The Next HotMail?
Tuesday, May 12, 1998 by Dave Winer.
Dan Shafer, now the top guy at CNet's builder.com has a fascinating article today, well worth reading.
He begins: "First there was free Web-based email. Several companies came up with the notion of free email supported by advertising, but a few refined the idea by putting email on the Web. Suddenly you could read and respond to your email from anywhere, whether or not you carried a portable computer. Cybercafes, all-night copy shops, friends' homes and offices, trade show floors, and local libraries furnished all the access you needed."
His theme is that email is just the first application to be brought to the web. Over time, more and more of the computer we use will migrate to a server, and as that happens the laptops we carry with us can get thinner and easier.
It's an oft-heard theme, from Larry Ellison and Scott McNealy and other proponents of network computing. Is Java central? Not yet. HTML works great, for some apps at least.
Calendaring, which Shafer singles out as a future opportunity, is a client-server app, and a web-based interface makes a lot of sense. Even if you could craft a better interface in C, who cares? Shafer posits that such a service would be as big as HotMail, and I agree.
On the other hand, web-based apps may not be the only way to go. Especially with more full-time net-connected users, cable modems and DSL, it's becoming more realistic to assume that at least some users have machines that can be "pushed" to, allowing for higher-level apps to manage the display of database-stored local information. It may end up in a web browser display, as an option, but other prettier forms of display such as Director apps, or custom apps (e.g. Doom) also make sense with high-bandwidth full-time net connections.
These push apps are sure to flow thru HTTP because of firewalls, with the client machine running a web server cast in a different role from the page servers that are so popular today. In my opinion, XML-formatted procedure calls are definitely the way to go for the next generation of non-browser-based distributed computing.
This relates to the discussion that's been happening on the last couple of Scripting News mail pages, centering around Microsoft's application strategy, how Office came to be, and how it left behind applications from Lotus, WordPerfect and Borland.
It's been mostly historic, there hasn't been much discussion of where the Office apps are going, probably because the opinions come from ex-Microsoft people who in the past worked on the Office apps.
No matter what, it's hard to imagine the Office apps migrating to becoming browser-based, nor does it make sense (to me at least) that it's worth it to convert a spreadsheet or word processor to run in the context of a web browser.
I'm fishing for ideas. Please speak up if you see how it could make sense.
Are Microsoft's productivity apps in a cul-de-sac? Can they go anywhere interesting from here? If there's room for new collaborative versions of the Office apps, at the top-level, will they work differently from their pre-Internet counterparts? Are spreadsheets and word processors enabling surfaces for collaborative work, or is something else needed? No matter what, Microsoft is in a race with everyone else to figure out what's going on here and to then implement it.
So Microsoft's we-need-to-innovate argument has teeth. They totally do. They aren't finished navigating. From this point on, no matter what the government makes an issue of, they pretty much have to figure it out for themselves, there's no visible competition in apps to boost them.
Of course a flotilla of small independent companies will explore all possible niches. Will Microsoft go down a singular path or will they launch their own flotilla? Or are they depending on buying up all the ships, or having the option to? Where are the other piles of money on this? IBM? Oracle? Intel? The cable companies? The TV networks? Everyone else?
With the success of HotMail, venture capitalists surely are funding dozens of startups to try variants of the HotMail approach. If there was a cohesive cross-company strategy behind the VC-backed products, they might not have to worry about Microsoft. Of course cohesive cross-company strategies are practically unknown in this industry, so time may be on Microsoft's side, as many of the VCs are betting it is.
There's big money to be made here either way. Microsoft paid $400 million for HotMail, and today there's a report of AOL purchasing a leading chat software vendor for $300 million. Whether or not such applications make the cash register ring, it's clear that they're worthwhile entrepreneurial investments, making it sure that we'll see a glut of such applications in a few quarters.
One of the Mail Page writers, Kornel Marton, says that Microsoft is on a treadmill, they must keep growing at 30 percent per year to keep their stock option game rolling. If it should fail, Microsoft is history, he says.
I don't agree. Microsoft is a big bank account, the biggest installed base, and the biggest commercial software distribution system. They own stock in Apple, Comcast and many other companies. A rising tide works in their favor as long as they invest, upgrade and distribute wisely. Not all of Microsoft's investments flow thru the campus in Redmond.
If their internal developers become inefficient compared to those outside, it's possible for them to let go of that model and shift to a new one. It'll be a big change, bigger than any proposed by the government, if and when it happens.
It's a line that Apple crossed in the early 90s when outside developers had become more efficient than inner ones, and the internal ones were stalled. Apple ignored the opportunity. Will Microsoft, if it should happen that way for them?
One more question. It all seems to revolve around Microsoft this week. Will it always be that way?