There's the Apache conference on 18-22 July 2005 in Stuttgart, Germany: http://www.apachecon.com/

Aslak is right. The web framework often is a minor thing in most serious projects. And at last, someone who feels the same about event driven and request driven frameworks. There is no either-or in those types. It depends on the website. Great article.

Stephan

Dion,

Great article. It's about time someone steps up and says it's not THAT a big deal what web framework you choose.

Aslak

"Some developers may still prefer something else."

Hmm. Not sure how that really qualifies as a "barrier to entry." That's simply: "not going to enter." A barrier might be "Ruby is hard to learn" or something like that.

Or perhaps what you mean is that some peoples' strong preferences for familiar frameworks/processes tend to keep new things like Rails from taking off quickly into the mainstream.

One of the really interesting things to me about Rails is how well it's been marketed. Most people who create frameworks don't give two hoots about publicizing or evangelizing them, or responding respectfully to their users' needs or suggestions. I think *that's* what Rails has done incredbibly well.

While I think you make a fair point, it seems a bit apples/oranges to me. One language is driven by a large corporation(s) and the other is free and driven by developers.

A better comparison would be Ruby vs Python in terms of how one framework got so much (relatively) mindshare--that is RoR in Ruby compared to a pretty splintered situation in Python web frameworks. The splintering hinders many people in python - interestingly, there's no "one obvious way to do it" (python's mantra). Rather, everyone writes their own framework. Which I suppose shows the language's productivity--at least for those with the time and skill to do it. Everyone else is told "there are 100 ways to do it, you get to choose the best for YOU." Which um, is great if you want to research 100 frameworks. Anyway, this is a tired story, but another contrasting example to Ruby or Java. And I think, proof that a strong leader (David of RoR) is very vital to bringing a project to fruition and mindshare/usage/enthusiasm. Maybe Matt Raible and Appfuse is somewhat similar in Java-space(?) [in terms of leadership, not framework]

I see your point. This is why I have always been in favor of competing solutions in the java world, no matter that some people may like the "comfort" of a solution being forced upon them, a la Microsoft and its minions. To me the definition of a vibrant community, one that long legs and a bright future, is one where competing alternatives provide fodder for selection by the "environment", such that in the end, you get only the "best" solutions winning out.

Another approach is to use a screen capture movie software program like camtasia (techsmith.com) or bbflashback (bbsoftware.com) to record a desktop demonstration. Record a decent demonstration and explanation the way you would show things to a colleague and give the movie to a tech writer. Let them reverse engineer the documentation. If someone is going to transcribe and correct it, you can stutter, hum and haw and it doesn't matter.

Just a thought.

Excellent article! We here at openflows.org have been pursuing an Open Source Services and Support Business Model successfully since 2000. Our strategy is to identify stable and innovative open source applications/platforms and then make the argument for their adoption and use. By providing support for said open source technology, it makes it a lot easier for clients to make the leap. What we've found is that the Open Source genre as a whole is quite infectious, and as soon as an organization starts to learn about the benefits, they then explore how open source can help them in general, and not just for the specific task that got them interested in the first place...

Important observations at a time when open source wave is rising. While open source documentation may be inadequate, there is an equal universe of inadequate documentation in the non-OS world that employ armies of technical writers.

Its a matter of being an user-oriented community, open source or not. For most parts, the documentation reflects the nature of the product too. There is a difference between JBoss and WebLogic docs, just as there is a difference in their orientation and overall usability.

Developers make very reluctant but not bad technical writers. One has good examples in the TLDP and the whole Apache range. We have had success in integrating engineering groups into technical communication channels at Pramati. It takes a lot of evangelising. "Its not about pages of documentation in good english, its about communication", is what is told. We got some tremendous communicators in the product engineering groups. They are speaking at events as well as writing documentation for product users.

I would like to think open source documentation, like open source software, has the stuff. It just needs the finishing, a round of edits, good coverpage and an exhaustive table of contents. Just the way the new crop of open source statups are giving finishing touches to open source software stacks.

Your opinion seems valid to me as I express in my blog entry at the given URL. I like your site, great work!

Best

Klaus

Hi Bob, interesting collection of articles showing up: keep up the good work! If you're looking for cases of small companies trying to make money on open source development, please review my recent interviews @ uk.builder.com: http://uk.builder.com/manage/business/0,39026582,39225378,00.htm and http://uk.builder.com/manage/business/0,39026582,39229656,00.htm

Cheers, Steven.

There is another angle or two to this. As in the OpenLaszlo 'crossing', where better to go for support than the originator of the software, in this case JBoss. The other thing I can see as a differentiator is that some smart corporates see the 10,000 pound gorilla for what it is, not a nimble 30 pound monkey. To elaborate, a smaller firm often gives more focussed and personalised service. They specialise and there is not the continous conflict of interest where the 10,000 pound gorilla is always trying to sell you more services/kit from their massive range.