Op vrijdag 18 augustus 2006 15:32, schreef Dries Buytaert:
Note that there are 3 systems here: (i) flexinode, (ii) contrib CCK and (iii) core CCK.
I can't speak for (i) or (ii) but if they are "hacked up" then their maintainers are to blame -- at least to a certain extend. I can only speak for (iii), and I think you're overly pessimistic about it.
My idea is that, if core releases as it is now then (i) and (ii) should use (III) as API, or for the heavy lifting. Especially since iii is done very nice. Since core does the node-type handling different from CCK-contrib node type handling, these two conflict right now. Two options: Rip out the node-type handling from CCK-contrib. Which, if done right, is a near-full-rewrite. Same goes for flexinode. Having both flexinode and CCK in head is very confusing, for a starters because there are two nearly similar "content type" interfaces which even share a menu-path! And yes: Metadata-based content is the way of the future. My whole point is that right *now* we (flexinode and CCK) are getting rather close to something that is *almost* production safe. Flexinode in its current hacked up state, CCK in its 4.7 branch (where slowlly some fields are released). But both *are* not ready. And if upgrading both to 4,8 pulls out all resources again, there will be no stable CCK/flexinode for a long time again. Not for 4.7 nor for a long period in 4.8. I am not talking about what-would-be-cool-for-core. When I am talking about building sites Now. Today. Yesterday. Bèr