Bruce Horn’s career started at the Parc where, as a student, he helped in the development of Smalltalk-based various projects.
In 1981 he started working for Apple where, in the next three years, became a key member of the Mac team.
His contributions include the design and implementation of the Resource Manager, the Dialog Manager and the Finder and also of system-level parts of the Macintosh such as the type framework for documents, applications, and clipboard data.

In the late 80s and early 90s he kept on working with Apple, as a consultant, and was with the Advanced Technology Group (ATG) but also worked for Adobe, on a variety of small projects, including a LaserWriter spooler. In the last years he has been working on Natural Language Processing Technologies and is now at Microsoft.

I contacted him as part of my research on the first year of the Mac and he promptly and very generously answered to all of my questions. In the following interview he shared his thoughts on the “computer for the rest of us” and his role at Apple, before and after.

Bruce Horn with Steve WozniakStories of Apple: What are your fondest memories of the work you did on the Macintosh?

Bruce Horn: Working with Andy, Steve Capps, Steve Jobs, Larry and Patti King, Bill Atkinson and the rest of the Mac software team, and Chris Espinosa, Caroline Rose, and the rest of the incredible Publications group. It would take a whole page to list all the amazing people with whom I had the pleasure of working on the Mac.

SoA: Any funny or weird story you were part of or you witnessed?

BH: Many–see Andy’s wonderful book!

SoA: Do you feel the original Mac’s legacy is still present in the current Mac lineup?

BH: Absolutely. The original Mac’s look and feel is largely still there, enhanced dramatically. I believe that the true spirit of the Mac has been maintained and evolved through the years. You can still tell that a Mac is a Mac and not a PC with its unique combination of ease, simplicity, power, and enjoyment of use.

SoA: Nowadays we have GUIs, object-oriented programming, networking, laser printers: has the personal computing world finally caught up with PARC?

BH: Yes, in some ways, but not in others. For the most part the Xerox PARC vision as embodied in the original Smalltalk systems has been copied and enhanced, but the essential dynamic nature of Smalltalk–the ability to change the entire Smalltalk system from within itself–is not something that you can easily do in current systems. Of course, the ubiquitous networking, wireless communications, extremely cheap color printing, digital cameras, and so on have surpassed the PARC technologies substantially.

SoA: On folklore.org and in his “Revolution In The valley” book Andy Hertzfeld states that the Finder and its spatial approach wasn’t influenced just by PARC but also by the Architecture Machine group at MIT. Is this true?

BH: From my point of view, my main influences were those from PARC, and specifically from the Learning Research Group–the group that developed Smalltalk, the first overlapping windows, popup menus, modeless editing, and so on. I don’t think that I had much knowledge of the Architecture Machine Group.

SoA: Some years after creating the Finder you worked again at Apple, twice: first on the Newton and later in Apple’s Advanced Technology Group. What stage was the Newton in at the time? Was it still in the form of a big tablet-like device?

BH: The Newton was a small-ish tablet device and was in the very beginnings of its evolution. I was involved in helping to determine what kind of dynamic programming language would be best for the systems programming of the Newton, and produced a report that had recommendations. At the time I thought that one of the best options would have been a streamlined Smalltalk system, since Apple had developed, in cooperation with Xerox, a very nice Smalltalk implementation that we could then leverage for the Newton.
Unfortunately due to political issues this never happened. An alternative programming system based on LISP was developed but was too large and inefficient for the needs of the Newton, so Walter Smith took on the challenge and invented NewtonScript. NewtonScript turned out to be a very interesting language and an excellent fit for the Newton.

SoA: How was working at Apple’s Advanced Technology Group as opposed to a “product” such as the Mac? How loose and “pure” was the research conducted there? Was Larry Tesler still in charge? And what do you think about Steve Jobs shutting the ATG down when he returned in 1997?

BH: Working in ATG was a lot of fun; we were able to explore a lot of interesting things that would eventually become part of the Mac OS, but much later. One project I worked on with Tom Bonura and Jim Miller, called LiveDoc, eventually led to features in Apple Mail where certain types of text are recognized (e.g. names, addresses, phone numbers) and become “live”: you can click on an address and get a map, click on a phone number and create a new contact, and so on.

LiveDoc in action

Our implementation was even more comprehensive and integrated into the main text editing component, so it was available everywhere. It is unfortunate that Steve shut down ATG when he came back, but you can’t be critical of the results. Steve saved Apple and we are all happier for it!

SoA: What’s your opinion of the iPhone and its UI?

BH: Just tremendous. I love my iPhone! I’m sure that if I thought about it critically I could think of some improvements, but it is such a joy to use I’ve never spent much time evaluating it critically :-)

SoA: What are you working on now?

BH: Currently I am the Principal Development Manager for Natural Language Processing Technologies at Powerset, which was acquired by Microsoft last year as part of the Live Search group. Our group is responsible for developing fundamental NL technologies for understanding language, such as finite state morphological analysis (part of speech, stemming, etc.); parsing of text; semantic analysis; and a whole range of development and debugging tools to make the process easier.

Some of the real pioneers and leaders in the NLP community are at Powerset, including Ron Kaplan who codeveloped Lexical Functional Grammars, the basis for our parsing technology; Tracy King, who is our lead syntactician and former head of the NL group at PARC; Dick Crouch, our lead semanticist and developer of our semantics framework; Livia Polanyi, one of our lead researchers and developer of a number of groundbreaking concepts in NLP; Martin Van den Berg, one of our semanticists and key architect of our original semantic index; and many others. This group consists of quite a few PARC alumni, so I feel right at home.
While we are part of Live Search, and thus responsible for improving search at Microsoft, our technologies have a wide range of applications, and we hope to leverage them in projects throughout Microsoft.

SoA: You put a lot of work into iFile: what is its current status? Are you still planning to release it? Did you try Mac OS X Finder replacements such as Path Finder?

BH: It’s on the back burner. Working at Powerset is tremendously demanding and I don’t have much time to work on other projects. When I first showed iFile to Steve Jobs in 1999, it was way ahead of its time, but now with specialized apps like iPhoto and iTunes, and enhancements to the Finder, like Smart Folders (which I showed Steve as part of my iFile demo), many of the functions of iFile are now available, although not in an integrated fashion.
I don’t think I’ll release iFile as it is. Maybe if I get some free time I’ll figure out what the next great thing is. It might be iFile v2 or it might be something completely different based on largescale distributed computing. Who knows :-)

Pictures are courtesy of Bruce Horn. The Livedoc example is taken from www.miramontes.com


Update: Per Helge Berrefjord has snapped a series of pictures of Horn at the end of the Eighties when he was at the University of Oslo