After listening carefully to feedback from our user group we decided to change the rating procedure.  We are currently implementing a new approach that will be launched soon. In our next version users will have to authenticate themselves (and therefore register before if they haven’t done so yet - it’s the same procedure as for the forums)  before they can rate.  To complement the rating people will receive both a feedback on what the rating actually means and the possibility to enter a comment to support their rating. With this we hope to receive valuable additional insights into usage experience and more objective rating input. We are interested of course to receive feedback from the user community on whether they like this new approach or not of course. The changes should be online in a few days.

Bookmark It

Hide Sites

Ross Mayfield argues that it is “Time for Web 2.0 to be Unleashed with Open Source.”

According to Ross:

Web 2.0 companies are largely built upon Open Source software. But how many of them do you consider significant contributors to Open Source? In general, there is an open ethic, and communities demand (and reward) it. But somewhere along the way, the focus shifted to APIs and Open Source wasn’t rationalized as part of the business model. Some call them Open APIs or Open Data, but until there is a legal framework adhered to as community standard (word is OSI will work to address this), they are just APIs with unilateral rights. And with the focus on APIs, instead of contributing code back to the projects you leverage, or contributing your own projects, cooperation has been limited (save a handful of great standards efforts like Atom) Business models have also been held back by the gradual evolution of Open Source licensing, until now.

Until, that is, the OSI approved the Common Public Attribution License.

While I’m not convinced personally that this change (availability of an additional license option) will single-handedly bring about a shift such as the one Ross describes, it will undoubtedly make possible additional innovation in the software-as-a-service arena, and hopefully will broaden the set of enterprises who feel comfortable opening (portions of) their codebases.

I’d love to see new social network platforms and applications, for example, take the step of making not only their APIs but their codebases available.

I’m reminded of the contrast (which I’ve blogged about elsewhere) during the O’Reilly Executive Briefing at this year’s OSCON between the Mozilla Firefox extension development platform and the Facebook API: how much richer the context was for Firefox plugin developers, and how controlled and limited the environment is for Facebook API developers.

Mozilla didn’t need the CPAL to enable such openness, but maybe other projects will feel that they can now release code they would otherwise have held private.

Bookmark It

Hide Sites

Good panel discussion from OSCON about the Art of Community moderated by Danese Cooper, with (left-to-right):

• Jimmy Wales
• Dawn Foster
• Sulamita Garcia
• Whurley
• Karl Fogel
• Brian Behlendorf

Bookmark It

Hide Sites

When defining the categories and selecting the projects there was a hot and long debate on whether to include Programming Languages and how to compare and rate the individual projects. For example should a programming language be benchmarked against the generic needs and requirements you would list or against the usual domain the language is used in? PHP for example is mostly used for web applications, so should only this domain be looked at? Many people in the expert group voted for not including programming languages at all. Others including me claimed that programming languages are actually one of the most important categories in Open Source software over all and can’t be excluded. We finally decided to include them but there was still the discussion on how to compare them and to decide whether for example PHP is really higher in “enterprise readiness” than Perl for example. Well, after long discussions we finalized the ratings and are following the answer of our users with high interest. From a popularity point of view we seem to have made the right choices with the exception of Ruby that seems to be very popular with our users and even beats Java in popularity. From a user rating point of view Ruby and Perl are more positively rated by our users than what we proposed. But this is just the current assessment and gathered feedback at this point in time. We can expect many more debates and discussions over the coming months. Stay tuned!

Bookmark It

Hide Sites