Issues which need to be solved so DeepaMehta becomes (for me) a recommendable (web) application framework for developers:
- #345, #42, #467, #510, #539, #290, #311, #562
- Input Validation Mechanism
- ACL.Read-Permission / user-private infos
- offer powerfuly data-query features (partial objects and paging)
- fix missing documentation on the most basic formats (see latest
I think ingeneral, a framework which wants to support the development of innovative user interfaces nowadays must at least offer good foundations for each of the following:
- stable (persistent) and personal (=user-centered) geometries,
highly individualizable item/entity renderingrealised in a contetx of a topicmap- out-of-the-box user-input validation techniques,
- offer a declarative way
for the creation of arbitrary and complex user-dialogs- this uses jquery-ui
A plus would be:
- utilities for working with 2D/2.5D space,
- allowing constraint-based development and/or data-binding
- offering a scene-graph like rendering engine,
- a neat and simple way to deal with i/o in the gui
- and if framework would support js and css aggregation /optimization
As well dealing with a pretty and stable API as with having a reluctant dependency/library management, saves work (and therefore lifetime).
I would consider any framework which supports this set of functionality already very helpful, if everything would be just about developing new user-interfaces.
About developing software platforms in general, I will keep it with (1) the Google+ rant of Engineer X on doing what your best at (in this case, help other people/developers to develop) and (2) doing that right as well as find some very good answers on why you need to start building this at all.
