Documenting a success: David Skok
In this part of the interview, Skok discusses his views on using Java and other technologies -- including open source software -- to get ahead in the marketplace.
Two guys in a garage
Skok: Yes. I've never seen a time where it was more possible to identify great new ideas to use technology for and to get funded for them and to get acceptance by the marketplace for small, young companies. So those are all the ingredients the two guys in the garage need to be successful. Some time back, you had a much harder environment for two guys in a garage to distribute their products worldwide. The Web solves that problem.
You need to have something hot, though, because now there's a lot of noise out there. You've got to be able to rise above that, but if people hear about you, they can find your Web site from anywhere in the world and if the Java community gets hot about something, then you frequently find everybody talks about it on newsgroups. And funding is another really important thing, and that's available.
The other part is all the seasoned execs you need to get into your company. They all want to join start-ups now, because they've all read about their friends making money out of stock options and everybody now knows that they're never going to get rich by living on even a decent salary; they're only going to do it by getting stock options in a hot start-up. It seems that this is the best time you've ever seen.
There are more places where new, exciting applications can be developed than ever before because of the fact of this ability to use the Web to get the application broadly used and not have somebody go to the trouble of installing it and putting a lot of emphasis on having to convince people they have to put it onto their machine to use it. That solves the problem of niche applications. To a desktop user, it's hard to sell them more than a spreadsheet, a word processor and a graphics package, because it's too much effort and energy. But via the Web, it's easy to convince somebody that once a month they should go to some human resources type of site and just do something they need to do.
I think we've got this huge range of new, possible business ideas, and we haven't even remotely started to see them narrowing down yet; we're in the phase where more is possible. So if you're in your garage, there's strong possibilities out there.
Fisco: What are your views on open source software?
Skok: I think it's a good idea. There's a variety of technologies where there's a broad interest in seeing that technology become completely standardized and used across the marketplace. So if that's the situation, then there's perhaps a rate of development that needs to take place that's quicker than any one small organization can pull off. Putting it out into an open source environment gives you that. Probably the best example that I can see right now of this working is the Linux operating system.
Apparently, though, the key to that working well is Linus himself. And one of my fears, if I was a corporation investing in Linux, is: What would happen if Linus stopped working, or decided Hawaii would be a better place for him? Is this feasible to continue? Is this totally dependent on this one guy? Maybe Linus is the illustration of the best way that open source can work because he's clearly shown that they're able to take that operating system and stably move from one release to the next. But I keep asking myself: If he's able to do this, why can't we [as a company] do software development that way? Break the rule which says that you can only do development in small teams, otherwise it gets out of control. So I think there's got to be a lesson there for us all to pick up and learn about development, which we haven't figured out yet.
Fisco: There's a lesson, but there's a contradiction, because that particular type of structure, based on the Roman Catholic Church, slaps in the face everything that we've learned about management in the last 15 or 20 years, with Total Quality Management and pyramid flattening. Taking the Linus type of a structure into a corporation, I think, would be a complete failure.
Skok: Disaster. Well, we certainly don't follow it with the way we do development. How is he able to pull it off? Deliver stable releases? Is there any other situation that you can think of where somebody has pulled off open source development with the same commercial-grade quality? We're talking about a server operating system. High up-time, more reliable than NT.
Fisco: Netscape is trying, obviously. I guess you could say some of this spirit is alive at Microsoft, or at least was. But Linus has celebrity status. There's a feeling in that community that this is a very special person. I think he commands by virtue of that reputation.
Skok: In other words, it's very non-replicable and it's very, very worrying that we've got this single source of failure, because when people buy an operating system, they want to feel that there's another 15 or 20 years of life in the product, not dependent on one, single person.
That's where the area of questioning for me comes in, around how you manage that. How do you take all these pieces of code coming in, stably test them, get out releases, and ship a product? Certainly, we haven't been able to figure out how to pull that off internally. We have to have very, very strict source code management and any kind of parallel branching that takes place, well, it can be a nightmare for us. It's not a linear system.
David Fisco is a Web software developer and consultant based in Budd Lake, N.J. He is the former editor of DOC/Java and editor-in-chief of The Java Report, as well as a special consultant on new media to several clients in the entertainment industry.
Page 2 of 2