[A longer than usual blog post, summarizing some strategy ideas I’ve been working on.]

In the early 90s at DEC, I had a colleague that worked with the cable industry.  I tried (unsuccessfully) to convince him that TCP/IP would be the winning network infrastructure for the fancy interactive TV services everyone was talking about.  In the end, “adequate general purpose” solutions always win.  Proprietary solutions can’t compete with economies of scale.

Moving up-stack, I’ve been thinking about how TV content delivery is going to play out.  We started with over-the-air broadcast, then shifted to cable (first analog, then digital), and now we’ve suffered through a decade or two of crappy set-top-box user interfaces fed from proprietary cable networks.  Along the way, our TVs have evolved from tiny, fuzzy CRTs to large, crisp 1920 x 1080 color monitors.

As household bandwidths have increased, we’ve got options that end-run cable companies to get content on the screen:  Apple TV, Roku/Hulu, Tivo, network-enabled DVD players, game consoles, etc..  But these are still closed & proprietary:   I can’t deploy my own apps (or even content, in some cases) without a lot of permission from other people.

On the bleeding edge, users are plugging their large-screen TVs into computers (the Mac Mini is a popular option), sometimes installing software like Boxee to provide a “couch-friendly” UI.  I like Boxee, but it’s only an intermediate step and feels like a response to the past.  Users don’t want “set top boxes” (STB) or “media centers” or “electronic program guides”; they want unrestricted access to apps and content.

And a general-purpose delivery mechanism already exists:  Web browsers.  As Apple demonstrated with the iPhone, HTML+JavaScript is an excellent way to get content on-screen, because it leverages a deep existing technology infrastructure.  (And adding Flash makes it even more compelling).

I’m betting the same will happen with TV:  HTML (and related technologies) will become the new “broadcast standard”.  I’m not talking about seeing Firefox’s File/Edit/View menu bar on your TV; I’m saying that content will be rendered by a full-screen browser engine using HTML+JavaScript+Java+Canvas+Flash+Quicktime+etc technologies.   Users will access any app or content they want by navigating to the appropriate URL.  (Innovators will adapt browser navigation to the TV screen & remote control, just like Apple adapted Safari for the iPhone.)

Here’s the test:  how hard is it to write Boxee’s UI or any STB UI as a full-screen JavaScript or Flash app?  Apart from accessing LAN content, it’s relatively easy.

I’m pretty certain this is how things are going to play out.  Now the entrepreneurial question is:  what to do about it?