It Keeps on Ticking!
Wednesday, August 16, 1995 by Dave Winer.
Are the current pack of web browsers the last word? Or, are they even adequate? I don't think so. There's a bright future, and it involves new ways of viewing the written word on the net. As Scoop on KFOG says "If you don't like the news go out and make some of your own." Right on! As Sheryl Crowe says "All I want to do is have fun." Of course.
After their stunning IPO, and the inevitable fall back to at least stratospheric valuation, Netscape has announced a deal with AT&T to sell servers and clients for the worldwide web. Nothing amazing in the announcement, but it was coool to see my old friend Tom Evslin quoted in the press release. Tom is at AT&T now, after a stint at Microsoft. Yo Tom! [What's his email address?]
More on the Netscape IPO -- based on nothing more than gut, I'd short Netscape stock right now. The multiples are way too high. True, now that they're flush with cash, they could buy lots of product, and become the networking giant that investors hope they will become. My question: can Silicon Valley support a new huge competent company? Is the local talent pool deep enough to support the kind of growth that would imply? I'd bet that it isn't. They'll hit a wall, the people wall, real sooon now.
I made a choice back in 1988, to go with Think C and the Macintosh API. It's been a long run, seven years, but I can see things changing for me. Which way to go? Hmmmm. Time to go for a walk!
I'm having trouble converting from my old Quadra 68040 setup to the PowerPC setup. Metrowerks is a big transition. I yearn for the familiar keystrokes and menu commands of Think C 6.0. I'm angry that Symantec dropped the ball and let Metrowerks take the cursor from them.
When I choose a development environment, I want to pick one that's going to move where I want to move. I think it's amazing, in 1995, that I can't flip a switch on my Macintosh development system and have it generate a Windows 95 app. After all this time, why can't a PC emulate a Macintosh? A lot of people had to drop the ball for this to happen!
What about those guys down in Monterey who had the Mac runtime environment working on a PC in 1989? Altura Software. Why didn't Microsoft buy them? Why didn't Apple? Such small-minded strategies. Microsoft wanted me to use their API. Apple didn't care if my software ran on Windows. Symantec was asleep at the wheel. Now, why doesn't Metrowerks make a deal with Altura and make me that much happier about learning their New Way of doing things?
I don't like any of the choices in front of me. The user interface that I'm comfortable with is obsolete. Using Metrowerks feels foreign to me. I feel abandoned as a user. I have exciting things to do -- much more exciting stuff than learning a new user interface for my primary development tool.
We're developing a tradition here. I write about PDAs and my friend Randy Battat, email@example.com, Motorola's exec in charge of wireless data products (i.e. PDAs) writes back. Yesterday I wrote about cutbacks in Motorola's PDA efforts, and Randy replied:
You mentioned the article in the Wall Street Journal about our staff reduction and what it means for PDAs. While saying goodbye to good people is painful, I think the article got caught up in a little too much PDA hype (what else is new?) and missed what's really going on in the wireless industry.
The wireless data industry is where LANs were in the mid 80s. Remember the Year of the LAN? Which one? 1985? 1986? 1987? And then we woke up in the early 90s and discovered that 3COM and Cisco and Novell are all billion dollar companies. People don't take wireless computing for granted today, but they will. All of us in the industry expected the market to be much bigger by now, but so did everyone in the LAN industry in the early days.
Note that the issue is not about PDAs. Yes, we do make a couple of wireless PDAs. We're also just starting to ship a wireless modem on a PCMCIA card for notebook PCs. For the next several years, we'll sell far more PCMCIA cards than PDAs, but that's always been expected.
Frankly, our number one challenge is getting the wireless data communciations industry off the ground. It's not PDAs versus PCs. Our industry is primarily serving vertical markets today. At Motorola, in addition to serving vertical markets, we're also going after general business usage like Wireless E-Mail. Perhaps we're nuts (because nobody else is following suit), but we really believe in this technology and what it can do for people.
The PDAs represent another belief. Major computing architectures don't scale. Mainframes and minis didn't. PCs can't scale to be handheld devices. Yet many people want something small and personal that they can use to help organize their lives and communcate. Not instead of PCs, but in addition to them. In this way, I agree with a point you made a long time ago that we can look at PDAs as accessories to PCs, you called them peripherals.
But PDAs are fundamentally different than PCs. They embody a different usage pattern (work on a piece of a task, such as scheduling a meeting or jotting down a note, rather than a whole task). They use a different user interface. They employ a different software model (high-level scripting languages and small apps vs. the giant monster apps of PCs). And, I believe, they'll adopt a different economic model (software and services via subscription). All of this means that PDAs will take a long time to mature, just as PCs did.
Thanks Randy! I have a few comments, of course...
I agree that wireless technology will be in our lives not too far down the road. I look at my desk, and under my desk -- it's a tangle of wires! I hate them. If you can help me get rid of some of those wires, and thereby increase my mobility, bless you. I'll get my VISA card out. And I bet a lot of other people will too.
Bless PDAs too! Let's keep the space open for something wonderful to happen there. In the meantime let's stay focused on the exciting things that are really happening, like all the wireless PCMCIA cards you're selling and more fun with notebook computers.
I don't buy the premise that we need new development tools to build small form-factor software. History doesn't support the assumption that a new scale begets fundamentally new approaches to software development. After 20 years, I'm still coding in C! Yeah, and I expect to be doing that forever. Carpenters still use hammers. Some things are too painful to change. I really believe that.
Of course I write lots of scripts now too. What goes around comes around. I did that on Unix in the 1970s! Gotta love it.
New friendly interfaces for the net built out of scripting software makes sense, but not just on PDAs, on desktop systems too. That's why I thought it was brilliant for Microsoft to choose Visual Basic as their development environment for their aborted attempt to produce a PDA operating system. A million developers. What a coup! Microsoft Network as the runtime environment. Too bad it didn't happen.
Also, the software distribution system for desktop software has already changed. Netscape is founded on free net-based software distribution. Creative distribution is the norm nowadays; it's the present, not the future.
If you're betting your company's PDA future on the assumption that desktop OSs can't scale down, could you support it with some specifics? Is ROM too expensive? Couldn't we IFDEF-out the incompatible pieces of the desktop OS? Couldn't I store files in a ramdisk on the PDA? Lots of software has TCP verbs built in, including scripting software. Meet them closer to where they live, and the Internet crowd will supply the scripts.
Glenn Davis, the Cool Site of the Day man, http://www.infi.net/cool.html, takes a contrarian position on Ceneca's PageMill product, announced last week at MacWorld Expo in Boston:
I saw PageMill at the Expo and at first was incredibly impressed. I remained impressed until I got to ask a couple questions of the authors. "Can it do tables?" As I expected, the anwser was no. I can understand that. The table code in HTML can be very confusing and it would be difficult to do WYSIWYG tables. However, in elaborating, the person demoing it from Ceneca also said that it doesn't do left/right justified floating images either. Now that, in my mind, is possibly the most used and useful Netscape enhancement and a very serious oversight on their part. He stated that it's tough to code. Word processors have been doing it for years. What's the scoop? In my opinion, without that feature PageMill is useless. Sure it makes creating pages incredibly easy, but that one feature is an incredible oversight. I was all set to order a copy, but not now. I'll wait for the PageMill update that handles all of the image features.
John Cameron Swayze, the man who sold Timex watches on TV in the 1950s, 60s and 70s, died yesterday at age 89.
Herds of elephants stomped on the watches. Whales swallowed them. These were Mission Impossible watches.
Swayze said "It takes a licking, but keeps on ticking."
How did they do it???
Along with You Can't Hurry Love, it's one of my main mottoes for software development. Software that keeps on ticking is the only kind of software I like to make. I'd quote Swayze to the developers that worked for me. When they showed me a feature that got rid of black holes in databases, or supported a new feature in the OS, I'd hum.
Takes a licking. Keeps on ticking...
Thank you John Cameron Swayze!
PS: PDA stands for Personal Digital Assistant. Make a fist. Raise your middle finger. It's digital! (Is it personal?)