Unity 3.5 UI

Hello,

Let's for the sake of this discussion assume Unity 3.5 will include the new UI system. My question is: Do you think we our old code will get the upgrade? I assume it will generate the same output in the new (hopefully much faster) system. Or do you think they will make it separate and keep this UI system in place?

I'm asking this because I'm deciding what to do with the UI, I hate to post-pone building my UI just because 3.5 is coming, I need to know if I can get started on it now or not.

Let me hear your thougts

Best regards,
Arthur

 

No, it will be a different system. In the same way that OnGUI is a different system than GUIText/GUITexture, but GUIElements weren't removed when OnGUI was implemented.

--Eric

 

I see (sorry, I'm pretty new to Unity). That's annoying. Would it still be worth it to write GUI code now although I would like to switch to the 3.5 GUI at some point?

 

You have to consider that 3.5 might not even ship with GUI stuff. If your app is going to be launched in the next 1-6 months, I'd probably just look for a tested and working GUI solution rather than pin my hopes on 3.5 GUI.

 

That is a very good point. It is a pretty small app, it should definitely be released withing 1-6 months. I will get going with the GUI. Thank you!

 

Given that 3.5 is supposed to be out within the next 15 days or so, that 1-6 month estimate is a bit irrelevant. I'm just now reaching a point where I have to make a decision on the GUI in my current project, and given the current timeframe, I am genuinely just waiting now to see whether or not this thing is going to happen.

 

1-6 month estimate is definitely not irrelevant as most publicly available information suggests that GUI code will not make it into a 3.5 release. I could certainly be wrong (and hope I am!), but I wouldn't hold out for 3.5 if GUI is what I needed.

 

The only information I've seen recently was from Unite where they mentioned that 3.5 would include the new designer-centric GUI system. Everything else has just been vague forum posts and guesses.

Either way, my point is that Unity 3.5 is due "late 2011" and there's been no announcement (that I'm aware of) to say it's been delayed. Therefore, the wait is a maximum of 15 days to see whether or not an alternate GUI system needs to be sought in any given project.

So unless someone's project is launching within the next 15 days, it's definitely worth waiting right now to see which way things are going.

 

Erm no the rest are no vague guesses.
It was not presented at Unite cause it was not ready and different UT developers took part in the thread on the 3.5 gui system where they implied that it might not make it into 3.5, as such its pretty dangerous and risky, if not ignorant, to 'simply assume its there' cause Unity has never said it will be in 3.5. It was on the roadmap for 'beyond 3.4' the last time it was mentioned prior Unite 2011

 

You're all missing my point. I'm not saying it's there or not there. Simply that we don't have very long at all until we find out, so therefore we should definitely wait.

If the only release date given for 3.5 thusfar is "late 2011", then by the end of this year it will either come out, or UT will announce that it's being delayed (hopefully with a new estimate on release window).

If it comes out, we'll then know whether or not the new UI stuff is in.

If it doesn't come out, and the new release date is 6 months off, then we'll also know waiting for 3.5 features is not logical.

 

Is this just an inference or do you know that for a fact?

Will it be a totally different framework or will it be easy to update old code to new. I wish they would tell us so we can plan appropriately.

 

There are no informations known.
Unity neither has talked about it nor did they show it so far so until there is a blog entry or press release on it, we know nothing and all such threads to is derail from this fact and raise questions, problems and confusion, independent of how well they were meant

 

Inference: if the new system was simply an optimization of the existing system, it would have been finished long ago and wouldn't be worth talking up like they have. It's possible it somehow runs on top of OnGUI code, but that seems awkward, and unlikely.

--Eric

 

I really wish they would at least end the speculation, they must know enough by now to chime in here, plaaning is a nightmare at the moment without knowing.

 

planing on unreleased stuff on the other hand is 'beginner style', no business person plans basing on things that are not around, have no timeline and no legally binding guarantee to be delivered. If guys like brady went with this attitude Sprite Manager 2 and EZGUI wouldn't exist, cause Unity 3.0 was meant to get a 2D system.

Its done when its done and until then for all projects started before its released, use what you have at hand. Cause there is no guarantee that it does not take a long time to finally work right (beast and umbra proof that well enough with 20 months counting on that end, lets hope 3.5 finally ends that nightmare)

 

Just make something that has no GUI whatsoever, then you're 100% safe.

--Eric

 

Sadly our plans involved almost 100% Gui. All the back engine 3d stuff is done.

The issue here in terms of planning is whether to tie ourselves in with a 3rd party solution that doesn't do everything we want, stick with current gui and expect to rewrite in 6-12 months time or wait for the release of new gui. If we knew that rewriting from the current gui wasn't going to be too hard that would be my preferred option. The system we are planning is going to be in use for a good few years so getting this decision right now is crucial and expensive to get wrong. Since this infomartion is obviously known by Unity I wish they would just share it and make our lives easier. It's not like it's some massive trade secret that another engine will steal surely.

 

I had a meeting about a month ago with some unity reps that said that the new GUI stuff may not even make it into 3.5 so don't plan on it just yet.

 

The 3.5 announcement topic specifically said that it won't.

--Eric

 

Are there any news? I'm looking for the confirmation that the old GUI code won't break after the new Unity GUI system is released.

As I understood, it's not gonna replace the old GUI system, but rather extend the existing one, or make a parallel new system. True? Also, I assume that the most of this is about having the GUI editor inside the Unity Editor.

Thanks

 

Whatever happened to the 2D system?

I know I'm in the minority, but I'd rather have that than the new GUI system!

 

The 2D system and the new GUI system are one and the same, as far as I know.

--Eric

 

+1, I'd like to have info about this aswell : will the new GUI share the same function names than the actual ?

 

I think that they are waiting for the Scaleform integration.

 

Actually what is the problem with the current GUI system? I am still reading some Unity books but didn't found any issue yet.

 

I found out that there's no problem at all. But you got to have a framework built on the top of it.

The low-level functionality that "immediate" GUI exposes is very well done. All the things that seemed as bugs to me are there by design, thus offering you a greater flexibility. I like all of it, because you don't depend on any complex Unity logic (events etc.), just a lightweight GUI routines, and you can plug in any GUI framework you like.

 

My primary target is mobile user, but I read some post about the GUI kill frame rate on mobile platform, but most tutorials still teaching OnGUI. So what impact exactly with OnGUI on mobile?

 

Any updates on the when the new GUI system might turn up? The feature set seems very promising :

From UT's Technology RoadMap 2011

Can't overemphasize the need for a more optimized, feature rich GUI system. We've tried a rather expensive, highly reviewed GUI library and found it a complete nightmare to use. Currently I have a GUI system based on Prime31's UIToolkit. I estimate atleast using this option until the new UI framework is stable ( 6+ months post release). But maintaining this GUI library ourselves is proving very time consuming.

 

My 2 cents : The problems we have faced relate to multiple draw calls ( one for each GUI element). On mobiles and to a lesser extent on netbooks this degrades performance significantly. There are atleast 3-4 different 3rd Party GUI libraries (you can find them on the Asset Store) which try and solve the problem by reducing the draw calls to one ( or a couple). But each has its on set of issues ranging from portability ( moving UI assets + code from one project to another), setup headaches in the editor, manual placement issues...

Other problems with current GUI system ( OnGUI) is the sheer lack of visual design tools . One has to manipulate position of GUI elements in code. There is no concept of "docking"( anchor this GUI element to left /right/ bottom/top of this other GUI Element) . You have to implement that yourself. It's just ugly .

The hope is that UnityGUI version deux solves these issues elegantly and optimally out of the box.

 

We recently open-sourced our GUIX system and visual layout designer (link in my sig), so you can use that until UT releases their new system. There is also SQUID which looks very cool.

-Jeremy