Voting starts in March for the Drupal Association Board election.
project requirements (very common, i believe):
- rich-text editing for adding content
- ability to insert images into content via a graphical interface
- ability to insert images in a non-templated way (e.g. size and placement may vary between nodes)
- ability to reuse images in multiple nodes
- whether images are nodes or not is relatively unimportant
- forward compatibility!
i've listened to the Lullabot "deprecated" podcast, and read their "Image and Image Exact Sizes vs. Imagefield and ImageCache" article. i know that lots of folks are recommending moving away from the Image module in favor of a CCK-based approach. but it seems to me that there are some things that the alternative CCK + Imagefield solution can't handle.
1) arbitrary placement -- with a CCK-based approach you'd need to theme your images in a standard way, e.g. for all nodes of a particular content type
2) wysiwyg/rich-text integration -- not possible to achieve w/ CCK and Imagefield, right?
3) private files -- a bit of a tangent, but files must be public to use ImageCache, and that seems like a non-starter for any site that needs privileged content (e.g. certain files only available to certain user roles)
in general, a CCK based approach seems fitting in situations like the Lullabot article above where you have a specific set of images needed for a given content type, that can be themed in one templated way. but it doesn't seem to be workable for a more open-ended situation where authors need the freedom to add images in various different ways. am i missing anything there?
so, it seems that for a set of requirements like those i've sketched out above, we're left with either:
A) Image + Image Assist
B) IMCE (if using TinyMCE for the rich-text editing)
and the choice between these two will come down to whether you want your images to be nodes, and ease-of-use.
finally, where does that leave us in terms of forward-compatibility? if Image is indeed deprecated, and some CCK-based approach is developed in the future that better handles the type of requirements i've sketched out here, how painful will it be to migrate our content?
any feedback is appreciated.