Gianugo gives us a CMS rant, talking about how the full-fledged CMS solution may be dead.
I'll declare up-front that I have a vested interest in this issue, as we built a fair chunk of the solution Gianugo refers to, in collaboration with Pro-netics - specifically, the publish / layout part.
Does this mean I think the monolithic CMS is dead? Er, no. It will ultimately come down to market forces.
For some people, the "do everything" solution from companies such as Vignette is absolutely the right one: they want to tick boxes on their requirements sheet, and have the theory of large company support behind them as they move forward into implementation. I'd hazard a guess that this fits well with medium-sized enterprises, where staffing is perhaps an issue and back-end solutions are limited and not overly complex. Ideally I think your "boxed" CMS should be kept away from any kind of integration - if you can run it standalone, it'll do an admirable job.
Others (perhaps those with more empowered I.T. staff?) find an excellent solution in products such as Plone, Apache Lenya, Daisy, and many others. A certain amount of willingness to tinker is required here - these aren't shrink-wrapped, and you need to be prepared to ask questions. One environment I've seen that is particularly well-suited to this is the academic one - the price is right, and most educational institutions should certainly have staff with skills to handle this. I'm not by any means saying this is the only environment - just one particular place I'd like to see wider adoption.
And then to go back to Gianugo's reusable components solution: the best place for this is surely where the organisation is facing tough integration issues, where any solution will need a significant amount of customisation. From what I've seen, these are really large companies caught up in the vendor myth ("no-one ever got fired for buying IBM..."), who need to be nimble in a rapidly-shifting and competitive marketplace, but are tied down to legacy back-end systems and legacy waterfall development models. Agile components could really save them - but will probably never be given the chance.
Posted by savs at March 21, 2005 8:21 AM