We spoke with Tim Bray, a member of the Project Management Committee for the Apache XML Project (xml.apache.org), to get his views on the project launch and its implications for the open source community. Tim co-founded Open Text Corporation in 1989, developed one of the first large-scale Internet search engines in 1994, and co-invented XML in 1998. He is currently at work on an Internet Mapping startup.

developerWorks: Why is the xml.apache.org project such a big deal?

Bray: The builders of the Web have voted with their feet for Apache; just look at the numbers. At the same time, virtually every new piece of Web plumbing being invented by anyone seems to have XML in somewhere or other. It seems that people who are working with Apache are going to be working with XML too. If there is a well-thought-out Apache-maintained set of tools for basic XML input, manipulation, and output, everyone wins.

developerWorks: What impact could this project have on the open source community and does this announcement give credence to the open source effort?

Bray: It's interesting to see commercial software developers, both large and small, donating their work to Apache. I think it's good sound common sense, but it's still interesting. I don't think open source's credibility needs any help.

developerWorks: What sort of enhancements do you think that the open source community will contribute to the project? Is there any suggested agenda for enhancement for them?

Bray: We don't know yet. A lot of people think that the best suite of XML tools available anywhere right now is that produced by James Clark, which is already in the Open Source framework. It's being used right now in the Mozilla project and in Perl. Since Perl and Apache are already more or less joined at the hip, we're going to have to figure out whether and how we wire in James' work.

developerWorks: Will this project have an impact on the W3C's standards?

Bray: First of all, I have to point out that the W3C produces "Recommendations" not "Standards." The difference is important to some people. Yes, this will have an impact. Standards are useless, counterproductive in fact, unless they're actually implemented and adopted. Having software to support these Recommendations reliably available to all of the millions of Apache user is a major step forward.

developerWorks: What does this new set of tools offer developers? What are the tools' strengths and weaknesses compared with other tools available?

Bray: The IBM and Sun tools, which have already been contributed, are as good as anything available commercially, and competitive with James Clark's work, which is saying a lot. I think I'm a fairly typical developer, and in the Web Server context, the first four things I care about are performance, reliability, performance, and reliability. So I think development should focus on making these things solid and fast, not on adding features.

developerWorks: What is the relation between Apache and Textuality?

Bray: None at all. I have been asked to join the project based on the fact that I'm a co-inventor of XML, am well-acquainted with work at the W3C, and have a good understanding of Apache (I'm basing my new startup venture partly on generating XML very efficiently in an Apache module).