This article has a nice tone, but I think the author makes a lot of generalizations by simply applying his current job/situation to the world at large. Specifically, I disagree with the following:

1) The world is going paperless. Alas, this is not happening. While it's true that some areas, like technical writing and review, have made progress towards this goal, printer supplies remain a truly lucrative market.

2) People prefer software with fewer features. Yes, it's true that people prefer a simple program that does simple things, organizations have laundry lists of things they want and tend to buy the thing that does the most for the cheapest, even if there is overlap (insert long rant about organizational purchasing here)

3) The page metaphor is dead -- it's all web now. This would be great if it were true, but I seriously doubt it. Lots of people (like me) prefer a hard copy of work when editing. In addition, pages give reference points during conversation, stuff like "check out the first sentence in the third paragraph on page 3" which gets very unwieldy without pages

4) Superior collaboration software translates into superior document creation software. Wiki is great. Love it. But in lots of places, like law offices, when you create a document, it has a very specific legal status. It must be kept for a period of time. It must not be edited without a clear trail of edits. Etc. For folks in the healthcare, legal, and insurance world (among others) documents are simply electronic versions of printed pages that have a lot of rules associated with them. This isn't a societal convention as much as it is a legal one. Good luck on getting that changed anytime soon.

Word is a feature-heavy monster, no doubt about it. Nobody uses 99% of all that stuff in there. But there's a long way from making that observation to officiating at its funeral.

The rumors of Word's demise have been greatly exaggerated.

In the law office environment, Word fulfills at least the following functions:

1) It avoids incompatibility issues in document formatting (i.e., you know that, if you created it in Word, others in your own office and elsewhere will be able to open, use, modify, and save it without any problems - a critical item as documents are exchanged among different parties in a variety of ways in such an environment).

2) It fits in with document-management systems that allow you to create and store your documents, in all their different versions, on local servers that you control and with which you can preserve the confidentiality of client documents.

3) It provides a baseline for determining that new employees will be efficient in their word processing skills (i.e., they must have a minimum set of skills that includes being able to work efficiently in Word).

4) It does all the things the you point out in your items 1-4 (particularly your observations about being bound to the idea of the "page").

Any substitute for Word would need to leave me capable of interacting with clients, staff, fellow workers, opposing attorneys, courts, government agencies, etc. in this same seamless manner without incompatibilities while using the "page" model through which we all are accustomed to relating to one another.

Call it the pull of the legacy, but it is an overwhelming pull. Everything else is bleeding edge from the standpoint of having to run a real business in the current environment.

This does not mean it can never change, but it will not change based merely on an idealized set of ideas of how it can be done better or on something that might work in one environment but does not port well into the standard office.

"In addition, pages give reference points during conversation, stuff like "check out the first sentence in the third paragraph on page 3" which gets very unwieldy without pages.

Isn't that why we have URLs?

As to legal requirements, perhaps Word is insufficient for a system with such strict legal requirements? Maybe a wiki-ish system that strictly enforced all the necessary requirements without requiring humans intervention? But I guess you are saying that the way the laws are written actually require actual pieces of paper to exist. In which case software doesn't help.

I'm curious, though, to know if that is the case or if businesses have sprung up specifically to meet the legal requirements of electronic document management for those vertical industries.

I think this is spot-on.

From a student's perspective, taking notes in class in Word is the de facto standard at this point. The outlining mode is intuitive and efficient.

Now let's say we were to replace this with MediaWiki. When I tab and shift-tab it won't indent and un-indent properly. While I'm sure there is Wiki markup to do outlining, there is no way you'd be able to do it fast enough in class while you're trying to get down everything the professor says.

The author, as parent points out, makes the common mistake of assuming what is true for one in a group is true for the entire group.

I still use word for some things, but taking quick notes is not one of them. Pure text is the best way to take quick notes, because you don't have to worry about formatting (and using the mouse to do formatting). If you learn how wikimedia works, it is probably easier than writing in word...

"Pages are dead"... but his article is divided into two pages?

I get the irony, but I suppose here it is so that the ads all get a second hit.

I completely agree with the premise of this article, but this line baffles me: ... Word's document format will always be complicated, even if it is natively XML and completely open, simply because it has to support all the thousands of features that Word itself provides.

This is akin to saying that binary has to be complicated because of all the numbers it can represent. Most features don't require a file format change in any way, even when storing, say, undo histories for new features. Word documents really aren't that complicated, unless they're made to be.

You must do some research on how complex OOXML is. http://en.wikipedia.org/wiki/Office_Open_XML .Initially there was a huge opposition in making OOXML a open standard, but eventually Microsoft succeeded in pushing the standard. Ever wondered why a empty word document's size is ~34 KB?

I'm not arguing that it isn't complex, I'm arguing that the feature-set of Word has nothing to do with the complexity of the file format by necessity. His assertion that the feature-set of Word is what makes it complex is simply foolish.

It's true that any arbitrary file format's complexity doesn't necessarily have to be tied to the feature set of the authoring tool (e.g. LaTeX), but in this case much of the extraneous complexity of OOXML is tied to Word's feature set. In particular, the need for backward compatibility with features from previous versions of Word make this format more complex than it needs to be, given the current feature set.

There's a Joel on Software article that discusses this. See http://www.joelonsoftware.com/items/2008/02/19.html

Historically Office file formats have grown from in-memory structures serialized to disk which helps in part to explain the complexity of them.

The other way of phrasing that line is: to garentee that you can display any Word document, you need to be able to support any formatting that Word supports.

Parsing a Word document is much simpler than displaying a Word document, often people mean the latter when they say "open a Word document".

I'm happy with Latex & Google docs.

Anyway, well done MS Office.

I agree, the main reason (besides the fact that LaTeX is the greatest for any type of document, and that greatness is extended by XeTeX,) is because I still believe one must have good file structure. As much as CMSes and Wikis are the rage today in web trendiness, you still need to manually organize them in some way, and this doesn't solve the problem that the article hints at about sticking them on a network share. If it's gonna be lost in a network share it's gonna be lost in a wiki or a CMS. The difference is you might not be running an indexing service for your docs on the network share but searching is intrinsic to the wiki.

And also, git for tracking changes, version control isn't simple, I find that MediaWiki's version control is more annoying to use than git.

Outlines in Word is excellent for ensuring that a document has excellent structure.

I use latex a lot (more than Word) yet people do not see its limitations. It would be best if Latex is replaced with something like open source Adobe Framemaker or something.

LaTeX has limitations?

You misunderstood file structure as document structure I meant it as in logical organization on a drive.

To split a document up into multiple files. But the disadvantage is that Latex only has single includes - so you can not split a document up that much.

(This is also a problem if you use images with equations such as those created by xfig (i.e. pure latex or PSTex PicTex). Usually most people would like to include an image as a separate file.

org-mode in Emacs is nice for outline-oriented note-taking, and it can export to LaTeX.

Gah that thing almost drove me over to Emacs. By almost I mean, I couldn't get full vim functionality in Viper and gave up, some commands just can't be replicated. But org-mode is a very very nice tool. The current vim alternatives like viki and vim-outline do not compare.

Office 2007 just got installed on my box at work, and the first thing I saw was a blue haze of controls and tabs that just confused me to no end.

When the feature set is as mature as what Office has, replacing the UI makes for more "look how old my current version is!" fodder for upgrades, but doesn't do much for making people on the fence want to stay with the product.

Then again, I am curious to know what the penetration levels are for various wiki products vs Word in the business market. I just can't imagine large non-computer companies embracing wikis terribly quickly. Word documents as data transfer still provide a simple and easy-to-understand model for how to write and share, even if it's a pain-in-the-ass in actual use.

If wikis want to replace word effectively, they'll need something like an Adobe AIR client. Web browsers are nice for lots of things, but I've nearly pulled my disks out and punted them across the room when an extra tab/backspace causes the browser to go back a page.

As I say above, a "real" word processor is much more than a text editor with WYSIWYG formatting. Anyone who works on reasonably complex documents would expect their software to keep track of cross references to tables, figures, sections, pages, etc. This is a feature than simply doesn't matter when writing a letter or a CV which is probably what most people actually use a word processor for. But Word 2007 has a whole tab devoted to just this, and another whole tab devoted to change tracking. The people who need it really need it, and ultimately they're the ones that decide what word processor the organization uses. You might get them to switch to OpenOffice, but Google Docs is long way from ready for prime time.

But Word's pretty terrible at that stuff, too. I've nearly thermite'd my machine when word screws up my styles or starts dancing with my inline images. I've stuck to svn/latex/emacs for real docs. Word tries, but it's just infuriating to make it put out a well-formatted, consistent document when you hit over ~150 pages.

If you can do it, you've got sk1llz I don't.

Yeah, that's bloody annoying. But my point is: people who believe that Google Docs is a viable substitute for Word (and OpenOffice, etc) are probably not trying to write particularly complex documents, because if they were they'd have run into missing features.

Though Google Apps has very much less features compared to Office, a small business could well operate perfectly just with Google Apps. I'm not sure what percentage of Microsoft's revenues are generated by small businesses but Microsoft is sure facing a tough competition this time. Any thoughts?

I'm sure a lot of buinesses (small or large) find the idea of hosting their documents (often containing business secrets or of critical importance) on the servers of another entity very, very frightening.

Zoho's online tools are, imho, better than Google's and their Remote API solves a lot of the storage problem(http://writer.zoho.com/public/help/zohoapi/fullpage#RemoteAP...) . I suppose there's still a transactional security issue but with more and more people working at home I think that's an issue no matter what.

I think this is a critical flaw, in this day of HIPAA and SarbOx. Any business that allows someone else to store their data (or even process it) risks personal responsibility for their officers -- and what corporate CTO is going to risk that?

I don't buy this. I'm not an expert on S.O. but I'm close to one in regards to HIPAA. As far as HIPAA is concerned as long as you have a Business Associate Agreement that clearly spells out the privacy level expected from the other party you aren't going to get in trouble for a HIPAA violation.

A Business Associate Agreement nothing to address the SOX requirements, which apply to more companies. "Google ate our documents" is not a defense to SOX filing and reporting requirements.

And yet are happy to type them on an OS that is connected to the internet but that they don't have the source code to

Not just operate well enough, but operate better. All their work in one place, up on the web. Easy team-working. Easy sharing with the customer.

Docs still has ties to paper, though. The Google product that fits this model the best is Wave.

Just had a look at Google Docs and it's not immediately clear to me how it manages references (e.g. "see figure x", "section s", "page n") where it keeps track of all the numbering for you as you insert new figures etc. Word has done this forever. It's a pretty key feature for anyone working on non-trivial documents.

Surely the modern equivalent is the hyperlink?

"Click here" all over the place is very much not the same as "see figure X".

Hyperlinking "click here" is terrible form. Just work the link into a sentence by putting it over a descriptive phrase like "our revenue last quarter", or even (more explicitly) "the graph of our revenue last quarter".

And that works when you print it how?

(If you're thinking "it doesn't matter" then you've not understood why Google Docs isn't ready for the typical business)

I've seen print transformations of hyperlinks that replace them with numbered footnotes. A sufficiently clever one could use its own pagination and put in "[1] see section X page Y". But honestly that's an obsolete format. If your business can't cope without, use Word. But it ought to be able to. What use is paper nowadays? Who prints stuff out?

What people "ought" to do and what they do do are often very different, and you can either pretend that isn't true or you can try to understand it and act accordingly.

I asked "who uses it", because I can't offhand think of anyone who has to. OK, there are the bosses who can't cope with anything invented since the 1980s and are too high placed to be told to suck it up. There are people in countries too backward to expect a reliable internet connection. And perhaps, some documents required by law? I don't know.

"Who uses it" is the question Google should be asking themselves - so they can turn their considerable market and political power towards obsoleting those uses.

Google Apps as distinct apps are dead men walking.

Their only future is as a collection of gadgets and robots for WAVE. WAVE simultaneously solves the problems of over-bloated office suites and Google Apps problem of not having a locally-hosted option.

In a small company I worked for, we decided to put most of the internal writing into our MediaWiki. Project plans, memos, design whitepapers, meeting minutes, vacation plan, internal encyclopedia, FAQs for locally installed tools etc.

Earlier, the company had installed a TWiki, which quickly filled up with poorly-named pages. With a mandatory naming convention and readily available copy-paste templates (I wish MediaWiki had MoinMoin-like templates), this was no longer a problem.

It worked well for most of those things except for project plans (no way of summing up those table cells). For external documentation however (we offered consulting and engineering for safety-critical software), we used DocBook.

By the features and functionalities that he writes, the software that can kill MS Word seems to be Google Wave.

I had never given this any thought at all but totally agree with the article. My company uses GoogleDocs for everything but rarely prints anything out. I'll have to look into the wiki thing.