I've seen the theme of Flex is changed in the new version of Flex (4!) if you compare with the last one (3!).
2. How can I use the old Fx3 theme?
Another question about the new version of Fx4:
1. Why is FX4 so different and not so compatible with Fx3?
2. Why is datavisualization free in Fx4 and not in Fx3? A little bit crazy, no?
My personal feeling about Fx4: it's worse than previous one. It looks like a step back! No good if you would like to use in a enterprise environment.
Please, take example from other frameworks.
Thanks a lot in advance for your attention.
If you don't know how much about something, what is your basis for deciding "No good if you would like to use in a enterprise environment."?
It's abit like deciding that Toyotas don't stop so well because they changed the paint colour options.. (hey that could be it!)
Nothing is stopping you from using flexbuilder 3 and if you move to flashbuilder nothing is stopping you from using the flex 3 sdk.
Why is flex 4 different - simple its an overhauled framework derived from Adobes's understanding of what the development community need/want, and part of improving the inter-operability of the Adobe platform apps among many other things.
If you feel you would be better of paying for data visualisation I would suggest the term crazy is being applied in the wrong direction.
Whether Flex4 is or isn't better than Flex3 is a personal assessment, those that don't feel that the 4.x derivative of the sdk is for them have nothing to be concerned about as the 3.x sdk is still available and there is no sign that it will lose support from Adobe.
Flex has improved in many areas of usability for developers and after all its a development tool so that makes me as a developer happy, yes like any tool I use the feature wishlist is always bigger than the provided features, i'm a developer I always want more .
If you can tell me what you consider an enterprise environment I could then make an assessment on whether or not you actually know what you are talking about.
The world is full of development tools for different purposes, at the moment Flex to me is one of the better ones for RIA, it is evolving with the industry requirements and thats what I want, so Flex may not be for everyone but its a good tool that serves its purpose well.
It doesn't depend on the "color".
When you create a new version you must "compatible" with old one!
See the Java example. Sun creates Swing after Awt but Awt stayed the same, without any changes.
In addition, it could be better to use "deprecation" method instead to remove a functionality, whatever functionality you can image.
Using Fx3, I was convinced to use it in my company, like a standard UI of intranet applications.
Now I have some doubts...
please, don't misunderstand me!
Flex is best tool that I tested for UI in last 2 years! I agree. And believe me, I'm pushing a lot in my company (very big!) to use it.
But what happens now? I should re-factor some classes and spent time (and money) to adapt my codes to version 4 (I'm not using Flex Builder yet, I've planned to use it and I'm trying to buy it). Do you know what I mean?
Of course I can continue to use Fx3 but I'd like to follow last version of framework that I'm using.
My "crazy" was oriented to a strange commercial choice (why free in Fx 4 and not in Fx3).
About developers, you're right! I'm system programmer too and I need every time of new features. But my approach (and wish!) when I face a new version of framework is do not change my old code if is possible and I usually ask to new framework parameters or something like this to compile my codes like the past and add step by step all new features which are interesting me.
I'm confusd; what has Adobe removed from Flex 4 that was in Flex 3?
The Halo component set is still there; and still functions just like it did in Flex 3.
Flex 4 offers a new component set (The Spark Components); which is designed to allow for more flexible skinning. There are things in Halo that are not in Spark; true. But that is not the same as removing functionality.
Or are you referring to something different?
Deprecation is fine for point releases and Adobe have done that, but in reality a new version should be just that, Flex 4 isn't missing any of the bits from flex 3. Many flex 3 apps will compile as flex 4 apps by changing the headers and then moving declarative type mxml into the new declaration tags.
I have a 2 year old app (25,000+ lines of code) that took less than a day to get running after a few hours of cut n paste and tweaking minor errors here and there even got deprecation warnings on several things that took no effort at all to change(even though it wasn't required), CSS was a bit of an issue but didn't have much to deal with, now I have a choice and can avoid css altogether. It was a much less stressful experience to a VC 6 app I had to update to visual studio 2008, even 2005 to 2008 isn't a pleasant experience.
The industry is used to new versions breaking code, do you ever wonder why there are so many legacy apps across so many different development environments, the worst is big enterprise, the y2k fears weren't based on the lack of deprecation, it was based on the fact that half the worlds programmers didn't even know where to start with software upgrades when it came to large enterprise like banks and insurance companies, and now a decade later half the worlds developers(notice I moved from programmers to developers, no one gave me any form of deprecation warning when they changed my title and managed to squeeze in a pay freeze at the same time) don't even know what the y2k thing was even about, we live in an industry that changes rapidly, there are 2 choices suck it in and stay ahead or sit back and pray you have a job tomorrow.
dropShadowEnabled attribute (if I remember correctly) was removed from the component mx:*
Continuing using old components (not Spark) it seems impossible to remove the shadow from ApplicationControlBar.
in addition, the look&feel is completely different and (maybe I'm not the best coder of world and a little bit stupid ;-) ) I can't force the old one.
just 2 examples, not inportant but anyway like a little stone in the shoe...
I still agree with you for teh second part of your reply.
I don't agree with first one...
You should take a compiler of new framework, compile the old suorce code (difference of release is 1, for more I agree with you) without errors.
and then, when I'm ready to improve my application, using new features, I'll change my application.
just for you: at the moment I and my team are responsible for 6 times your amount of code... so .... :-(
I have to say, I've experienced some similar issues to what Stck is describing here. Halo seems to have become the 'second class citizen' component set, and this is hardly a good thing if you need to upgrade an app's SDK lib. I don't know if this was intentional or not, I'm just stating things how I see them.
(Stck may have meant this: The style 'dropShadowVisible' is only supported by type 'mx.controls.List' with the theme(s) 'spark'.) ?
Upgrading to Flex 4 isn't as seamless as going to a new version of the Flash Player. Unfortunately it's not just a matter of changing headers and moving a few mxml tags. It would have been nice if Fx4's backwards compatibility was as good as the Flash platform's. Perhaps that just wasn't possible this time around, but I do hope this upgrade proves to be the exception rather than the norm.
I still don't get the point of the argument, Flex 3 sdk is still being developed as 1) a fp 9.x supporting framework and 2) upgradability for fp 10.x,
mx based applications built with 3.0 sdk will compile with 3.5 sdk, and sure there will be an occassional glitch where deprecation has been overlooked but compared to other environments I have worked in, Flex is well within exceptable update behaviour.
Now stepping up from 3.x to 4.0 has to be considered as a major shift and once again flex is a pleasure to deal with compared to say VisualStudio or Delphi, both these environment have had minor version updates over the years that completely destroyed any chance of compiling without a significant overhaul of the code.
The fact that a pure mx application will compile under spark headers with minimal effort is a great plus for Flex. The ability to mix n' match mx and spark is a nice touch. We will all find areas that don't cut it and sure flex 3.x to flex 4.x will cuase some major headaches for some but the reality is unlike say a shift from delphi 5 to delphi 6(1999) which cuased every application and component to fail under compilation from the surface right down to the core.
We have the 3 sdk which allows us to extend and maintain legacy code, we have the 4 sdk with which we can move forward with, if we are in a hurry at least we can keep a great deal of our legacy mxml and almost all our legacy as3.
Did Adobe get it perfect ... no, did they get it right yes....., now for the 5.0 sdk
I realize it's nothing new to software, but excuses based on the failures of older technology are weak. I hope Adobe is not losing focus on one of the prime advantages Flex/Flash has on it's competitors. That said, Fx4 is not a show stopper, but... it needed to be said.