I SAY TOMATO, YOU SAY RTOS
Have You No Standards?

by Denise Caruso

c. 1990 Media Letter

Tracking the interactive media industry is a practical application of the chaos 
theory -- trying to find meaningful patterns in seemingly incomprehensible, 
hopelessly complex, chaotic data.

If that seems like a pessimistic and slightly depressed point of view, consider 
that for the past couple of weeks my head has been spinning, trying to figure 
out why so many intelligent people who seem to know the score also seem to 
be hell-bent in about a gazillion different directions.

What prompted my latest foray into the media tornado was my musings on 
computer-based video editing. Digital F/X finally announced its long-
anticipated Video F/X product to cheers; MacroMind is scrambling to ship its 
MediaMaker product, which sprang from the loins of Max Whitby and The 
Multimedia Company/BBC; the PC-VCR, a NEC product driven by Light 
Source software, is garnering attention; and Farallon is readying its "Kubrick" 
editing product, an extension of its "multimedia for the rest of us" approach.

Although I believe these products certainly have market potential across the 
standard range of applications -- from creating in-house corporate videos to 
allowing you to wreak havoc on loved ones by editing your very own home 
videos -- the fact that a digitally-driven industry seems to be embracing the 
concept of analog video (edited via computer or no) is something which gives 
me pause.

You may be surprised to learn that there is a wide range of interactive media 
producers who have, at least for the next couple of years, washed their hands 
of digital interactivity. The (comparatively) ancient technology of the 
videodisc player is where it's at for them. As far as digital is concerned, they 
say, the emperor has no clothes. These producers can create a good, solid 
interactive videodisc -- using an easy authoring tool like HyperCard -- in 
about a week. It looks great, doesn't lose one smidge of resolution because of 
compression and requires no heavy equipment investment. The disc holds a 
bunch of full-motion video, and if you can work some magic along the lines 
of Warner New Media's Magilla project (Media Letter Issue 2, July 1990), you 
can even lay lots of streaming tracks of audio and text on top of it. Just about 
the only thing bad they have to say about the videodisc is that it usually 
requires both a computer screen and a monitor for playback.

But they say that interactive digital video technology -- from CD-I to CDTV to 
PC-based CD-ROM and DVI applications -- is a royal pain in the hind-end. It's 
splintered, factionalized, and chasing its own tail. The hardware that's 
required to produce it is expensive. The playback platforms are not 
compatible. The authoring tools are obtuse (despite the hype). The amount of 
time it takes to take a product from conception to mastering on a single 
platform, using the hardware and software that's available today, requires 
gut-wrenching resources amongst a group of people who have a very low 
threshold of pain, financially speaking.

So who are all these analysts talking to, some of whom claim that the 
consumer market alone for multimedia will exceed $6 billion in 1994? I think 
they must be talking to each other, or maybe to wistful vendors, but I can't 
imagine that they're talking to producers or to the people who are going to 
buy the products. Here's why.

STANDARDS, STANDARDS, STANDARDS

Everybody says they've got standards, but you have to pin them down about 
what they mean. Do they mean the same compact disc that works in a CD-I 
machine will play in an Apple CD-ROM drive? An IBM CD-ROM drive? On 
an FM-Towns machine? On Commodore's CDTV system? Nooooo.

Are the products upwardly compatible with HDTV? Nooooo. Do all of them 
use the same (or even a similar) operating system, to ease portability between 
platforms. Nooooo. Do they use standard architectures to store video, 
compressed NTSC signals, CD sound, compressed CD sound, text, 2D or 3D 
still images? Is there a single way to do data transfer? A single user interface? 
Even just two or three? Nooooo.

How that saying go? "The nice thing about standards is that there are so many 
to choose from."

In an article for Design Management Institute, Tom Peters wrote that "Design 
is one of the keys to business success in the 1990s . . . (it) is only secondarily 
about (creating) pretty lumpy objects and primarily about a whole approach to 
doing business, serving customers and providing value." Putting this into 
practice, from the very first line of code, is critical to the success of interactive 
media.

For now, I'm going to ignore the controversy around PC-delivered television 
vs. TV-based computing, another cornerstone of chaos, and concentrate on 
what the market should be trying to deliver, both to its developers and to 
consumers.

In the world of standards, doing business" has to do with the relationships 
hardware vendors have with their developers. Especially in multimedia, the 
ability to grease the skids for your developers and create an atmosphere of 
mutual trust is critical to getting support for your platform.

If you're wondering, then, why more developers aren't foaming at the mouth 
to create interactive media products for the myriad of platforms out there, 
think about operating systems and authoring systems. One report I read about 
the last CD-ROM conference in San Francisco contained one sentence on the 
subject, and I paraphrase: "Sony didn't show any new products, but Toshi Doi 
(Sony's director and senior general manager) talked about the need for a 
standard operating system." If you look at history, it seems to me there's 
something very foolish about not listening to Dr. Doi.

Right now, if a developer wants to run products on, say, three platforms, all 
the picture, text, graphics, sound and control information is done in different 
sequences with different machine codes. That means three complete 
development cycles for the same product -- and one developer I know 
estimated the cost of a CD-I title at $300,000, including hardware, software and 
the time it takes to author the thing and interleave all the pieces together (see 
"Word of Mouth," pg. 8 for update).

Here's a partial listing of what they've got to choose from. For the proprietary 
Macintosh operating system, there's HyperCard, MacroMind Director, 
Authorware and Farallon's MediaTracks. For the PS/2 line of 80286 and 
compatibles, there's the "winhelp" engine, based on Microsoft Windows 3.0, 
Owl's Guide, Asymetrix's Toolbook and MediaBase from Attica Cybernetics. 
commodore's Amiga has its own OS. Sun has UNIX.

The "consumer" platforms also have their own operating systems. CD-I has 
CD-RTOS, also known as OS9. Commodore claims its CDTV player conforms 
to the ISO 9660 standard, but it doesn't play in all CD-ROM players, does not 
run CD-I discs and requires developers to learn its proprietary OS. Intel's DV-I 
has its own set of authoring tools as well. Not one of them is regarded as easy 
to use or learn.

"Unless and until there are user-friendly operating systems that don't require 
a degree in software engineering from Carnegie-Mellon, there will not be 
enough titles to make anyone reach into their pocket," says my producer 
friend, an interactive video veteran who is fed up with trying to develop 
products for both CD-I and CDTV. "And that's a fact."

The last part of the Peters equation is serving customers and providing value. 
One way to provide value is to make sure that the end product of every 
interactive media developer can work across platforms. As a corporate user of 
multimedia, I want to know that the document I create -- an amalgam of text, 
video and sound bytes, whatever I choose -- can go with me from department 
to department, and from job to job, if need be, regardless of the type of 
computer system I'm presented with. This freedom requires a standard 
multimedia document architecture. There are at least a dozen at the moment.

The same is true for schools and entertainment. Who on earth would invest 
in a library of expensive multimedia products (and $50 is expensive) with no 
conviction that they'll be able to use them past the life of whatever little black 
box they happen to own at the moment? How can you expect people who 
can't program their VCRs to learn a new user interface every time they open 
the box on a new product? If that's what you expect, your business is doomed. 
The diversity of interfaces has already scared a lot of people from anything 
that even hints "computer." If you intend to make me do that, I will no only 
consider that you are not trying to serve me or add value to my life. I will 
consider that you are doing me a disservice and making my life more 
difficult. And I will not buy your product.

Of course, I've been around the block enough to know that there has always 
been incredible resistance to implementing these kinds of standards in the 
electronics business. I remember when Sun Microsystems first started 
bludgeoning the industry with its "open systems" marketing campaign. The 
big problem was that other workstation manufacturers sure as hell didn't 
want to shift their platforms over to plain ol' vanilla UNIX. That meant -- 
horrors! -- they'd have to compete in an open marketplace based solely on the 
price-performance curve, and not hold the trump card of incompatibility.

And in fact, users of technology owe Sun a debt of gratitude for that. Having 
called the question, they've made a lot of customers ask just exactly what 
vendors are willing to do for them, a noble and necessary question.

COGNITIVE DISSONANCE RULES THE DAY

All of this stuff serves as examples of what psychologists call "cognitive 
dissonance." That's the practice of altering your perceptions to suit your 
decisions, a perfect description of the multimedia industry today. If you give 
me choice between an apple and an orange, and I choose the apple, I will find 
lots of things wrong with the orange and suddenly discover wondrous, 
magical things about my apple.

It's not that the industry sees everything as "either-or." It agrees that any 
successful consumer platform has to cost less than $1,000. It seems to agree 
that compact discs, in their many "standard" permutations, will be the 
cornerstone of the interactive media market, whether for entertainment or 
for delivering new business solutions.

And it also agrees that it's confronted with a chicken-and-egg situation today: 
developers don't want to invest the money in creating interactive products 
for a flash in the pan, consumers don't want to buy products without 
reasonable assurance that their investments are safe, and platform vendors 
don't want to make any real big moves toward mass-producing interactive 
players and peripherals unless they're sure developers and customers will 
follow.

But where cognitive dissonance reigns is that individual companies, having 
invested money and energy into proprietary solutions to the above-
mentioned problems, see their personal solution as magical and everyone 
else's as slightly askew. So unless it changes its perception and tries to be 
more holistic about standards for platforms and interfaces and data types and 
operating systems, this industry will be chasing its tail while progress on all 
fronts is significantly stymied. If they aren't willing to take the broader view, 
vendors are indeed asking the impossible of their developers and customers.