeX sources or output. :)

    Seriously though, I think it’s crucial to realise that LyX is not a word processor and isn’t trying to be one. It’s a front end to LaTeX, LaTeX is itself just a sort of front end to TeX, and TeX is an output oriented document typesetting system and language – not a word processing core. If LyX can enable some people outside of the mathematical sciences on the input (and output) side, and publishers and aesthetes on the output side, to benefit from the excellent TeX system, that’s splendid, but it’s hardly surprising if it fails for many as an alternative to a word processor.

  52. #52 Greg Laden
    November 10, 2009

    Word is not the universally preferred format in the publishing world as far as I know, though it is increasingly an option. And, converting from LaTex to word should not be difficult, using OpenOffice or a couple of simple commands. I’ve not experimented with that yet, but there is a LaTeX to HTML conversion and HTML to DOC is easy.

  53. #53 Ian Tindale
    November 10, 2009

    For the past couple of decades or more until fairly recently I’ve been a graphic designer / art editor in electronic publishing, mainly involved in putting magazines together as well as many other types of publication. My tools started with phototypesetters (Bobst Eurocat, Quadritek) then various DTP solutions as they started to appear, culminating in QuarkXpress, InDesign, FrameMaker. Our first prefacing procedure as we load up an author’s “word” file is to strip all formatting and style from it as we take a gulp of coffee ready to start the real work. We style it all up ourselves, unless we know an article is from one of those rare authors that know about DTP styles, and their “word” implementation features the same styles in name and structure, not necessarily in rendering intent – when it hits our DTP, it picks up the same style at our end and looks like we want it to upon import. For most authors, though, it’s a case of getting rid of what they thought it will look like on publication and putting in a sensible structural style treatment instead – one that works. Naturally, any graphics they’ve embedded in their word processor file will be lost, and tables need individual treatment, and tabs – don’t get me started.

    Okay, I will. Authors don’t know how to tab. They don’t know anything about tabs. They think that the tab key just moves the cursor over by a predictable jump, so they’ll lay out a tabular portion of text by typing a bit of text in the first column, then hitting tab, then the bit of text in the second column, then hitting tab, and the bit of text in the third column, etc and so on until they’ve got all the columns. Then they’ll go on to the next line.

    But at some point the first column might contain something a bit longer than one tab column, so that’ll push all the other tabs to the right. So they go through their prior tabbing, adding more tabs to make it look right on their screen. They’ll add many tabs where there is no data in a part of a column, to get to a far column. As long as it lines up on their screen, they think they’re doing it right.

    But when it gets into FrameMaker (for example) there’s no ‘columns’ – only a set of text with tab characters (initially invisible) sandwiched inbetween them. I then set up some tab columns for the tabs to line up to. Except of course that I’ve no idea how many tab columns there should have been. If I end up putting in eighteen tab positions, I might find all of them are used. Except that the data in each column is scattered across the table apparently randomly. It might turn out that there was in fact only six columns in the table, but the author simply kept hitting intermediate tabs to get the six columns to ‘line up’ because of the different sized content in some of the columns. They lack the knowledge regarding how to actually set up tab columns in the word processor that they earn their living from. They’ve simply left the default columns in. They haven’t cleared the tabs and drawn up their own columns that are correctly sized to accommodate the text or data they’re typing into these columns, and hence, they’re in fact scattering the text across a bunch of fictional columns that simply shouldn’t be there. But I’ve no way of knowing, unless they’ve printed it out and given me the printout (a practice that used to be more common than it is currently).

    If any of you are writing for a publisher, and are not going to be responsible for the styling and layout, then don’t style it at all for layout. Simply give the styles structurally meaningful names (preferably that match the style names of the DTP people at the production end – ask them). Your headlines, subheads, body, lists, etc should look like you want to see them on screen for the purposes of writing. When they’re imported, those same structurally meaningful style names will match up to the DTP people’s stylesheets and come out looking how they want it to.

    If you’re responsible for your own rendering and output, then it’s a different story, but nevertheless, I’d still say to leave all styling and layout to the last step. Authoring, editing etc should be done in a typographic environment that you actually like. Laying out the page for print is a completely different task, and is actually quite a pain for the subeditor to have to work in once it’s styled (though not impossible, and not so difficult that they’d otherwise have to learn the easy to use inline text editor in InDesign).

    By the way, in Linux, Scribus is actually quite a competent DTP package – I’m used to the top end DTP packages professionally, but looking at Scribus, assuming I didn’t have to do any colour-critical four-colour work (Linux ICC profiles are still the weak area), I wouldn’t hesitate to use it if I had to (which I wouldn’t, because it removes interoperability with the rest of the design world – but let’s say it was an isolated task, then I would say yes). But of course, you wouldn’t start authoring directly into Scribus. That’d be insane. Possible, of course, but ridiculous. We don’t do that. But the import text format might be from something fairly straightforward, such as tagged-up copy generated in Vim, Gedit or whatever you want. However, don’t underestimate what an absolute pain tagging things up becomes – especially as you’re writing. XML is a container format, whereas the old inline slash codes aren’t (necessarily – until you cancel one). Doing things in DocBook or similar can become masochistic. Open Office or Abiword now starts to look like a sensible and easy way of applying styles, by comparison.

    There are no clean-cut solutions to all of this, and to be honest, I don’t think there are supposed to be. The only guidance I’d offer is to author in something as low-tech and comfortable as possible, which offers the opportunity to structurally mark up. Don’t use any clever features, don’t embed any media that isn’t text, and don’t rely on the way your particular page or screen is set up to you right at this moment.

    Personally, I write a lot of my stuff these days in Google Docs. I even did my dissertation in it (it had just been released at the time, having disappeared from being ‘Writely’ a short while earlier) (I authored it as separate chapter docs, then put it all together at the last minute in a different program). It has just about the right scale of features for me. I don’t want it to do any more than it presently does.

  54. #54 Ian Tindale
    November 10, 2009

    Oh, and I’m against trying to write an entire book or something of that scale as one huge file. Of course, it can be done, and DocBook etc are happy to offer pagination solutions to demarcate chapters, sections, etc. The reason I’m against it is nothing more than the same reason I’d learned to never write the label on a floppy disk before copying the data onto it. If you do it that way round, the floppy disk will definitely end up defective and you’ll only throw it into the bin, with your nice neat written label on it, laughing at you. Leave it until you can believe it’s copied properly and the disk works, then write the label.

    Same with the huge single integrated document files. They always get some ridiculous and show-stopping corruption or EOF error or something, which renders the whole thing a waste of time, laughing in our faces. Authors that chop the work up into separate chapters are on safer ground. Much safer. That’s the only reason.

  55. #55 Leandro Mattioli
    January 27, 2010

    Ian, I’ve read “Where word processors went wrong”. That was an interesting view of the problem. There are no bounds between creation and consumption nowadays, but we should consider that many documents must be prepared by a group. In the cases where each person is responsible for a section of the document, the support for modularization could be enough. But there are situations that you must make a revision of other people’s content (generally, they send you a poor content and you have to decide if you want to make it better!! :) ).

    I’ve used LyX with a team to write a small book. Each member was responsible for one chapter. Using the “child document” feature of LyX, we made id! The cool thing about it is that there were no specialists in LaTeX in our group, and LyX do the basic typesetting just fine.

    About the cover page, although I’ve accomplish to do it in LyX, I guess it could be done with another tool (maybe a WYSIWYG word processor! :) ). We have tools suited for each purpose.

    I have deeply used the styles features of Microsoft Word and I do believe that it is possible to make astonishing documents in MS Word. But it’s very easy to do something weird in MS Word, and that happens o lot. Many people align text with whitespaces (TeX blocks two whitespaces, unless you explicitly “ask” for it), most of people don’t use styles appropriately, most of people don’t use page breaks and insert a sequence of “newlines” in order to put something in the next page. I had to format other people’s documents lots of times and in big documents this is a real problem. I’ve got a material with almost 150 pages and MS Word was crashing trying to open it. The solution was to open in OpenOffice Writer and it was very slow.

    In my opinion, LyX is still a work-in-progress. But it’s already enough for most of my typesetting. I think the perfect processor would let you markup the document and (re)define some formatting, just like HTML+CSS try. The feature I most miss in LyX is the ability to format Sections, Subsections, etc. For example, I’d like to define bottom border for sections without any LaTeX command.

    About the LyX file format, it is almost LaTeX, and LaTeX is a well known format. No problem with it at all…

    When using LyX, I’m not so sad to sacrifice some formatting features, considering all the benefits I have… You can include Chess Games, music scores, chemical molecules, electrical circuits and whatever you want, without having to export these elements to figures. This is great! Also, you have lots of templates, including “holywood”, “curriculum vitae”, etc… BibTeX (for references) and the math typing stuff are so cool that many people start using LyX just because of these features.
    You can include code and enable the syntax highlighting (adding some LaTeX you can do this with colors).

    And with LyX you markup documents without losing the GUI facilities. (And yes, people like front ends).

  56. #56 Greg Laden
    January 27, 2010

    Lyx might be a work in progress (and I tend to agree) but LaTex is not, and Lyx is just a tool.

    In collaboration, there is no reason that every member of a group has to share the responsibility of formatting. IN fact, that is a nightmare in some cases.

  57. #57 Michael
    January 6, 2011

    So if most people do not us MS word correctly, the simple question is:-

    What is the best MS Word tutorial available?

    Mick

  58. #58 Greg Laden
    January 6, 2011

    Michael: Seriously? I’d head right to Woody’s and see what they have, between their books and their web site, assuming that operation is still up and running (Woody’s office watch, Woody’s word watch, Woody’s Windows watch, etc.)