 |
|||||||||
|
|||||||||
|
|
You may have seen some of my posts about Acropolis both on my blog and in the Acropolis forum. One blog post in particular that I made provides a general overview of what Acropolis is and what problems it intends to solve. Yesterday Microsoft released the July 2007 CTP of the Acropolis framework. According to the Acropolis team blog, here are the new features enabled in this CTP:
- Transition Animation improvements – You can now add transition when changing page in TabLayoutPanes!
- Better Design Time support
- Better Custom Theming support
- Various bug fixing driven from community feedback
Ok, so while I realize that this is just an update to a CTP, and not a conversion from CTP to Beta.. my inner architect is writhing in agony as I read this. What did they add after all of the initial feedback they got from the developers (myself included) on the forums? More transitions, better animations, and custom themes. The "Better design time support" actually refers to the design-time ability to change a theme and that's all they refer to in the blog.
When I scroll back through the forum posts, I can't find a single person asking for more transition animations. I found one occurrence where someone wanted access to more themes. Most of the concerns I found were deeper than that. People (like me) wanted to be able to send loosely coupled notifications between two parts without requiring the use of a parent part container as a notification router. When you force a parent part container to be a notification router, you lose the loose coupling because the container needs to know which parts fire which events, and to which part those notifications must be delivered.
There were also a lot of concerns from people that a lot of the functionality from CAB was missing in Acropolis, despite their obvious similarities and nearly identical goals. Also not addressed in this CTP is the fact that you can't create additional external (e.g. non-MDI-style) windows that contain parts without resorting to some ugly hacks.
The previous CTP of Acropolis felt bloated, slow, and the development was tedious. I kept telling myself that the suffering is for the greater good because a composited, loosely coupled, building-block style application will be easier to maintain and easier to upgrade and be more reliable in the long-run than a "straight-up WPF" application. The problem is that really isn't the case... there are a lot of ways in which Acropolis could be simplified to make the development of composite applications more simple, faster, easier, and far less confusing than it is now. I would like that stuff, which I consider core and fundamental to be addressed in upcoming CTPs... None of the developers seriously considering using Acropolis to build frameworks by which teams of developers can become extremely productive actually give a rats ass about lipstick features like transitions and themes - that's eye candy that can be dropped on top of a fully functioning application later. No amount of animated transitioning is going to save an application built on a weak house of cards, whereas an application built on a solid building block foundation can be easily skinned to look beautiful later.
Anyway, it appears as though I'm passing on the July CTP of Acropolis for now. We'll see what August brings us :)
It's still in CTP form, looks to be due for delivery sometime in Q1 2008
since Orcas itself is due for Q4 2007. That means we still have a lot of
time to try and convince Microsoft to turn this thing into a valid
application construction framework... or enough time to be royally
dissappointed, depending on whether or not you're an optimist.
I'll volunteer for the pessimistic side and let you be the optimist. :)
