Aral Balkan wrote: > 1. Focus Handling? > > ^^^ Big, big, big, huge! :) (Usually the last thing considered when > constructing a set -- believe me, it's experience talking -- and the > hardest thing to retrofit!)
Ya, i agree - this seems to be something that is overlooked all too often. > 2. Ease of skinning via MC's (Ideally also allowing for programmatic > skinning too.) Skinning seems to be so simple, but as soon as you start to look at it a bit more closely it becomes a nightmare. It always seems that the more flexibility you want to add to the framework, the more complicated and complex the system has to become. Finding that balance is extremely difficult. There is also the reality that some things are just better done dynamically via code rather then using pre-baked MCs. I think i might have to start a thread about skinning - not just GUI components, but in general, and how people deal with all the issues that come with it. I have a hard time holding it all in my head... > 3. Simple to use API / polymorphic with MM components (as much as possible) This was what i was thinking too - make it simpler for people to switch. It has drawbacks too - how much do you want to inherit the bad bits of the v2 components just for compatibilities sake? Cherry pick the good bits, leave the rest... >> - CheckBox >> - RadioButton >> - ScrollBar >> - ComboBox >> - TextInput >> - Button >> - ToolTip > - Containers (Box, Panel, Window, etc.) > - Data grid > - Splitter > - Loader (with Preloader) By loader do you mean like a progress bar, or something like the MCL class? I notice that you don't mention Data Binding. How important is this feature? What widgets gain the most from it, which _require_ it to make them usable? What do we loose from not having it? Thanks for the feedback. j -- Jos Yule Digital Hyakugei "Machines take me by surprise with great frequency." - Alan Turing _______________________________________________ osflash mailing list [email protected] http://osflash.org/mailman/listinfo/osflash_osflash.org
