I want to get started with my game idea but I know 4.3 is coming and will finally officially support 2D.
I kind of would hate to spend a lot of work in learning plugins and such that will quickly become left in the dust when this 4.3 update comes out, but how long can I wait for 4.3?
Should I just get started and do it anyways?
Just my two cents; I have ALWAYS found it is better to just jump in and work with what you’ve got rather than waiting for the next big thing. Some of your work will always be able to port to whatever new system or technology comes out, even if that “work” is just finding out what not to do.
Everything is always right around the corner all the time.
If you wait for tech or tools to fit your needs you will never get anything accomplished. The primary survival skill/tool for success in the game industry (and tech in general), is the ability and desire to adapt and always be learning. Tech will change, new tools will appear or current ones evolve. Often multiple times during a project. Focus on the game, and using the tools you have (or build) currently, adapt as needed.
If you start building a 2D game today in Unity, everything you learn building it will apply if the tech changes. How things hook up, some structures may change, but that is all trivial. How 2D content is handled shouldn’t affect your actual game logic too much at all.
And don’t worry about learning things that may change. It does and will happen, its something that you just have to accept in this field. But make the best of it, when learn new tech and tools, just focus more on process than details, that is the part that will change. The new 2D Unity tools won’t drastically affect how you build a 2D game, it will just make parts to the implementation easier/better.
+1 to what they said.
An asset like 2D Tool Kit is pretty cheap and it does everything you need (probably be better than Unity built in 2D for some time).
In this case I would wait. There are other things like graphics, soundfx, menus and cutscenes you could work at in the meantime
Depends on what you have to do. I’m progressing other project while 4.3 doesn’t come out. However, if you have absolutely nothing to do, then go for it.
(It was a happy coincidence, when I finally thought about doing a game in 2D, I saw the 4.3 announce in Unity main page.)
A very long time ago I designed graphics hardware. Since making chips is so very expensive, chip design involved a series of prototypes. In my case I made emulations in c, then abstract designs in VHDL, and then gate level designs, which could be sent off for fabrication. Each of these designs was thoroughly tested, so that by the time a silicon foundry fabricated the chips I could be relatively sure the chips would work correctly. This hardware-style design is how I think of game design. You just make series of prototypes, each more advanced than the previous and using whatever technology you have today. If you make a 2d game using out-of-the-box Unity features today, or use ngui, or one of the other plugins on Asset Store, you can make rapid progress with your game - today. If in a few months using a 4.3 version of Unity looks a better approach, simply rip out what you’ve got, and re-implement with the new 2d features. You’ll have learned a huge amount about Unity, and you’ll almost certainly make the new version ten times quicker than the earlier version.
As @nggs and @zombiegorilla said, never wait for the next best thing.
As was said above, don’t wait to start. That said, it also doesn’t seem very wise to invest in something like 2D Toolkit with something coming for free in a few months that looks way better. I would personally recommend starting your game with Orthello, which is a free 2D framework. That way, you don’t waste time or money!
Tell me how Unitys 2D solution is “way better” than 2D Toolkit.
The physics will/should be “way better” :). But other than that, who knows.
Well, allright, I’ll admit this is really better, but 2DToolkit will adapt and work with the 2D Physics. Personally I won’t turn to the native 2D solution, I’m way to happy with 2DToolkit.
we are always waiting something. unity makes it better and we are waiting assets to be compatible with new versions. i test unity 4.3 . it looks nice with it’s 2d features. so far i used 2d toolkit. i am happy with it. now i need physics in my game. i use 3d physics engine of unity. 2d toolkit is not compatible with unity 2d engine because 4.3 is not released yet. i will be waiting it to be compatible with 4.3. i have ragespline and ragetools. nowadays i learn to use it. also it is compatible with 3d physics and Farseer. Farseer is different from unity internal system, so i do not want to learn and use it. so i will be waiting that rage products will be compatible with unity 4.3, maybe it will never be compatible with it.
so always there i confusing items 
i think i will go on using 3d physics until they will be compatible with unity 2d system.
Yeah, like everyone’s pretty much said: just go for it.
That being said, I’m waiting on 4.3…but I’m not just “waiting”. I’ve been working on the art and furthering my animation skills (which were in need of work for 2d). I’ve also worked on implementing a Unity-like ECS in XNA and a Unity-like environment between XNA and Tiled.
The takeaway? You can wait for 4.3, but don’t just wait. DO something. If nothing else, start planning out how your game is going to work using psuedo-code. Or start writing the physics/rendering-agnostic code. But DO something. Time spent just waiting is time wasted.
I’m on the same situation. I’m working on a 2d project with ragepixel. It’s a very neat package! but for my specific needs I found it inadequate. With a funding of zero dollars my options are a little limited. Therefore I’m on the waiting team, it’s not like there’s nothing else to work on in the mean time 
I just hope it really comes out this fall.
+1
A few weeks ago, a Unity evangelist came to a developer meetup and demoed the Unity 2D features live with the beta software. The way they’re integrated into the editor is incredible and the initial setup of all the art, animation, and atlases* is an order of magnitude faster than I’ve done in the past with 2D Toolkit. I was spoiled in the matter of 15 minutes just by watching it.
That said, once the set up is done, the game-building and scripting workflow is probably roughly the same, with the exception of using Box2D instead of PhysX.
*And by that, I mean there is no setup of atlases. It just automagically generates them somehow.
I know this thread is a bit old but wanted to provide my feedback.
I’ve had the pleasure of trying out the 4.3 beta, and having used Orthello Framework for Unity3D before… I must say that the integrated solution is way easier, faster, and better in any sense in regards of workflow easeness…
As Dallon said, after the initial setup the rest is very similar, but setting everything up for a 2D game in unity3d 4.3 is a real breeze.
Cheers!.
I imagine Unity’s 2D will be faster than any other third party system because it’s integrated internally in compiled C++ rather than in managed scripts, much like how the particle system is faster than you could make third-party. The imminent release of Unity’s 2D does raise question of how deep to get into 2dtoolkit etc… for me too, but for now I’m just going ahead with what works now with an eye toward making it possibly easy to upgrade/transition to the built-in system later. Otherwise all this stalling just holds up a project.
Till 4.3 is released you can try working on your artwork/artstyle. I find it always good to have a mockup of what you want your app to look like exactly, and aim towards that. That’s what I’m currently doing and it’s apparently quite a task.