A very late response, but the subject is important for MMBase if it wants to evolve. Here are my thoughts about the subject. First, the document is not clear about what problem it solves and that makes it harder to respond to it. That is ok, because like Gerard already wrote it is a start to create a vision about the subject.
The problem is complex, there are different types of problems. But maybe we could add it as an introduction to this document.
MMBase has evolved to a good competitor of other (commercial) CMSes. The
main difference with other CMSes is that MMBase (core) is more a content
engine instead of a full featured system. A CMS/site builder has to
integrate pieces to get the desired system. Every site build on MMBase is a
good example of these systems. Well known 'open source' ones are wiab and
didactor.
(where can I download this one?)
Examples of pieces are editors, editwizards, security
implementations, remote cloud communication, workflow, polls, forums, chat,
and email. IOW, the power of MMBase is that it does not limit the system to
the features which it provides. The weakness is that you always have to
integrate pieces on the engine which takes time and knowledge. IMO, the
first outweigh the second.
For this integration we need (a) tool(s).....
With this in mind, the question remains what organisation and infrastructure the MMBase community needs to be most effective. IMO it is not in the first place the hosting of full systems. The first step is to have an infrastructure where the pieces can be easily provided to the community. At the moment, a commitor can add something to the 'speeltuin' and announce it on the mailinglist. The MMBase community is a lot larger then this elite group of people and the pieces are hard to find (Not everybody wants to browse the cvs to see what is there). A site where you can find and add info about pieces would be an improvement. Info about a piece can consist of a description, author, mmbase versions tested on, homepage, license and maybe a download link. These pieces can also be divided into community supported and non-supported pieces.
I agree this could be the first step.
A next step might be to provide server facilities (cvs, mailinglist, webspace) for pieces/systems build by people on different locations. This step is almost like what apache and sourceforge are. The difference will be that the systen has to add value to the MMBase community to get the facilities. People building those don't necessarily have to be a core commitor of MMBase.
Yes, that's why I think we need different groups of commitors, like apache has.
Gerard
