Monday, March 25, 1996 by Dave Winer.
As you know, I took a few days off from writing about my online travels. I even took a trip into the real world -- I went skiing! I found my center of gravity. My Texas-sized body hums and floats and twists and hops. I have balance and poise on the mountain, where gravity is something you can play with. Walking on ski-less feet can be real trouble, when you're accustomed to flying, mere walking is, er, well, hmmm, different. Yah.
I took a break from writing DaveNets not because I ran out of ideas, more because I had become overloaded with them. I wrote a few pieces last week, but they were all confused. In the end, I didn't want to ship any of them.
The geek in me was buzzed. I think I became distracted from what really matters. There's an ecology in the software business. If Apple leaves a vaccuum, someone probably will fill it. So as I zooom back into the net world, I find that I have some updated observations. Chair lift conversations brought out the truth again. As usual, I want to share my conclusions, and then I'll move onto something else.
The Mac is a minor client platform, current numbers are steady at about twenty percent of the browser hits at www.hotwired.com. Twenty percent is nice market share, but it isn't enough to drive standards. I don't think this is going to change because Netscape drives this market, not Apple, nor does Apple have a chance at taking the lead here, because content developers aren't developing unique things for Mac users.
The Mac will lose ground as a client platform unless they match the new integration between Microsoft's browser and the Windows OS. Microsoft has taken the lead here, it's Apple's turn to follow.
The Mac is a minor server platform, even though the numbers may be impressive. Server sysops prefer to serve web pages from Unix boxes or NT machines. I try to find a competent ISP who will host my pages on a Mac, to run my CGI scripts, and the answer is always no. So unless I want to buy a huge pipe, I have to be happy serving stuff on a Unix box.
Apple could let go, and make an investment in smoothing out the FTP-connection between the Mac OS and other systems. There are lots of Unix people at Apple these days. Working with other platforms is good stuff. File sharing over FTP? Yes, we gotta have it. All our scripts would work!
If Apple wants to be taken seriously in the server world, new system software is needed. The Mac file system is creaky old legacy code. We need a new low-tech file system that works with WebStar. Performance is everything. In the meantime, the Mac makes a great CGI platform.
Another thing Apple must do if it's serious about the server business -- start from scratch and design and deploy a server farm of Macintoshes. Make the economics work. The next big rage in the web world is going to be sites built out of hundreds of NT machines. If Apple wants to play in this world, we're going to have to have a way of adding Macs to racks ad infinitum. Maybe Compaq wants to do this? DEC? Hmmm.
Think content. Think content. Think content. All indication from Apple is that they think Adobe has the problem licked with SiteMill and PageMill. I don't think this is true. Adobe's products are opportunistic. They recognize that people want easy HTML tools. But people also want to do frames and tables. Sure you could do a great editor for frames and tables. But people want to call CGI scripts and embed ShockWave stuff and Java stuff. It'll never end. Until Adobe catches up we'll be using BBEdit, if only to add these features to PageMill files. And after all that trouble, why bother? -- use BBEdit. Save trouble, save time. For now, BBEdit continues to rule this market.
Otherwise, people who write and draw still like to use Macs in numbers disproportionate to the total market share of Macs. I've said this many times before, no need to repeat it.
There's danger in trying to wake Apple up. When they spot an opportunity they tend to fire up an internal development team. But they can't give away their software for less than I can. I think my tools solve problems that need solving. I've invested lots of time in understanding how this stuff works. And it's beginning to get easy, as the art of Mac net scripting gets better, and as the problems scripting can solve best become clearer.
In a perfect world, I should fear nothing. If I want to bring new power to Mac users, the platform vendor should be excited and help and stay out of the way. I think the chances are virtually nil that it will play out that way! It's a fun ride anyway.
Apple's foray into the web browser world is more indication that things haven't changed. Why invest in a new web browser? Didn't they see what happened at America OnLine? Cyberdog is the Apple's next eWorld, I'm sure of it.
The world will value simple highly leveraged improvements from Apple for the Mac platform. There are no home runs left for them to hit. Make the trains run on time and all will be right. We need a solid fast file system. Integrate a file system browser with Netscape and Microsoft Internet Explorer. No more rock and roll on TCP transports. Do a big Mac server farm with huge pipes right on the net backbone. Share it with Mac developers.
In the media campaign surrounding the Microsoft's developer conference, Bill Gates was showing up everywhere. I must have read a dozen Gates interviews last week. A common repeated theme -- there is no specific Internet market. If he's serious, he's right. He's serious.
With Gatesian irony, he asks us to think about the RAM market or the hard disk market. The net is just another computing resource. Correct. All evidence is that they're supporting this well in their system software.
Another observation from the interviews -- Microsoft isn't trying to get in front of Netscape, they are trying to *be* Netscape. An incredible example of humility. To win everything, Gates has to acknowledge that the market leader is a startup that he ignored for too long. He said that Netscape has 110 percent market share. If I were Jim Clark, I'd bronze that quote and put it over the fireplace. It's worth $500 million. Maybe more! It's better than an Oscar.
Gates advises his customers to switch over to TCP as their networking protocol. Good advice, and a good competitive move. TCP is what Novell should have been. Why use a different addressing scheme for LANs? Getting rid of this bump will be good for flow.
In 1994 I said that Bill Gates was betting his company on MSN. He said it wasn't true. Now, I think even he would say I was right.
But the battles between Microsoft and Netscape are distractions.
Even the censorship debate is a distraction.
The thing that really matters is moving lots of people onto the web.
I actually like looking at people's pages with pictures of their boyfriends and wives and kids and cats and bosses and boats and motorcycles and houses. I'm really tired of corporate presences on the web. They're all so silly! Someday everyone you know will have a website. It'll change the way people stay in touch with each other. And hopefully we'll make new friends this way too.
It's great that AT&T and MCI are competing to get Internet access into more homes. The next step: make a website an automatic free feature of a phone account. If your number is 510-555-0552 then your web address would be http://www.phoneco.com/5105550552/. Use FTP to move stuff in. Your password is your voicemail password. It's an easy feature to add.
Here's another idea -- take a roll of film in for developing, and have scanned images stored in GIF files on the net. Nothing to upload, you don't need to buy a digital camera. Point to the pictures from your web page. Won't this be loony! Everybody's going to be doing this. Sooon, I hope.
PS: In Gotta Be Aretha! on 1/8/96 I dreamt I was the new CEO of Apple. There were some good ideas in that piece!
PPS: A term I dislike -- Third Party Developer. Too arrogant. Who's first? Who's second? Another term -- Independent Software Vendor. Independent of who? Instead, call us developers. It's humble and respectful.
PPPS: There was a largely unreported system software disaster in the Macintosh world this year -- Open Transport -- which was the only way onto the net for high end Macs, and it didn't work. They shipped it anyway. Amazing. Sounds like something Ashton-Tate did in the late 80s.
PPPPS: Heard in a PBS debate on net censorship, Esther Dyson asked "If you don't trust your kids when they're 14, what about when they're 18?" Yay! Esther says she was a kid once. Gotta love it. She stopped the censors cold with that line. It's kid logic. Simple and truthful and potent. Right on Esther.
PPPPPS: An interesting quote attributed to Marc Andreessen: "We're going to build voice telephony into our Navigator software. We can get it out to 25 or 30 million desktops in the next six months. That's a big enough critical mass for it to take off." Rock & Roll! But what about bandwidth? Can the net support 5 million simultaneous phone calls?
PPPPPPS: BBEdit is a text editor produced by Bare Bones Software at http://www.barebones.com/.