Like it or not, everyone uses MS Word. Very few places have a specialised wordprocessor for publishing, and they usually have MS Word for technical staff and SMEs to write the drafts. Most web sites have their content written in MS Word. A republishing tool for the web or online help or anything else that does not take MS Word docs as input is just not used. Get used to it. Bob T On 21 April 2011 08:36, Christine Kent <cmkentau@xxxxxxxxx> wrote: > James, you are ignoring the organisational reasons why a business goes with > a single product. Only large and dedicated tech writing departments can > really benefit from having a dedicated software, and for them it is valid > to > argue for such. However, in most organisations I have worked with, the > materials I produce have to move backwards and forwards between myself and > the corporate wide users. Messy and totally horrible though it often is, > particularly for compiling large documents, most organisations will make > the > call to have only one product. > > To be fair to MS, the changes made in recent years have led me to believe > that MS is not remotely interested in developing Word as a tech writing > tool, nor do they ever promote it as such. They do not even have a "book" > template, meaning they are not promoting Word even for the production of > books, which it does perfectly well. > > All the significant changes of 2007/10 have been to make it more user > friendly for the average user. Features that are important to high level > niche users, such as master documents, cross references, mail merge and the > like have not been touched since at least 2003 and probably go back very > much further. This neglect implies quite strongly that MS is not interested > in them. We can only speculate as to whether they will eventually improve > them in one big hit, or perhaps come out with a niche product, or whether > we > are too small a component of their market for them to care less what we > think. > > Christine > > > > -----Original Message----- > From: austechwriter-bounce@xxxxxxxxxxxxx > [mailto:austechwriter-bounce@xxxxxxxxxxxxx] On Behalf Of James Hunt > Sent: Wednesday, 20 April 2011 11:52 PM > To: austechwriter@xxxxxxxxxxxxx > Subject: atw: Replacing Word (Long...) > > The proposition that Microsoft Word is not a good tool for technical and > corporate writing is widely accepted, and the shortcomings of the product > have been described in detail in many places. According to a British > technical writer in a recent blog comment - > > "Once again, someone has pointed out the fact that MS Word is not really > suitable as an enterprise technical writing tool unless it is heavily > customized. Have programmers create special document-creation interfaces, > create locked purpose-built ribbons, use SEQ fields instead of native > numbering... Here we go again. I've seen and heard this exact same thing > too > many times... People keep saying Word will work for tech writing IF you > change this and customize that and renoberate the other thing. Face it, > this > is like saying that a car will work as a helicopter if you totally rebuild > every part of it. If you have to spend that much time, effort, and money > making it work the way you want, it would be worth your while if only in > avoided irritation to go get something that already works the way you need > it to work. And if you have to try to change your corporate culture to make > the tool usable, it's NOT worth the effort. You are supposed to use tools, > not the other way around." > > Outbursts like this, and there have been many of them over the years, never > seem to change anything. We keep on using a product that has serious > problems, and reduces our working efficiency quite noticeably. > > I do not propose to add another lament to this pile. Instead, I propose to > point out that a superior alternative exists, that the alternative is > stable, is bug-free for all practical purposes, and has been tested by > millions of writers over several decades. Furthermore, it is free, and in > use involves far less effort than does Word. I will describe in a little > detail how this tool, in conjunction with a freeware version control system > and a few other tools, can be used to perform documentation tasks more > efficiently than we could with Word. > > Technical writers are usually in control of a complete end-to-end process: > we research our material, write it, rewrite it, and eventually check the > formatting and deliver documents to end users. > > Are there any other writers like us? The answer is yes: engineers, > physicists, other physical scientists, mathematicians, and even computer > scientists, who write journal articles, books, and reports for publication > in scientific journals or by scientific book publishers or by their own > employers. These writers are found largely, but not exclusively, in > universities and research institutes. There may well be a few million of > them - no-one knows for certain. > > These scientific writers all use a simple markup language and typesetting > program called LaTeX, which has been around since the 1980s, and is based > on > a system called TeX that was developed in the 1970s. (LaTeX is not used by > all scientists: biologists rarely use it, but it is also used by > non-scientists: I have come across theological tomes typeset in multiple > languages with LaTeX.) > > Microsoft Word is a WYSIWYG system, which means that text is typeset as it > is entered: the writing and typesetting tasks are conflated in a single > window. This has the unfortunate result that the input file is continually > modified by the typesetter, and errors in the typesetting process will > instantly corrupt that input file. This is not the case with LaTeX: writers > write in text editors, not word processors, and insert formatting tags into > their text in the form of text strings after escape characters (such as > "\chapter", "\section" or "\list"). Input files are pure text files, with > no > hidden formatting information. When a typeset version of a file is > required, > the text file is sent to the background typesetting program, which reads > the > tags in the text as commands and produces the required PDF output. (This is > not hard: usually clicking on the "Typeset" button on your text editor will > do the job.) The input files remain unchanged by this process. If the > result > is unsatisfactory, will have still have your unmodified input text files > for > checking. > > The typesetting algorithm in LaTeX is much more sophisticated than that > used > in Word, and the result is quite noticeably better: it is even better than > anything that Adobe InDesign or QuarkXPress can do. In fact, LaTeX > typesetting is of book-publication quality, and is used by many technical > publishers. > > Is it possible to learn from all of those LaTeX-using writers, and build on > their experience? Yes, and I will argue that we should do exactly that. > LaTeX is not hard to learn: you can learn the basics in a day and become > quite skilled in a week. The mastery so obtained is yours for a lifetime, > because the core of LaTeX does not change. Books on LaTeX are often > dauntingly thick, because LaTeX was developed for typesetting mathematics, > and large chunks of LaTeX texts are devoted to that topic. We don't do > that. > We don't need to learn about the complexities of academic referencing > either, because most software and engineering user guides do not contain > bibliographies and citations. When these items are left out of the average > LaTeX book, the remainder comprises a discussion of about fifty tags - not > many more than basic HTML. This is what is learned in a day. Going beyond > the default style options, which give very good results anyway, will take a > few more days' work. Authors of books on LaTeX are often apologetic about > the "steep learning curve" involved, but it is steep only if you need to do > some mathematical heavy lifting. > > This site is a good start: http://latex.silmaril.ie/formattinginformation/ > > A PDF version is available at: > http://latex.silmaril.ie/formattinginformation/beginlatex-a4.pdf > > For instructions on downloads, see: http://ctan.org/starter.html > > For a more substantial reference, see: > http://upload.wikimedia.org/wikipedia/commons/2/2d/LaTeX.pdf > > I won't duplicate the material in those sites here. The core of LaTeX is > small and easy to learn, and if you want to do something outside the scope > of the core, then you will need an add-on package designed to do that task. > There seem to be thousands of these, all designed for a particular purpose: > drawing Feynman diagrams for physicists, adding hyperlinks to the PDF > output, setting out manuscripts for science fiction and fantasy publishers, > making page margins visible, writing in all manner of alphabets, and so on > and on. You can ignore most of them. > > Our basic products are books and book-like objects, ranging in length from > a > few pages to several hundred pages: user guides, installation guides, > reference manuals, and so on. These were once actual printed books, but are > now mostly PDF files on a server. We also produce lots of ancillary > material, most notably on-line help systems, but also quick-reference > guides, instruction cards, training manuals, presentations, and so on. All > of these can be produced using LaTeX. > > ---------- > Tools - summary > ---------- > > The tools required for writing our materials comprise a text editor, a > LaTeX > distribution (the basic typesetter and a collection of interesting add-on > packages), a version control system, a screen capture utility, and a > graphics editor. All of these are available as freeware. > > ---------- > Text editors > ---------- > > There are quite a few editors, and they are interchangeable because they > all > produce only text files. Most are freeware; some are minimalist (TeXShop), > and some have more buttons and control panels than a nuclear power station > (texmaker). TeXShop, TeXworks and Kile are commonly used. Prehistoric > products like emacs are still around, and do the job quite nicely. > > ---------- > Revision control > ---------- > > There are a number of freeware revision control systems available. > Subversion may be the most widely used, but there are others, such as Git, > that are just as good. Revision control systems are most useful when > handling plain text files, and these are what text editors produce. > > Revision control systems are useful if successive editions are required for > successive software releases, or if several writers work on a single > project. > > If you are working on a software development project, the simplest approach > may be to use the system selected by the project manager. > > ---------- > Illustrations > --------- > > Illustrations in various formats may be included by reference in LaTeX > documents: PNG, JPG, and PDF screenshots may all be used. > > The open-source scalable vector graphics editor, Inkscape > (http://inkscape.org/), has many of the capabilities of Adobe Illustrator. > Inkscape illustrations can be included in LaTeX documents. > > ---------- > What about XML? > ---------- > > The major problem with XML tagging is getting the files to print. Now a > major feature of LaTeX is that the tags inserted in the text are later read > as commands by the typesetting program. This is not the case with so-called > XML formatters: some of these work by substituting LaTeX tags for XML tags, > and then processing the generated LaTeX files in the usual way, while > others > have their own typesetting programs - which never seem to be anywhere near > as good as LaTeX. The former approach merely adds an extra processing step, > and the latter reinvents (part of) the wheel. > > ---------- > How far can this model of work extend? > ---------- > > The work model described works very well for a single writer producing all > the documents required for a project, where that writer may or may not be > using a version control system. Can the model be extended to projects with > multiple writers, or even multiple writers producing multiple documents in > a > range of languages? Of course. > > ---------- > Multiple authors writing different chapters > ---------- > > Let us suppose that you are the lead writer on an enormous software > development project, and that you have assigned different chapters of the > various books to several other writers. The writers (and you) write in > text > editors, and at regular intervals they test their work by making sure that > it can compile to PDF without errors, and then commit the text files (not > the PDF files, which are only temporary) to the version control system. You > may then extract copies of the various updated files from the version > control system, and compile the full PDF versions of the various manuals. > (This process could even be automated as part of the daily build process.) > The write-test-compile cycle is the same as the one that characterises > software development. > > ---------- > Successive releases of a software product > ---------- > > Let us suppose that your company has released version 1.0 of the software > and started to develop version 1.1. The files under version control are > suitably marked as "Version 1.0", and the writing process continues until > development is frozen and the final PDF documents for Version 1.1 are > produced. > > ---------- > Managing translations > ---------- > > If your company decided to localize the product and sell it from Iceland to > China with local-language manuals, how would you go about preparing those > manuals? > > First of all, the text strings displayed on the various windows of a > software product are all stored in one file, and imported to the program > windows by reference. To produce other-language versions of the software, > new files of text strings must be prepared, and the user must be able to > select a language and store it as a preference. Thus the software > developers > can produce twenty or more different language versions of the product, and > it is up to you to provide the required manuals in those twenty-odd > languages. > > If you used a number of screen shots to illustrate the base-language > manual, > then you will have to capture the same screen shots for all the other > language versions. A tedious business. Name the new screen shots > systematically - for example: "screenshot_01_1_de.png", > "screenshot_19_1_fi.png", ... (If you used Inkscape to prepare drawings, > and > used LaTeX tags to label them, then the text in the drawings may be > translated with the text documents.) > > Send copies of the base-language text files (for Version 1.0, say) to the > chosen translation house. The translators will not need to change the tags, > only the text: LaTeX tags remain in US English no matter what language is > used in the document. Eventually, you will get back sets of text files of > the translations. Edit the various "\includepicture" tags to make sure that > the screen shots all go into the right manuals. > > Next, compile the PDF versions. There will be errors, no matter how many > times you tell the translators to Leave the Damn Tags Alone: correct the > errors, insert a tag specifying the language at the start of each master > document (to ensure that the correct hyphenation algorithm is used for each > language) and compile, and you will have a collection of elegantly typeset > manuals most of which you probably cannot read. > > Producing the first set of translations is hugely expensive. However, when > it is time to produce the various translations of the Version 1.1 manual, > the cost can be much lower. The aim is not to translate anything that was > translated before, and translate only the changes made in the base-language > version. The changes between Versions 1.0 and 1.1 can be pin-pointed by any > version control system. The translators can find the matching lines in the > foreign-language text files, make the required changes, and save the > results > as Version 1.1 for each of the various languages. Again, the files come > back > to you for checking and compilation. > > Projects of this scope and complexity are not common, but the proposed set > of tools could certainly handle them. > > ---------- > What about document reviews? > ---------- > > There are two ways to allow reviewers to comment on drafts: allow them to > have access to the text files in the version control system, or give them > PDF versions into which they can insert comments. The latter procedure is > preferable. > > Adobe Reader alone cannot be used to insert comments in PDF documents > generated by LaTeX. Adobe Acrobat can do this, but it costs money, and I > prefer an all-freeware solution. > > Foxit Reader for Windows is free, and can be used to annotate PDF documents > generated by LaTeX. You will need to download the software for your > reviewers from http://www.foxitsoftware.com/pdf/reader/reader4.php. > > Under OS X, the supplied PDF viewer Preview contains annotation tools. > > Comments by reviewers can be incorporated - or not - into text files as > part > of the revision process. > > ---------- > Is there a PowerPoint replacement? > ---------- > > Yes: there are LaTeX document classes that generate PDF files whose text > can > be revealed line-by-line. > > ---------- > What about those embarrassing hidden-text problems in Word? > ---------- > > Lines and blocks of text can be hidden in LaTeX in two ways: put a "%" sign > at the beginning of every line that you want left out of the generated PDF > document, or just move the text after the "\end{document}" tag. Nothing > after that tag appears in the PDF document. > > ---------- > Conclusions > ---------- > > LaTeX is a stable program, proven in use by many writers in scientific and > technical fields. It is easy to learn, and the core of the program has not > changed for many years. In conjunction with other freeware tools such as > version control systems and PDF document annotation software, it can be > used > confidently in documentation projects of considerable complexity. > > So - > > Why not make our lives easier? > > ---------- > ---------- > > ************************************************** > To view the austechwriter archives, go to > www.freelists.org/archives/austechwriter > > To unsubscribe, send a message to austechwriter-request@xxxxxxxxxxxxx with > "unsubscribe" in the Subject field (without quotes). > > To manage your subscription (e.g., set and unset DIGEST and VACATION modes) > go to www.freelists.org/list/austechwriter > > To contact the list administrator, send a message to > austechwriter-admins@xxxxxxxxxxxxx > ************************************************** > > ************************************************** > To view the austechwriter archives, go to > www.freelists.org/archives/austechwriter > > To unsubscribe, send a message to austechwriter-request@xxxxxxxxxxxxx with > "unsubscribe" in the Subject field (without quotes). > > To manage your subscription (e.g., set and unset DIGEST and VACATION modes) > go to www.freelists.org/list/austechwriter > > To contact the list administrator, send a message to > austechwriter-admins@xxxxxxxxxxxxx > ************************************************** > -- Bob Trussler Phone 0418 661 462