Brian Behlendorf, Founding Member of the ASF and Apache HTTPD, Talks How Open Source Developers Can Save the World

Brian Behlendorf is one of the original developers of the Apache HTTPD Web Server, and a founding member of the Apache Software Foundation.

Behlendorf discussed with The Bitsource about the early days of Apache, the future of the Web, and how open source developers can save the world in his upcoming Keynote Speech at ApacheCon 2009.

Brian Behlendorf: I have been on the ‘net since I started at UC Berkeley as an undergrad in 1991.  While working at a new magazine called Wired in 1993, I helped them launch the first major publisher’s website, then later one of the first ad-supported sites, called HotWired.  I left that in 1995 to co-found a company named Organic which built commercial websites, then launched another company in 1999 named CollabNet, which creates open source development tools and communities.  I spent 2007 and 2008 doing quite a bit of public speaking and travel to promote Open Source software collaboration practices across the tech sector, and then since February I’ve been in Washington DC helping the Federal sector understand where Open Source software can help address some big challenges.

Q: What is your personal account of how the Apache  HTTP Server came about and why it was successful?

Brian Behlendorf: In 1995, acting for both wired.com and my own website company Organic, I was part of a group of early-web webmasters and technologists hanging around the Internet Engineering Task Force working groups on HTTP and HTML; many of us were also users of the free NCSA (National Center for Supercomputing Applications and University of Illinois Urbana-Champagne) “HTTPD” web server project, produced by the same office that produced the original Mosaic web browser.  Some of us had fixed bugs in NCSA HTTPD and had sent patches for these fixes back upstream for inclusion in the next version.  When we found out that NCSA lost all their developers in one fell swoop to a new company named Netscape, and the future of NCSA HTTPD was uncertain, we created a “fork” of our own, and decided to call that Apache and commit to collaborating amongst ourselves.  My code contribution to Apache was actually fairly minimal; I mainly focused on development process, development tools and infrastructure, and then on organizational and legal issues, and then advocacy, as we incorporated as a 501c3 organization and then scaled up to the size Apache is today.

Q: What inspired you to write the Apache HTTP Server?

Brian Behlendorf: To be clear, my actual investment in terms of code is de minimus, compared to the tremendous contributions made by many others, in both the web server and then other projects later (really, this isn’t just me being humble, it’s really true!).  The inspiration for this initial group of us to get together was partly practical – we had this pre-existing server, NCSA, that we were all using, and were interested in improving together – but also somewhat idealistic: we wanted to make sure that the Web remained an open, flat place, where anyone could set up a web site, and where Open Standards were reinforced by software people actually wanted to use.  We didn’t want to see the “Netscape web” be a separate thing from the “Microsoft Web”, etc.  We weren’t against vendors large or small at all – we just wanted a place where everyone could play on an even technical playing field.  It worked.

Q: Why is open source software important?

Brian Behlendorf: Open Source software is all about communities with common interests banding together to write common software, and sharing in the results. Participants can have a diverse set of interests and yet still be interested in the same code – a web server is a web server whether you’re a for-profit company, non-profit, government body, or individual setting up a blog.  Not every user will want to look at source code; even fewer will want to modify it.  But giving everyone the ability to do so is essential; and with that ability, big technical challenges can be addressed, software can evolve over time to meet evolving needs, and interesting things can happen.  Without exaggeration, we have an Internet today thanks to Open Source software built to implement the standards that define the Internet, and shared widely.  Today we see Open Source software not just in business and consumer applications, but in disaster relief software, education tools, microfinance, and more.  These projects exist because the Open Source approach allows a community to solve a problem that other models don’t.

Q: Where do you see the Web in 20 years?

Brian Behlendorf: From a technical perspective, mostly like it looks today.  We’ll probably have IPv6′s 128-bit addressing, and we’ll always have instability in one form or another (instability is a fundamental part of any complex system). But we’ll have HTML in some form, and email, and hopefully a decentralized approach to the new micro-blogging and social-networking tools that today are too centralized (more like identi.ca and OpenID rather than Twitter & Facebook) but I suspect that’ll arrive much sooner than 20 years from now. Thing is, we had things like that 20 years ago, too, on the net – we called it USENET and IRC.  So that suggests what will change isn’t the tech, but its impact on society by the way it’s accessed.  In 20 years, who knows?  Intracellular Ethernet connections?  Beats me.  I just hope that if society has to depend upon it, it’s open source.

Q: Which aspect of the current Web’s architecture do you believe is in the most need of improvement?

Brian Behlendorf: I’m worried that at this very point, what we are lacking most is a common “operating system” for the emerging “Platform as a Service” model in the whole cloud-computing space.  If I write to Google AppEngine, I’m bound to that service.  If instead I write to SalesForce’s Force.com, I’m bound there.  It’s not portable.  I don’t mind companies competing on the quality of their hosting services, or on introducing innovative new features I can optionally use if they provide some cool value, but I need service portability.  Most people understand the need for data portability from these cloud applications – export & import – but portability of apps across PaaS vendors is the next important step.

Q: What are some topics of discussion in your Keynote speech at ApacheCon 2009?

Brian Behlendorf: Here’s by abstract: Open Source communities like Apache have done far more than just create awesome, no-cost, flexible software – they have demonstrated a whole new way for communities of interest to band together and solve big problems. Beyond operating systems and web servers, Open Source software is now being used to organize disaster relief efforts, enable microfinance banks in the developing world, integrate electronic healthcare systems, and change the way citizens collaborate with their government.  How well are these projects going?  What should they emulate or attempt to inherit from the more “infrastructural” kinds of projects and organizations – such as Apache – and what new kinds of challenges emerge?  And where can you help?

ApacheCon 2009 Homepage: http://us.apachecon.com/c/acus2009/
Brian Behlendorf’s ApacheCon 2009 Keynote: http://us.apachecon.com/c/acus2009/articles/2009/09/02/featured-keynotes/

The Apache Feather logo is a registered trademark of the Apache Software Foundation. Images Copyright the Apache Software Foundation.