Dan Wood: The Eponymous Weblog

A few people have e-mailed me asking for my reaction to the recent Macworld article that says "Apple Resurrects Sherlock -- and Does It the Right Way." In many ways, it's flattering to Watson, but the author comes to a conclusion that I just don't get.

Matt Deatherage's back-page opinion piece in the May 2003 issue (no link available) starts with a significant amount of praise for Watson, but right away he states:

But if you're expecting a lengthy battle for the Web-services crown, stand down. It was over before it began, and Sherlock 3 won.

---

The funny thing is that this article doesn't really justify this conclusion. The author goes over the history of Sherlock and how Apple let Sherlock 2 languish, then discusses all the cool things that Watson does, then reviews how Apple copied so many of Watson's ideas in Sherlock 3. But then he compares Watson's Cocoa-based architecture with Sherlock's mixed scripting-language architecture and then comes to this conclusion: "Only Cocoa developers can write Watson plug-ins, but anyone with advanced Web-site-building skills can write a Sherlock 3 channel." This, and the supposed plethora of Sherlock 3 channels written by third-party developers are why he comes to the conclusion that Sherlock 3 "won."

I have to say, there are a lot better arguments that one could make about how Sherlock 3 may be the winner of the "battle". For instance, it's very difficult for software that users must pay for to compete with something bundled for free with the operating system. It's also difficult to compete in mindshare -- there are hundreds of thousands of people with Sherlock 3 installed on their desktop, and they may not know any better. But his reasons seem pretty weak, and ill-informed: Creating a Sherlock channel is definitely not a simple matter (See my detailed comparison); one needs to know two different scripting languages (with Apple's proprietary extensions), as well as a pretty good understanding of Cocoa. I have yet to see a Sherlock 3 channel that is as feature-rich as a Watson tool. As far as comparing number of tools provided by third-party developers (How many Sherlock 3 channels are you using?), that's about as useful as comparing the number of software titles available on the PC verses the Mac. Sure there are more programs available on the PC, but are they the ones you'd actually find and use?

Which brings me to the last part of his article, in which he suggests that "Watson's best bet" would be for a future version to be compatible with Sherlock 3 channels. Sure, it's feasible from an engineering standpoint, but what exactly would that accomplish for anybody? And from a marketing standpoint, this is akin to suggesting that since the Wintel platform is the clear winner in the PC race, Apple's best bet at survival would be to embrace the Wintel standard and start selling Wintel PCs.

I just can't come to the same conclusions as Matt. Sure, it's an uphill battle for a company with the superior product to compete against the 800-pound gorilla, but it doesn't mean that it's time to give up.