The icon style exists in the most recent version of Flex 4 (Flex 4.5 hero), just grab the prerelease.
I need to create custom skins? What if I dont want skins. I was perfectly happy handling all my styles through css, it was plenty powerful for our uses and adding icons is simple. Things have gotten way more complicated in flex 4. It is not an upgrade but a downgrade.
I've been saying the same for probably a year. You're not the only one! But I don't recall ever getting a single response from anyone at Adobe to even voice their awareness that SPARK and the supposed "designer-developer workflow" it's supposed to support does not fit all use-cases, such as ours. In cases like ours (if I may include yourself) we don't have dedicated designers (because our users just don't care how the app looks)--the developers are the designers for the most part--so Spark and Catalyst, a crummy broken/slow and ignored Design View feature, and more "powerful" (yet much more complex) skinning does nothing to help us at all.... and actually makes our workflow more complex. (Not to mention the migration/update hell Spark has left in it's wake.)
Enterprise developers like Flex because of increased productivity, platform (browser) independence, consolidation of front-end development technical knowledge needed, avoidance of extra tests (to support new browser builds), never being forced to patch anything to support a new version of a browser, consistent development platform, flexibility to run offline in AIR or web based, and so on. It has little or nothing to do with fancy effects or the look and feel beyond making things look presentable and modern for executives and management to be happy with. Either Adobe thinks we're a small minority of the Flex developer base (doubtful, anyway enterprise is where the money is), or they've just miscalculated the amount of extra work SDK 4 has created for those of us who were 99% happy with Halo.
Heck at this point I'd be happy if they just "un-deprecated" the Halo components that the API docs currently badmouth now. I don't like developing apps with a virtual dagger hanging above, but it's either that or Spark. (Or Silverlight...geez.) Preferably Halo would become a first class component set once again instead of supposedly being Spark's mentally challenged relative that you should avoid at all costs (yet we're forced to use it for missing components!)
I understand that Halo had some issues if you need to do heavy customizations, but Halo was not a failure in every case. User experience matters, but efficiency matters more!
4.5 is still way to buggy in design view, in fact there are still a lot of design view bugs in flex 4.1
I agree with what was said above. Flex was perfect in many ways for us. We do care about app design and put a lot of effort into our interfaces but the halo components worked really well and were very easy to skin. I like some of what they did, I like boarder containers and dropdown vs combobox. I like not having vbox and hbox but a single component with different properties. But overall spark is a PITA. WAY WAY more work to do simple things.
In addition the way state code works is awful. We have a HUGE app with many modules and each module has many files with multiple states. It worked well in version 3. What is bad now is, lets say i build a base state, then I add some more states. If I now add an item to the base state, the default in flex 3 was it was inherited. Now its not, it makes it specific to that state, which creates more work. I do not know when we possibly will make it to flex for due to the time involved in some of this.
Design view needs work, please!! I am our designer/front end coder. Catalyst is just too cumbersome, way more work. Why did they need to mess with a good thing.
Design view was one of my biggest pro's for choosing Flex over a lot of other choices. It definitely needs more attention, performance fixes, bug fixes, and better support for displaying what you're actually going to see in Flash. I'm having a heck of a time just getting a custom font to show up correctly right now.