Microsoft is bipedal; its legs are Windows and Office. I’ve always thought that Office was the more important, and less open to attack. But there insurgents lurking out there in deep space; their attacks are just pinpricks. So far.

Windows, at the end of the day, is not a very good operating system. Some of the pieces of Office, though, are superb, and would probably have been very successful over the years even without the file-format lock-in and abusive competitive practices. In particular, I’d put Excel on my personal list of the five best pieces of software ever. I want to bust the file-format logjam not because I think it’ll sweep Office away, but competition will make Office cheaper and all the competitors better.

I hadn’t really paid much attention to the Google Docs thing; but then, a couple of times in recent weeks, I was in gmail and there was a message with a .doc attached. Normally I’d save it and open it with MacOffice; I have an elderly but still satisfactory version, but that’s all screwed up because my new Mac came with a trial version of the latest release, which I accidentally opened once and it’s bollixed up my fonts somehow.

Then I saw “Open as as a Google Document” and I tried and, you know, it’s Good Enough. And snappy. And free. And internationalized. And I’m not the only person having this experience.

Yow... thought-provoking.

Now I see that Apple is taking a run at the fortress with Numbers. Dunno yet whether it’s going to be a triumph like Keynote or a stinking piece of crap like iCal, but the fact that they’re trying is significant.

I’ll be honest, I’d thought that MS-Office on the desktop was going to be a central feature of the ecosystem for as far forward as I could see. But maybe not.


Comment feed for ongoing:Comments feed

From: Kevin Marks (Aug 08 2007, at 21:11)

Tim, this is something we've discussed for a while- the possibility of HTML+CSS as a document format:

The thing is, that is just what the Google Docs team have implemented:

For an encore they made it auto-save to redundant web-based storage and support simultaneous multi-user editing too.


From: Mike Kozlowski (Aug 08 2007, at 21:24)

I suspect that Office at home has little future; people just don't need an office suite that much, and the limited free things can do everything they need and then some.

But at the actual office, there are so many business processes built up around Office (Excel is for office workers what Perl is for sysadmins -- the glue that lets everything work together) that I don't see it going away soon.

Also, I'm genuinely amazed at the quantum leap that Office 2007 represents. After a bunch of versions that -- despite welcome improvements -- all felt a bit samey, it's nice to see a huge innovation in a place that nobody expected to see it.


From: Jacob Kaplan-Moss (Aug 08 2007, at 22:27)

After a couple of nights with Numbers, I think it's an unqualified success. About a million times more intuitive than Excel, and nearly as powerful. It's certainly going to replace Excel just as Keynote replaced Powerpoint.


From: paul (Aug 08 2007, at 22:36)

I am using a beta app called Mailplane to access my Gmail. It's kind of minimal browser, if you like, or a browser that only views one site. Simple and effective, skinnable. More things like that will pry apart Office's grip . . . it will be interesting to see how a increasingly mobile workforce gets fed up with fat file formats and the heavy tools they require, vs lightweight versions of the docs and apps. An example: the one man band that is Hogs Bay Software has rolled out a GTD app, very simple, just checklists that you can move/slice/edit. But the document is just an ASCII list: the presentation is in the app, not unlike modern html/css. So you could create a list anywhere, and it would be comprehensible to anyone, but when open in TaskPaper, it becomes a formatted, active document.

As for Numbers, the descriptions have made me think of Improv, the amazing financial tool that ran on (wait for it) NeXT, back in the day. I haven't seen it in forever, nor have I seen Numbers, but I wonder if Numbers has its roots in Improv's features.


From: Peter Sefton (Aug 08 2007, at 23:25)

Google docs might be good enough for previewing stuff people send you, but it has serious problems importing and exporting 'real' word processing docs, which it will mangle very badly. For example it kills styles.

Tim, I know you think that most business docs need to be made web-ready as a matter of course, but have you tried to use a mainstream word processor to make HTML lately?

I recently did an experiment to see if I could write an academic paper in Word / Writer / Google docs, the terrible results are listed here:

HTML export in those three word processors is stuck in 1994: they make very feeble attempts to convert formatting to HTML and none of them bother to have a default set of styles that converts cleanly to HTML. This is not rocket science, but nobody seems to bother to do a bit of basic work to make their word processing package a first-class web citizen.

We have shown that it can be done on the ICE project at the USQ: - there's a web site made entirely from word processing docs with a web-ready style sheet.

There's a big opportunity here for MS, Sun or Google to take the lead in making decent HTML from their word process, I've been corresponding for years with all three off and on and none are interested.

I'm not sure that MS Office can remain a central tool without tackling the web properly. The blogging tool in Word is a start, but it too makes bad HTML and can't be used for much more than plain paragraphs and headings.


From: Ted (Aug 08 2007, at 23:59)

I’d put Excel on my personal list of the five most hated pieces of software ever - because the world uses Excel to handle and distribute tabular data, but Excel lacks proper relational operators. So when it comes to connecting tables together, everyone lashes up their own ad-hoc solution, which for various reasons makes my life a misery. One of Microsoft's Great Crimes was not to bundle an RDBMS into Office.


From: John Cowan (Aug 09 2007, at 05:06)

People only like Excel because they've never seen Quantrix Modeler (, or its precedessor Lotus Improv, in action. It costs way too much, because they are marketing it to quants (and their employers) instead of geeks, but it is the Real Thing.

The number of bugs in spreadsheets is linearly proportional to the number of cells (excluding blank, pure data, and label cells), and only painful code review by several people will catch errors. Quantrix uses a few exposed formulas instead, et voila.


From: Marco (Aug 09 2007, at 05:54)

I second your opinion on the outstanding importance of Excel for the working environment. In my opinion this doesn't come simply from features, which can be replicated by competitors, but from the "ecosystem" of add-ins which were developed over the years around it.

Working in Finance I've realized that Excel and its add-ins (think about Bloomberg datafeeds, pricing libraries, etc..) are probably the single most important tools available, possibly the only bit of the whole MS infrastructure which cannot be replaced.

Playing around with Apple Numbers was refreshing because for the first time after years I had the feeling that somebody was trying to go beyond the Lotus 1-2-3 model... very basic features are missing though (think about grouping, pivots, freezing, ...) which would probably prevent its usage for "serious number crunchers".


From: david (Aug 09 2007, at 07:44)

For me using Office became an unending experience of dodging annoyances and trying to figure out why it wasn't working today the way it had yesterday. The day came when I realized that I worked on the fringes of the Office ecology, most my work didn't require Office and happily iWork was released shortly afterwards.

These days I open Word less than once a month, PowerPoint never, and Excel quite often. Numbers looks like it is suitable for much of my spreadsheet needs but not all. It is missing important features, freeze frames and split screen for example.

Two things seem clear to me: first, Office's days in the home and home office are numbered. Second, as people discover that alternatives like iWork and Google's apps are perfectly fine MS will lose even more customers.

If Microsoft is a bipedal company it better start looking for a good pair of crutches....


From: len (Aug 09 2007, at 14:26)

It won't be the pricks who take Microsoft down. They've been at it for years. It will be the inability of Microsoft to deliver high quality precision development tools to the middle-tier of enterprise developers.

Having the best geeks despise them is really inconsequential. It is like having the finest jazz musicians say nasty things about rock bands. No one really cares. Really. The audience won't leave a show because of it and the rockers laugh all the way to the bank.

But if the middle tier, the so-called long tail majority begin to revolt because they simply can't get their job done in the eight hour day (yes, Virginia, it is normal to want to go home and NOT code), then MS support will evaporate very quickly.

Today their challenge is not their Office formats. You are dreaming. The penetration is so large there that even if every government in the world decided to ban MS, the bootlegging to keep the very large majority of office documents in play alone would keep them in business. That won't be what does it no matter how desperately Sun wants to dig out and IBM wants to take the government service revenue.

No one cares about monopolies that don't make them personally uncomfortable.

But every day every government is issuing and processing the requests for enterprise software based on inadequately performing web technologies. Sun and IBM are inadequate here too. Why? The web is an inadequate underpowered stateless and slow platform. Everyone is swimming in the same shallow pool in the hot sun.

No what will kill MS is that Visual Studio 200n is buggy in ways that waste hours out of every middle tier eight hour day. The frequency of that is far more important than the amplitude of the manufactured bitter butter battle over office formats. It wastes the time of those who ensure the density of MS penetration: middle tier, average so numerous developers who can show their bosses that VS 200n is crapware.

And that will be deadly if they can afford any other realistically productive and reliable toolkit matched to their average skills. The challenge for the open sourcers is to push development tools to the middle tier that match the skillsets. The rest is noise.


From: Noah (Aug 09 2007, at 14:55)

Ack, I know you use Macs a lot, but I really hope you're talking about Excel for Windows when you mention it among the top five programs ever written.

I don't use [Mac] Excel too often, but when I do, I find it requires much more time than it really should to do what I want it to. While the interface is decently Mac-like, it never works the way it seems it should. Editing graphs is a good example: I usually have to click in several different locations, and enter the properties dialog before realizing the option I'm looking for is somewhere else. A simple drag of the graph usually only moves the actual axes, whereas moving the object is really what most people will want to move most of the time.

Worst of all, the program crashes so frequently, that it would be useless without the auto-save function. This, on several different computers, and with only a few hundred cells and several graphs. No excuse.

I readily acknowledge that it is an incredibly powerful program, but it makes doing the simple things really difficult. I would like to think the Windows version is better implemented, but I have no experience with it myself.


From: Aristotle Pagaltzis (Aug 09 2007, at 23:30)

> The web is an inadequate underpowered stateless and slow platform.

That’s why it withered on the vine.


From: len (Aug 10 2007, at 05:57)

When you only have one vine, you water the grapes before you water the horses. Wine over efficient transport.

The precedent established is that a global information network is useful not that its current implementation is efficient or the last word in technical evolution.


From: Aristotle Pagaltzis (Aug 10 2007, at 19:48)

The web was the only contender in town?


From: Colin (Aug 10 2007, at 21:00)

"One of Microsoft's Great Crimes was not to bundle an RDBMS into Office."

It would have made more sense for most of what is Microsoft Access to be part of Excel instead. The parts of Excel which deal with "lists" are constrained by the facts that Excel doesn't separate data from presentation in any meaningful way and that there is no good way to enforce relational constraints on the data.

The most common solution I have seen to make Excel work with relational data is a lash-up of VBA using DAO/ADO to access an external .mdb database (the Jet DB engine being pretty much ubiquitous) and populate a worksheet with the results


From: Eddie (Aug 10 2007, at 21:53)

Why didn't Apple include round-tripping OpenDocument Spreadsheet (ODS) formatted speadsheets with Numbers? Likewise Pages doesn't allow for import / export of OpenDocument Writer nor does Keynote work with OpenDocument standards.



From: Tim (but not THE Tim) (Aug 13 2007, at 19:03)

Responding to varied points...

"more intuitive than Excel": I think, but could be wrong, that the days of needing software to be "intuitive" are somewhat past. Nearly every office worker knows what a spreadsheet is and how it's supposed to work, and nearly everyone uses the basic functions fairly easily. Advanced functions will probably never be intuitive - but then, they're not used as frequently either.

"linking tables together": a spreadsheet is not a DBMS though many think so. Microsoft, and others, rather than just bundling a real RDBMS into their office suites (which they'd never do because DBMS software can generate serious cash, but I digress), ought to ship the "Express" version of the RDBMS, but graft spreadsheet functions on top of the DBMS - it makes more sense that way.

iCal as a stinking mess: nearly every calendar product is a problem. Calendars today are like e-mail two decades ago. The simplification efforts for calendar interoperability are making progress (RFC 2445 revisions, etc.) but who knows when and if the UIs will improve??


From: Ted (Aug 18 2007, at 12:47)

"Excel doesn't separate data from presentation in any meaningful way"

The enormous value of this kind of separation seems to be appreciated at Apple but not at Microsoft.

In most Mac applications, the default paste operation pastes characters but not their style, whereas in most Windows applications it pastes the characters and their style. And in Outlook there seems to be no alternative to this behaviour, which is intensely annoying when you are trying to assemble a large message out of fragments copied from other documents. I asked a friend who works at Microsoft Research for a workaround, and he said assemble the message in Notepad first !


From: Paul B (Aug 20 2007, at 09:27)

Ted wrote:

> I asked a friend who works at Microsoft

> Research for a workaround, and he said

> assemble the message in Notepad first !

I use a clipboard manager that has a "paste plain text" option. Takes a couple of extra keystrokes, but that quickly becomes muscle memory. And I can copy several items into its storage, then switch over to the destination and pull them off. Only one context change.

Development appears to have stagnated in the last year, but I've found it quite solid and, until some MS API change cripples it, I'll continue to use it.


author · Dad · software · colophon · rights
picture of the day
August 08, 2007
· Business (121 fragments)
· · Microsoft (19 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.