AppleCORE: Where Are They Now?

Where Are They Now?

This is the first installment in an ongoing, in-depth series that will focus on former and current OpenDoc developers and software companies.

When Apple Computer placed OpenDoc (and by extension, the company’s own OpenDoc crown jewel, Cyberdog) into maintenance mode on March 14 of this year, many folks who had put a lot of time and effort into the system software, as well as the products that took advantage of the component technology, were left out in the cold.

Software developers such as Digital Harbor (WAV), Corda (C-Table, C-Graph, C-Text Box), Nisus (NisusWriter), Adrenaline (Numbers & Charts), MetaMind Software (Dock ÔEm), SoftLinc (LEXI) among many others were left holding infant OpenDoc parts and editors — a.k.a. Live Objects — and started building a pretty good chip on their shoulders. And really, who could blame them? Apple is well-known for hawking new technologies as "the next big thing," and then, after sinking untold dollars and man hours into such projects, abruptly canceling their support of or involvement in development of said "big thing." (Witness Taligent, a collaborative effort at creating a new OS, and Dylan, Apple’s stillborn programming language.)

But OpenDoc’s story isn’t over quite yet. Despite being put on the back burner, faithful users and supporters of the technology did not allow developers and their software to simply fade away. Perhaps due to Apple’s decision to include OpenDoc (and Cyberdog) in its white-hot Mac OS 8 system software revision this summer, an ongoing trickle of interest in OpenDoc remains, and a fervent campaign by several developers is afoot to get Mac users thinking seriously about component-based software once more.

One of those die-hard campaigners is Brad Hutchings, a principal partner of Hutchings Software. Brad and his company are perhaps best known for creating Rapid-I Button, an OpenDoc control you embed in any other type of OpenDoc container document, and whose appearance (such as a graphic or a QuickTime movie), label and action you determine. I recently interviewed Brad concerning his past, present and future work with OpenDoc.

Mike Lambert: When and how did you get interested in OpenDoc development?

Brad Hutchings: I went "technology shopping" at MacWorld SF ’95. Steve Roussey (of Kantara Development and PartBank fame) and I talked to Mark Thomas, former OpenDoc and AppleScript Evangelist for about 3 hours at his little OpenDoc station. In February 1995, I took the (Microsoft) OLE 2.0 Software Development Kit (SDK) and the OpenDoc with System Object Model (SOM) SDK and experimented. While OLE was more mature, OpenDoc was just done right. I decided to start a new project around summer.

When August came around, I had finished some contract work and an original product. I borrowed the Cyberdog video from WWDC ’95 from a friend and watched it. I called up Steve Roussey and told him he had to see how incredibly cool it was. I think we watched it, rewound it, froze it, rewatched it for a whole afternoon. So I wrote a note to the Semper-Fi mailing list saying how cool it was. Jim Black, former OpenDoc and Rhapsody Evangelist, saw the note and FedEx’d me development tools, documentation, a cool letter which is still framed in my office, and an OpenDoc T-shirt. I guess I’m easily amused, but that’s when I was hooked.

ML: What kind of OpenDoc projects or products have you worked on?

BH: Using a development version of OpenDoc Development Framework (ODF) in fall 1995, I built a few simple part editors, mostly as proofs of concept with the intent of shopping for an investor in early 1996. The most capable part, Signaling Flags, appeared on the OpenDoc DR4 CD. It did speech, scripting, GX, and every other feature you could imagine.

Rapid-I Button came about because I wanted to demo at WWDC in 1996. Nobody was doing controls, which I thought users would really like. So, I merged ODFEmbed, ODFButton, added Cyberdog functionality, and showed embedded QuickTime movies on buttons controlling an Internet kiosk at WWDC. At MacWorld Boston, I showed the first beta. The first beta test was brutal, with performance being the biggest problem. Rapid-I Button had appearances written in AppleScript, embedded labels, and other really cool, super-customizable features which just weren’t efficient. Starting in September 1996, I rewrote it from scratch and released it on November 20.

Early in 1997, I worked on a few part editors for use with Cyberdog. I’ve got a lot of neat features lying around the office, but they didn’t gel into shipping products earlier this year. March 14 threw a wrench into my plans and necessitated that I spend time "rescuing the business."

I’m currently working on "Carousel", which will be a HyperCard/AMT/Director-like editor. It has a hierarchical arrangement to its "cards" (called"Sels" in Carousel). I hope to have public betas within a month.

ML: What was your reaction to Apple’s placing OpenDoc into maintenance mode?

BH: That was actually a tremendous victory at the time. The NeXT people just wanted to can it completely, and Avie Tevanian said that the e-mail campaign we waged in early March saved it at the last minute.

The challenging part of the process was pinning Apple down on what "maintenance mode" meant. Nobody actually knew. In the minds of Apple management, "OpenDoc" encompassed a lot of things — OpenDoc, ALOE, ODF,Cyberdog to name the big ones. There is still an ongoing process to ensure that committed developers can continue to play with OpenDoc. It’s a very difficult process because Apple changes direction every couple of weeks, key Apple employees become unavailable, and individual developers lose interest. But we’re all (Apple employees and developers) giving this process our best efforts, and I think positive news is on the horizon.

ML: How does Apple’s decision affect your current work?

BH: From a development perspective, I couldn’t ask for anything better. I’ve already licensed ODF from Apple, and that has given me the ability to tune it so writing my part editors is easier, and the results are better. There were changes I made for R4-HS that a team could not do, because those changes required days of control over all parts of the project, and that’s not how team development works. The downside is that small changes can’t be done in parallel, since there’s only one of me. Over the long term, I expect that by improving ODF and keeping it publicly and freely available, a few brave souls might write part editors and help build the market.

The market perspective….ouch. "OpenDoc" as a term is going to have to be ditched, because it symbolizes a cross-platform, industry-scale failure. I’m not looking cross-platform. I think OpenDoc part editors offer unmatchable value to MacOS users. Thus, the "Component X" initiative. I want "Component X software" to supplant "OpenDoc Live Objects" in user/press vocabulary by January 1998.

ML: Where do you go from here?

BH: I am excited about the WWDC Rhapsody story. I’m also quite excited about the BeOS. The most important 90% of OpenDoc functionality could be built with a much simpler API on the BeOS in a couple of months. I’ll even admit that I’ve been experimenting with it. But even the simplest, coolest architecture needs software running in it. OpenDoc on the MacOS is more viable than any hypothetical, new implementation on the BeOS based on that fact alone. I am learning a lot from OpenDoc, and plan to remain involved in the user component crusade for a long time.