Sysops

Thursday, November 16, 1995 by Dave Winer.

Goooood morning!

Last week I wrote about how invisible I'd become. What does that lead to? What comes after invisibility? I got lots of responses. Still not sure. Checking it out. But one thing it definitely can lead to is a bad cold! I must have caught it in Boston. I was totally grounded last weekend.

At 40, I'm surprised to find that a cold has a bigger effect than it used to. Or maybe I'm paying more attention. Both things are true. I prefer feeling healthy! That's true too...

So I called up my inner mom and held the hand of my inner child and proved that it's going to be OK. Lots of Vitamin C and soup. Rented movies. Saw Pulp Fiction again. Coooool. Let's just quiet down and get healthy.

Uh huh.

Boy did I crank out code! Lots of shit happened. Check out http://www.hotwired.com/userland/yabbadabba/newfeatures_383.html.

Yeah. It's interesting that my body slows down but my head and my spirit don't! I'm totally excited by the stuff we're doing. And it's moving really fast.

Here's a flash I had the other day -- I think I've got the most powerful computer on the planet. If they could only watch me use my tools. I can make big things happen. And it's just me! Yes it is. I swear.

My local network is more powerful. I have three computers with their own IP addresses, and room for nine more. The information superhighway runs thru my house! I can pick it up, look at it, show it to my friends. "Look -- this is the Internet!" Hmmm. I always thought the Internet was the ultimate in vapor. I'm fascinated when I see something you can touch and see. A wire. This is the Internet. Coooool. This is a server. It looks so small. I grew up in the days when big computer presences were actually physically big. Not no more!

Looking at this coax cable running across my workbench, it's pleasant to think of all those little ones and zeros floating and humming along the wire. Put together they form a personna called Dave. Everything you know about me comes over those little coax cables. And they're just wires! Yea.

The net is so damned fast. Now I'm debugging at the TCP level. Maybe it's so amazingly fast because for the first time I have a real sense of how it works. Are phones fast? Yes, of course -- but I don't think of them that way because I have no idea how they actually work. But a Ping call across the Internet happens in a fraction of a second. I could be talking to a computer in Sausilito or Florida or Virginia or Austin. It happens in an instant. There's so much talking going on. And it's so simple. It's very low-tech stuff, but it works! Keep OpenDoc and OLE -- I'll go for a no-brainer any day. Thank you.

On the other hand, the net is so damned fragile. I'm trying, finally, to totally automate my production system -- so I point at a folder on my desktop and run a script from the Finder's menu bar. I have totally smart scripts. But the whole system gets dumb when the result is FTPd across the net. It's fragile. It might not work. So the last step of the process remains manual. My flow of ones and zeros needs me to hold their hands while they travel from my desktop to a place where you can quickly download them.

Someday, hopefully sooon -- I'll just be able to point at something I want to share with everyone and the computer and the net can do the rest. It's almost there for me.

Pragmatically -- we need a friendly Internet Service Provider that's willing to work with us to debug the connection between desktop software and their server software. Everything has to be right -- high bandwidth connection, lots of managed disk space, and good people to keep it running and to work with us to get the glitches out.

There's a position for quality in ISPs. Everyone's thinking coolness is the commodity, but they're wrong. A great on-the-backbone service-oriented service provider that emphasized throughput and reliability would clean up. Let guys like me create the glitz and glamor -- I want my server guys to be really square and professional and calm. And to get the job done.

I've written about this before. The Compaq of service providers. Please!

If you build it they will come.

Sysops 

Who are they?

Sysops. System Operators. People who manage online communities. People who develop websites. The talent is developing. A lot of very young people. They need tools and they want a bigger ballpark.

We're talking about this on the claybasket-talk mailing list. We call it clay.webfarm.com. People develop and manage websites using my software. I come with a community of sysops. They need a stadium to play in. A show floor for their booths. Of course I have competitors, and their users will need this service too. We'll be compatible. FTP is a great platform. In other words, if a Compaq-style service provider provides the stadium, I want to supply some players. That's my dream.

We can get it to work if we want it to.

The big financial entities in this business are missing a big opportunity. Apple has lots of cash. They're doing a dividend. Why not use cash to open up the most brilliant service provider? It's the kind of thing cash does best. Adobe could be a big winner here. AOL could attract a developer community if they wanted to. And Oracle, looking for a niche to own -- why not buy a bunch of SGIs or Suns or Nexts and set up the most massive network operations center in the world? And make it really work.

By the way, to make it really work, you have to answer the phone. Don't miss that. When we call and offer to work with you -- you have to say yes. If you don't say yes to all reasonable offers -- you aren't the winner.

BeBox and Next 

When I wrote about Jean-Louis Gassee and his multi-CPU BeBox and its multi-tasking OS I got back a lot of mail comparing it to Next.

I agree! I see Be as very comparable to Next and, guess what, that's why I think it will kick butt.

Huh?

First, I think the Next platform worked. Too bad the platform vendor wasn't tuned in enough to realize it. The web was invented on the Next platform. The web is hot. Next still has a valued and respected position as a web server. Lots of Next boxes out their running websites -- check out http://www.webcrawler.com. And a bunch of sysops and webmasters who like their OS and their tools.

Anyway, I had lunch with Gassee yesterday, and I was very emphatic that he should configure his system as a web server and sell it to sysops and service providers. A new OS is an advantage here. No legacy code! It should be absolutely totally fast. Multiple PowerPC processors. New disk operating system. No installed base to support.

If Be went for it, and really got with the state of the art in web serving they could clean up.

Jean-Louis -- please think seriously about doing this.

Thanks!

The next OpenDoc? 

Is the loop closing on OpenDoc yet? I hear they shipped.

Now the question is who wants it?

When these proposed standards are launched, usually years before they're available, no one can say what they'll be good for. Eyes glaze over, everyone is impressed with the power. The nausea factor. A lot of money is spent, years pass, and eventually people start forgetting to ask about it. In the meantime lots of energy is spent worrying about what it means to this that or the other thing.

Negative energy -- the proposed standard isn't a growth thing, it's a limiting thing. The Fear part of FUD. Dave Nagel is wrong -- there's no such thing as positive FUD. The irony escapes me. FUD is a negative thing, almost by definition.

So, Netscape is still where the action is. Are they the next Finder? Could be. In the meantime, HTML looks like the platform. It's still a crock, and it gets crockier. But it's the best crock we got.

It's great to have a standard to feed on. It's how you have fun -- wrestling with a crock that has become a standard. Solving a piece of the puzzle and then executing.

As Eric Schmidt of Sun said -- everyone he knows is working on HTML now. Me too.

So why bother with OpenDoc?

And this raises the next question -- is Java the next OpenDoc?

I'd like to run a short list of things Java is good for. Let's get this out in the open, early in the process. What are the top-10 compelling applications for Java? Why should we devote our energy to understanding it and developing tools for it? These are serious questions. Like a lot of developers I'm thinking about investing in Java.

To Java's proponents -- what applications do you envision?

Steve and Apple 

Cathy Cloud, mcleanpr@applelink.apple.com, called last week to make sure I had seen a piece in the SF Chronicle by David Einstein, einstein@sfgate.com, about Steve Jobs possibly returning to take charge at a new Apple hypothetically owned by his friend Larry Ellison.

Cathy wondered what I thought of this.

My answer: I don't like the idea.

I can see why it appeals to Steve. He's always wanted the Mac to be the information appliance, the computer for the rest of us, a limited closed machine with very little power. A toaster of a PC. That's exactly what Larry Ellison is talking about. There's no doubt that this idea came from Steve. What goes around comes around. Steve still wants to hit the home run, to top the Apple II, or what he thought the Apple II was.

Would Steve coming back make Apple stock more valuble? Probably. Maybe by a lot. There would be a Steve euphoria. He's an electrifying speaker. On stage, he can make you believe. I'm sure he can still do it. If he were chairman of an $11 billion company, people would pay attention. Steve would know what to do with the attention.

Steve Jobs is growing older, so maybe he's wiser too. But I've always found that there isn't a lot of room around Steve. So if it's either-or -- if everyone will look to Steve for brilliant product ideas, I think that's bad for the Macintosh platform, because they probably won't look elsewhere, which is where the powerful ideas are being implemented.

Steve doesn't actually write code. He doesn't speak geek.

The ideas the Mac platform needs to win are already out there. The Macintosh net community is thriving. Steve doesn't understand the net community on his own much smaller platform, and he's had lots of time to get it. So I think Steve would be bad news for the Mac platform. It would distract our attention, again, from what's really going on and where the opportunities for growth are.

To make its stock more valuble, Apple should tap into the growth of the net. Maybe Apple needs a visible symbolic break with its past, but bringing back Steve Jobs is the wrong way to do that.

If Apple continues to dance the Microsoft dance they'll continue to have an undervalued stock, and Macintosh users will continue to feel like they're on shaky ground.

Whoever runs Apple, they should pay attention to the needs of content developers, the tools they use, spotting opportunities for improvement in those tools, and recognizing the breakthroughs when they happen. It's a simple quiet job -- make the trains run on time. Keep the flower bed fertile and wet. Make machines, distribute them. And keep the noise level down so developers can make their own noise.

And please Apple, take better care of the operating system, which is your exclusive domain and protected from competition. Make the file system faster. Add a Windows menu to the Finder. Simple little things that show you're tuned into the needs of your users and developers.

Dave Winer

PS: A great website, showing how software docs on the web can really work -- http://web66.coled.umn.edu.

PPS: I think sysops are a lot like gardeners.

PPPS: Once again, FUD stands for Fear, Uncertainty and Doubt.