Evolution of the Cathedral

In his post about the changing cathedral and bazaar, Jon Udell asserts that the cathedral and the bazaar are both changing. And of course, they are -- that which stands still, dies. At the same time, my core point is that the "sharp edges" I'm talking about are manifestations of user experience. User experience is one of those things best tackled by people who aren't expert users but by people who are experts in watching other people use their software. User experience remains the last frontier -- one where aesthetic meets science. Hillel Cooperman, lately of Max and previously of AERO, describes user experience as more than just the User Interface. The user experience is the complete end-to-end interaction that the user has with a piece of software, including:

  • the user’s state – the aggregate of their goals, expectations, mental model, emotional state, physical and cognitive abilities

  • the individual elements of the interface, including all the controls and widgets

  • the user model, including the interaction model, the data model, and the concepts and abstractions

  • the underlying system bits that lets the user move between different pieces of the software

  • the aesthetic experience of the interface, including the visuals, audio, the tone and language of the text, and the general emotional impression the software creates

  • the physical hardware and devices connected to the PC (and their aesthetic experience)

  • and the environment in which this entire combination is being used

Hillel's team also did a huge amount of research into who our users are and, like Jon, found that they're evolving and raised the point that you need a huge amount of information in order to see precisely how they're evolving. This touches on everything from traditional usability testing to gathering information from error reporting services to even starting to measure emotional response.

That kind of investment is hard. Very, very hard. Novell recently released 200 hours of usability videos. Watch them and see how many sharp edges there are to remove, and think about how much work it will take -- some of it core architecture work -- to address them.

So can the bazaar, in theory, fix this? Sure. In practice? I'm much less sure.

Comments (2)
  1. Mike Dimmick says:

    I never accepted that Windows was an instance of the Cathedral in any case. It was a strawman that Raymond set up to attack, and represents the model of the 1960s and 70s much more than the commercial model of today. The fact that the API is open means that the Windows ecosystem is a bazaar – much more so, in fact, than the Cathedral structure of any given open source project.

    Indeed, in most closed source/commercial projects, developers all have check-in permissions, with a greater or lesser degree of pre-check-in review by a fellow human. On the other hand in Open Source projects very few developers have the right to check code into the master sources, all others having to submit code through the official maintainers. That tends to concentrate power in the hands of the maintainers, and they can be difficult to dislodge. Power is much more distributed in the closed-source case where the aim is – or should be! – to get the most out of every developer, since developers are a limited resource.

  2. johnmont says:

    Totally agree. It’s argument by analogy, and not a good analogy.

Comments are closed.

Skip to main content