Is it going to be just the rewritten User Interface layer, acting as a combined gui/2d game system?
What will be the features?
Nobody knows for certain how it will work yet outside of UT - I’m sure we’ll see some sort of press release as we get closer to the 3.5 release window though.
Been wondering this myself as 3.x is supposed to have some sort of full blow 2D interface engine in it, supposedly, am seeing a few 2D systems show up in the asset store, but holding off until something official is released by UT, I want more than a GUI remake, I want something that can blow Torque 2D work out of the water.
@zumwalt:
me too (or a very smooth hybrid mix of 2d and 3d each rendered with it’s own dedicated renderer) 
Yes, Unity must be made into a convoluted mess trying to be all things to all people. A 2D Unity outside of a new engine is a terrible idea when there are already good 2D engines. If people insist on making 2D games with a 3D engine, let them, but don’t accommodate them.
@JRavey:
Nobody said anything about a seperate engine for 2d…
then you might work forever potentially.
thats a massive endevour and nobody is gonna starting it until UT finally gets their lacking roadmap handling fixed as nobody invests months just to be backstabbed.
We know that a 2D sys is gonna happen but I personally don’t suspect it to be much more than early SM2 / SM1
@saymoo: nobody did nor did JRavey but he said that anyone who wants a good 2D has the option to use one of the many available instead of hoping / dreaming of getting a 3D engine bent and ripped enough to do a good 2D job …
I interpreted
“A 2D Unity outside of a new engine is a terrible idea when there are already good 2D engines.”
as an besides the 3d version an outside version that does 2d. but if he did not: i withdraw my comment.
Anyway, many engines do have a “subengine” build specific for 2d rendering (e.g. animated GUI’s and sprites).
So you don’t have to fake it, with projecting it on a plane which is rendered in 3d space (costs unwanted resource).
The current “2d subengine” is barely present in Unity. Iif you want a full featured 2d like setup, faking it is the only descent way. Sure there is GUI implemented, but it lacks many things.
Looking at the roadmap, at least something is going to be done about (although not fully, which not to complain for now) in 3.5
Many folks here, ask for a good 2d featureset very often. I’m happy UT starts working on that (piece by piece, release by release)
Sprite + GUI existed for an eternity actually in free form, no need for unity to develop it. (GUI + Sprite Manager 1 - or SM1 + 1 of the gui systems that used it)
Problem is people expect 2 hells more than that with stuff that can compete with T2D or Cocos2D as implied above and on any other thread related to it and thats what I doubt we are going to see on 3.x at all.
And the blog is no roadmap. Thats a “what we are thinking about” listing and not to be taken too serious. I just want to remind that the 2D system and GameKit support were publically announced for Unity 3.0, and if you look around you realize they didn’t happen and there are various other things.
So take it as what it is, a potential foresight into the unity future, don’t even consider using it for any business relevant decisions
I’m partly not disappointed in Unity’s 2d lack of offering, because it makes room for developing my own 
I feel that they will probably spring the 2D library off their new gui stuff. The new gui stuff is intended to be as fast as it can possibly be with atlasing and so on, which makes it a perfect fit for future 2D games. I bet I am not wrong.
cocos2d for the win ^^
Normally you do it the other way round.
You need a working performant 2D system to develop a GUI on top, you can’t develop a performant or even feature complete enough 2D system on top of a gui system (otherwise 2D systems would be done in on gui or with guitexture ;))
Reason is simple, trivial OO relantionship definitions:
UI gadget IS A 2D Object
but 2D Object IS NOT a UI gadget
They will do this because it isn’t anything like what you’re describing. They’re doing a fast atlas based batched quad system and then fitting gui into that. You’re just inventing how it will be
The basic system will of course be 3D hardware driven, which secondary means that at some point there will be quads for images/gui-elements/sprites etc, which thirdly means there really will be NO difference between a gui element and a sprite. From what I can tell I think it’ll just be a new gui interface with an editor and it won’t really sell itself as a 2D game platform in its own right, it’ll be positioned as an improved gui which you happen to be able to use for 2D games.
I kind of bet it won’t have any vector graphics though.
I would love to see it, but agree not to hold our breath. UT is riding way back in the saddle and appears to have little interest in a gallop.
UT is clearly bowing under the weight of a dated framework and lackluster roadmap. The only innovations we see come from 3rd party license agreements, and the hard work of indy developers who have to pony up 30% of their income to UT. The ‘asset store’ was a great addition, but to UT’s bottom line. It’s a convenience to be sure, but a feature? One that Indy developers foot the bill for. At least it represents the latest in E-commerce technology, and Devs can count on their updates getting uploaded quickly… oh, err.
UT promised a 2d system prior to 3.0, they appear poised to deliver a badly needed update to their GUI functionality. There are a couple of solid contenders to the throne of ‘thee’ 2d sprite system, it’s just irksome that UT dips their hand into dev pockets and calls the privilege a ‘new release feature’.
Swimming in boatloads of cash doesn’t seem to have lit a fire under their butt, just the end of some expensive stogies.
To be fair to Unity they can’t just keep innovating new stuff all the time, they’re already doing many many things well and it doesn’t leave a whole lot of room for new things to be done. When the engine gets all the features you want/like are you going to still want them to keep finding new things to be added? Probably, but there are limits. I don’t think it’s totally fair to say they’re just resting in their laurels, there are few companies who I’ve seen put more effort into keeping improving things than Unity does.
@brendang - I don’t know, I feel like the 3.4/3.5 roadmap changes are some of the most exciting changes in a long time, even eclipsing 3.0 itself.
The only thing I don’t see in the roadmap that I would really like to see is a move to Mono 2.8
I’d also like to see that, but I would guess a Mono update won’t happen until Unity 4.
–Eric