francescociriaci's Blog

I’ve been there… and it’s all true: the JBoye conference is addictive.

Graham Oakes, a friend and a guide for me, told me not to miss the JBoye conference, that he (as many others) consider as the best worldwide CMS conference. So I planned my trip and after last preparations for the World Plone Day, which was unfortunately the same week, I took the plane to Aarhus, Denmark.

The amazing volunteers at the J.Boye conference.

The city itself is very nice but the welcome evening was even better: Mr. Janus is an excellent host, personally welcoming his guests and remembering details of everyone (he tagged me with the keyword “Plone”, of course). The whole conference was perfectly organized: friendly, with lots of occations for networking, every detail taken care of, and probably the best food I’ve ever had at a conference.

The tracks, talks and keynotes I attended where excellent. A real benchmark for the whole CMS industry. In particular I appreciated an excellent open debate on Social Software by Tony Byrne, the networking (across and beyond the organisation) by Bjørn Guldager and the case studies of “killer” intranets by Toby Ward.

At the social dinner I also had the chance to sit next to Tony Byrne, who managed to surprise me on how deep understanding one person can have of the extremely complex CMS landscape. I also met Jim Hobart, a (uber) user interface design consultant (internationally recognized, yes, but I did not recognise him at first).

With them and many other people I could discuss about the future of CMS and trends in the market, about interesting user scenarios and … Plone. In fact I was the only Plone evangelist at the conference and I really did enjoy to discuss “where-Plone-is-heading” and some of my favourites and evergreen stories about Plone.

Seth Gottlieb (Content Here) was there, we already met, at the Plone Strategic Summit in California; he told me “believe me, this conference is addictive”. He is right.

Last but not least Mr. Janus invited me to submit a talk on Plone at the next year conference: I think it’s a great opportunity for Plone to show up and be more visibile. And believe me, after a full immertion in the CMS world the feeling is once again that Plone is a great CMS, especially in regards of usability (present and future), key factor for the success of any CMS project.

The conference has motivated me to tell the world how Plone can be interesting and to tell the Plone community that Plone still need to evolve and grow. A lot.

Here we go with the last brainstorm in Reflab, started from both an observation of what the (Zope3) technology allows us to do, recent product developments (such as geolocation or relations) and some works and thoughts by Martin Aspeli.

The basic idea is to rethink the concept of Content Type itself.

In a few words current idea of a content type in CMF/Plone is the sum of a data structure (the Schemata), plus some behavior (Note: the word “Behavior” is used in other technical contextes with other meaning: in this particular context meaning is purely non technical and related to what a content behaves in a CMS from the User perspective.)

Examples:

A News Item has Title, Description, a preview image, a text body, etc. as main fields (data). A News Item, because of it’s type, shows in the News portlet when published, and shows it’s preview image in listing, cannot be commented, shows in searches, etc. (behavior).

An Event has Title, Description, a text body, a lot of contact information, etc. (data). An Event shows up in the calendar portlet and Events portlet when published, shows in searches, can be commented, etc. (behavior)

This are very basic examples of course, but the should alreay give the sense of it.

So the question is: why content cannot simply be “newsable” or “calendarable” or “mappable” or previewable”, etc. despite its basic structure? In fact many simple behaviors are already separated from the data structure itself.

Here is a  uncomplete overview with some add-on products as well:

What I’d like to work on is on complete separation of the two (for all contents, defaults and add-ons): from the user perspective there are loads of use cases and scenarios where an extremely decoupled approach makes sense and would make the life of integrators and custom products developers a lot easier.

In fact this kind of new products development has already started: no more content patching or subtyping when you can adapt. And very interesting, tiny, small and simple products are popping up.

Still I’d like to point out that the process started from the very developer-oriented perspective and often the user and the integrator are left behind.

This is only a first braintorming on the idea, I’ll stop here for now.