Author's note: This interview with Dan Wood, chief architect of Watson, is the first in a new collection of articles we're branding as "Developing for Mac OS X." In this ongoing series we'll cover the ins and outs of how to successfully bring your ideas to the marketplace, from both a technical and business perspective.
It's appropriate to launch this endeavor by publishing a conversation with Dan Wood, who has achieved great respect among Mac users and developers alike for bringing Watson to our computers, then continuing to innovate as competition increased.
Derrick Story: Hello Dan. Thanks for taking time to visit with me and our audience here at the Mac DevCenter. I've noticed from the Karelia site that Watson 1.5.5 is available for download, and it has some impressive new features. But before we get to Watson itself, I'd like a little more information about Karelia and your role in the company.
Let's start with Karelia. The home page says that it's a region in Finland and Russia, and that the company name is inspired by the Karelia Suite, a classical composition by Jean Sibelius. Could you tell us the story about how this became the name of a Mac software company?
Dan Wood: Well, we hired a very expensive naming consultant, who convinced us that the new trend in software company naming is the classical music motif. Seriously, it was a domain name I had picked out for my own personal use several years ago. When I was looking for a company name, I decided to turn that into a company name, rather than having to hunt for a reasonable domain name in today's crowded namespace.
DS: And what's your role in the company these days?
DW: I'm still the chief architect of Watson, and I still do programming on it, although I've brought in another talented Cocoa programmer to work on our latest plug-ins. I hire others who do the other pieces like icon design and marketing on our Web site. I'm designing a new product that I hope to have ready early next year. Lately I've been spending a bit of time talking with other companies about partnerships that will help us move Watson to new levels. Oh, and I also find time to write some technical articles for MacTech magazine.
DS: So tell us a little bit about the inspiration for Watson, and how you continue to muster the energy to go toe to toe with Sherlock?
DW: Late last year I was looking for ideas for a program to build in Cocoa. I was inspired by one of the keynotes at Apple's developer conference that talked about "Deep Internet Integration." I had run across some small utilities that did web "scraping" (extracting and reformatting useful data from HTML pages) such as monitoring eBay auctions and collecting headlines of SlashDot.org. I decided that the world needed a handy container application for these kinds of utilities, and from there the individual tool ideas started coming left and right. Somewhere along the lines I recognized that this kind of program was to services what Sherlock was to searching. So I came up with the name Watson to complement Sherlock and to help people understand what the program was all about.
DS: Elementary...and brilliant!
DW: As far as having to compete with Sherlock 3, well, originally it looked like I was just going to have to put Watson into maintenance mode and accelerate the next project. But a few weeks before Jaguar shipped, when Apple developers were able to first experience Sherlock 3 for themselves, we started getting countless emails from users asking us not to give up on Watson because it's so much faster than Sherlock and has so much more functionality. So I decided to breathe some new life into Watson. I figured that since Sherlock had moved into the services game, not just in general, but pretty much all of the specific services that Watson has provided, that it made sense to add general Internet search capabilities to Watson as well. Hence the new Google searching tool. So it's really our user community that has given me the energy!
DS: What are the features about Watson that excite you most?
DW: Well, I've been a fan of good user interfaces, which is why I've stuck with the Mac through the dark times. And I love what the Web has done to affect people's lives, but I hate its non-standardized, minimalist user interface. In a sense, the Web is a great big "dumb terminal" with fonts and graphics. Using the Web, I yearn for the subtle user interface behaviors and consistent controls you get in a real desktop application,like buttons enabling and disabling when it's appropriate to use them, scrollable lists, dynamically adjusting displays, things like that. To me, Watson is a way of taking back the great user experience that the Mac perfected. I still get a kick out of zipping through a site using Watson's columns browser that would otherwise require page after page after page on my browser.
DS: Well, not to pile on here, but I've heard lots of developers complain about the Web experience, but I have to say, the Watson interface really leaves the browser in the dust for certain online activities. In large part, I think many people will agree with me that you've set a new standard.
DS: That being said, I'd like to shift gears here for a second. We all know about the stir caused when Sherlock 3 was released right after Apple awarded Watson the "Most Innovative Mac OS X Product" for 2002. Many people felt, myself included, that Apple should have handled the situation better. But they didn't, and Sherlock 3 sure looks like it borrowed much from Watson without any formal acknowledgment.
DS: My question to you is, based on your experiences, what suggestions do you have for Apple in terms of developer relations?
DW: Good question. I think Apple has a long way to go in terms of developer relations. Apple has helped some companies build themselves up, but others they have harmed greatly. From stories I've heard from other developers, the Sherlock/Watson issue is just the tip of the iceberg. I respect Steve Jobs's desire to make the Mac the most useful tool it can possibly be, to win over converts from the Windows side, but instead, Apple seems to be nibbling at their own developer community. Perhaps there needs to be a better agreement that both parties enter to avoid these situations. Apple should be thinking win-win in terms of developer relations--helping companies build great software so that both Apple and those companies will succeed--but the reality seems to be "We'll provide you some help to make a product, but we might just bundle your ideas into our OS or our next iApp."
|
DS: Obviously Apple needs independent developers to be productive and creative in order to move the platform forward. My personal experience at the ADC level (Apple Developer Connection), is that those folks have much respect for independents. How do you think that respect could be better translated in to support?
DW: Well, it's obvious to us, but I wonder if it's really obvious to the leadership at Apple. I've had wonderful experiences with some individual people in ADC, but their leadership seems disingenuous, pushing Apple's anti-developer interests rather than really trying to help out developers. Apple could do so very much to build up a healthy economy of Mac developers and cross-promote their products, but they only seem to push the products from the big companies that they need in order for the Mac to survive, like Microsoft and Adobe. Just imagine if Apple were to take the one million dollars that they spend on a fancy staircase in their Manhattan store, and instead invested that in their developer community to help some small companies with fresh ideas get off the ground? Or heavily promoted third-party applications instead of writing replacements for them?
DS: Speaking of developers and writing software, let's talk for a moment about what's happening around Watson. As I understand it, Watson has an open architecture. Is that correct, and if so, could you say a few words about it?
DW: Right, we publish an API at http://www.karelia.com/developer/watson/ that describes the protocol that an individual Watson tool must follow, and the big toolkit that comes with Watson that makes it easy for the tools to load and parse web pages or XML streams. We provide header files, class documentation, and sample Cocoa code. We also have a developer community email list with over 150 members, so people can exchange ideas and help.
DS: Does this mean that there's opportunity for developers to write plug-ins for Watson?
DW: Absolutely! Many have, in fact. Some people are writing in-house tools that aren't for the general public, but we have a handful of smart programmers that have made available some additions that I couldn't have conceived of. One guy, Sujal Shaw, is a big sports fan, and he's contributed an amazing tool to show baseball scores and stats. And now he's working on a similar Football tool that's almost ready for release. Terrence Talbot wrote a tool for recipes that shames our built-in Recipes tool! There are a few others that are hovering near release, including a great genealogy lookup tool.
DW: I think it's really important to have an application be pluggable. For those who are curious, we actually have a comparison online--slightly biased, of course--between Watson's architecture and Sherlock 3's, which was documented online for less than a week. Being able to easily create a front end to an Internet service is a lot of work, and Watson's architecture should make is as painless as possible, because our toolkit has been evolving for quite a while.
|
Related Reading 
|
DS: Since you've mentioned comparisons, how is Watson faring now that Sherlock 3 is out and in use?
DW: Sales dropped when Apple first showed off Sherlock 3 in May, and dropped again significantly around the MacWorld Expo New York time frame, when Steve Jobs showed off Jaguar. However, sales are strong again since our new release now that we've taken Sherlock 3 head-on. The jury is still out for the long run, but I think we're back to being a success again. Heck, maybe Sherlock 3 will inspire more people than ever to buy Watson as more users get a taste of the concept but wish for a faster, more powerful alternative.
DS: Do you think Watson is going to help change our idea of the Web? We're still very browser-centric in our thinking. Watson represents a shift away from the browser. More like a Web services thing. What's your take on that?
DW: I get a lot of comments from users that Watson opened their eyes that the Internet doesn't have to be just a Web browser. It's very easy for people to get used to one paradigm and get stuck in it. A year ago, there was nothing like Watson to quickly access the most useful services. And now we're starting to see a few other applications that break the boundaries of the Web browser and use the Internet in completely different ways. Take Spring from UserCreations. It puts a canvas of interconnected icons representing real-world objects like people, books, places, and foods on your desktop, and each object is connected to the net. NetNewsWire from Ranchero gathers up news feeds and presents them in a simple UI. WeatherPop from Glucose puts a quick weather display on your menu bar. I don't think that the Web as viewed through your browser will go away anytime soon, but people are starting to realize that it may not be the best way to view structured information.
DS: Is there anything else you'd like to cover today?
DW: Well, since the readers of this will mostly be developers, I suppose I have something to say to that community: go for it! It's actually possible to get a new product built and selling with a minimal outlay of time and money. Of course, you need a good idea, and you have to work hard, but it is possible to make a living and build a successful software company, even in today's economy, even deploying on a "minority" operating system like the Mac, even Mac OS X only. The new generation of innovative applications like the ones I just mentioned don't come from the Adobes and the Microsofts, they come from the hearts and souls of the little companies with fewer than 5 or 10 people. Apple has provided us a powerful system for building great applications. It's called Cocoa, and the rest is up to us. Just be prepared that a big company with intimidating lawyers might come along and build their own "cheap knock-off" of your idea, and then stay ahead of the game!
DS: Dan, thanks for your time. I'm thrilled with the success you and your team has had to this point, and I wish you all the best for the future.
Derrick Story is the digital media evangelist for O'Reilly. His current book is The Digital Photography Companion. You can follow him on Twitter or visit www.thedigitalstory.com.
Return to the Mac DevCenter.
Copyright © 2007 O'Reilly Media, Inc.