Strange bedfellows
Friday, September 29, 2000 by Dave Winer.
ICE working group 
I went to a meeting of the ICE working group on Tuesday. I'll have a full report on Scripting News probably sometime next week. It was a good meeting. A room full of people who understand content management and syndication. After going in different directions in 1998, we now seem to be headed in the same direction. I would have met with them earlier if the RSS community hadn't been locked up, now I regret not meeting with them earlier.
I wrote about two models for syndication a year ago, one model based on distributing content, the other based on distributing links. I bet on the latter model, it follows the grain of the Web, and imho the other model follows the form dictated by print. I also wanted to support what I call "amateur journalism," more on that later.
I think there will be good forward motion with the ICE group, we agree on the need for simplicity. A good group of people, powerful companies, professional and respectful and open to new ways of thinking. There was even some humor! Right on.
SOAP rules
In the last week I've seen three new companies with interesting products that are basing their future on SOAP. It's really happening. In the hype over P2P, this may be being overlooked.
XML is getting mature, and the killer app for XML, clearly, is SOAP. Some have said that SOAP is just a formalization for CGIs. That's very true. The key word is formalization.
We can build a more useful Internet if we focus on building systems that cross all kinds of boundaries with ease.
Everywhere you see a screen-scraping application, there's a SOAP app waiting to be developed. How to monetize it when there are no ads? Charge for the service.
At the same time, let me express my love for XML-RPC; SOAP's lightweight older brother, a Model-T of distributed computing. Yesterday I heard about a new JavaScript system that does its object serialization with XML-RPC. At a conference earlier this year, the CEO of a $70 million technology company hugged me. Their software system is based on XML-RPC and they like it. David Galbraith, chief architect at Moreover.Com, says it's magic, it's so simple. I've had the same experience, many times.
I can't tell you how satisfying it is, after many years of struggling with Apple over this technology, to see it take root in the market. Whether or not UserLand profits from it is not as important as the compatibility that's happening. Systems that expose interfaces through SOAP or XML-RPC offer more chances for compatibility, and partnerships that actually mean something for users and investors.
Looking back, the turning point for SOAP was IBM's implementation for Java. Lots of people use Java, and apparently there were serious limitations in RMI that SOAP doesn't have.
Looking back even further, credit Microsoft, for having the guts to get behind the idea. To this day most of the gorillas of the software industry only want to work with each other. They're doing it the hard way. When gorillas team up with independent developers, sparks can fly.
People with big company attitudes complain about guys like me, but instead of complaining, Microsoft embraced. As they say in Apache-Land, "It worked!"
Halsey Minor
Yesterday I met Halsey Minor, the founder of CNET, for the first time. It's surprising to me that if each of us sticks to our plans, we're going to be in the same market someday, in a compatible way.
I always think that disconnects are permanent. I have to remember to relax about that. The ICE meeting on Tuesday was more proof. If people keep their minds active and learn from all their experiences, me too, we eventually arrive at the right place, even if there are detours along the way.
Halsey sent me a followup email, "I feel like I just travelled back to 1995 when this 'industry' was more fun!" Right on. I feel the same way. I sent him a pointer to Radio UserLand. We're going to meet again in a few weeks.
BTW, this is one of those times that I can't tell you what's up. It's happening more and more. For example, when we were working with Microsoft on SOAP, I said publicly that we were working with Microsoft, but I didn't say what we were working on.
I have to respect their processes, and I'm also learning to do more of that with my own work. It would be great if everyone could always be open about what they're doing, but there's an imbalance if your competitors don't do it.
Strange bedfellows
Earlier this week I was asked to write a piece for Fortune explaining, from my point of view, what's going on in the technology industry in Y2K. (I spoke the piece into a voice recorder, for a half hour, they're going to edit it for a special feature where a few other people from Silicon Valley do the same thing.)
I explained that the Internet is not what the dot-commers said it was. Strategies that gather people behind specific domains and put ads on their writing is not what the Internet is about. The ads don't generate enough value to justify the market caps for all the companies that IPO'd in the last few years. It had to collapse.
Now where will the technology industry go? It's so obvious, back to technology. Who cares what we have to say? That's not our job. Our job is to supply technology. Did word processor vendors have a stake in what was said with their tools? Our systems have to work, the users come first, and we'll ask them to pay. If they do, we can have a technology industry. If not, we'll have to try something else.
Anyway, it occurred to me that now that the blush is off the rose, now the east coast business pubs are going to tear Silicon Valley a new orifice. We deserve it. However, we will have the last laugh. Highly produced and sanitized content is what they're good at. Delivering the truth is what the Internet is good at. Once we get over our latest attempt to centralize it, we'll be back on course, supporting the route-around that will create a new competitive environment for delivery of news, information, opinion and art.
CollabNet
I have been mystified about the business model for CollabNet, but a few days ago I think I figured it out.
To understand what's going on, look at IBM and SOAP as a case study. With a relatively small investment in engineering resources, at the right moment, IBM was able to lock in SOAP 1.1 as a standard, which gives the distributed computing market a chance to grow right now.
As I understand it, they just had two engineers working skunkwork-style, in suburban NYC, for a few months. Then they backed the work, and did something a bit surprising, they contributed it to the Apache project, as open source.
In that, is the value in CollabNet. They are organizing the work of dozens of open source developers as employees, and then doing contract work for big companies like IBM. This could easily be a profitable business. And it gives the big companies a way to efficiently turn dollars into strategies. No one will begrudge CollabNet their profit.
Dave Winer
PS: ICE is an acronym for Information Content Exchange.