Release Candidate for Torque3D

#21 Features that would / could get pushed off if we clamped right now:

OSX Advanced Lighting - Still investigating workarounds, possible for a future (post 1.0) release.

OSX PhysX Support - Nvidia doesn't provide an OSX-compatible SDK, so our option would be to license it from them ($50k) and port it for them...yuck. There's always the option, later, of integrating a Mac-compatible solution like ODE. Considerable work has already been done on that anyway, and it's possible we could release that a resource early if people wanted to take a stab at using it, improving it.

Wetness - Pat Wilson actually created a neat PostFX that looks like it will be a much better solution than Gerhard's. This *might* make a 1.0 release anyway, but surely it would be available post 1.0 if not.

That's pretty much it. Not bad right? Most of the stuff we're unsatisfied with for this as a release candidate involves stability, polish, and bugs, not features.

Stuff that you guys haven't seen yet (build imminent):

Particle editor
Tons of GUI editor improvements (thanks Rene Damm!)
Lots and Lots of feature improvements, usability improvements, bug fixes, etc.

#22 @deepscratch:

Quote:then everybody at GG should go on a well deserved vacation for a few days

Thanks! I think we're going to lose a bit of time in December while the studio plans for the move to Las Vegas, and hopefully everyone can squeeze some vacation time in there.

#23

Quote:
In fact, I'd even be okay sharing our dev road map with you guys provided is stayed here, and it wasn't treated immaturely as a promise to deliver down to the last detail.

The phrases "stayed here", "wasn't treated immaturely" and "internet" are not compatible.

You might want to keep the exact details to yourself and just issue something ... vaguer. eg. "we'd kinda like it go like this ... and maybe a little ..."

wow do I type slow...

#24 oh, and fix the forums main page.
please

#25

Quote:I want you guys to be *ecstatic* about Torque 3D when we're ready to release. I want you to feel, genuinely, like it's something worth evangelizing yourselves and I want you to evangelize it yourselves. I want you to believe in the product, where we're going with it and in how incredible this team building it is. I'm amazed watching how fast these guys are moving Torque forward.

I think everyone already is. If you've seen Torque3d, (and everyone here has) you can't help but be ecstatic about it. It's an awesome engine. As for feature clamp as the engine stands right now, that would be fine. I think it has most of what people asked for anyway.

#26 It's just great to hear that there is indeed an end in sight. I know I can't really say at any point when T3D should be feature complete as I don't know what's planned, but as long as there is a checklist keeping track of what's needed and what is wanted for later then it's all good.

I personally don't care much about the OSX support and would much rather have official Linux support, but being a longtime Linux user I've come to expect it to be ignored for anything commercial. Having done some OSX development of my own I can definitely understand how it would slow down development, but assuming it's the OSX platform and not the necessary OpenGL support I think that a Linux build would be even easier then an OSX build (hint..hint). Windows is where the majority gaming marketshare is so as long as that goes smoothly and the OSX support isn't dropped and eventually reaches the same state as the Windows build I can happily wait and then port the OSX build to Linux.

There definitely needs to be more documentation, but so far it looks like it's on track.

#27 @Steve: I'm aware of the incomptibility. I'm just hoping for something a step or two above 4chan :)

#28 @Bryan: I'd be very interested in getting you some official help with the Linux port if you're serious about it. While it's a failure for consumer desktop machines in today's market, having Linux compatibility out of the box does make a lot of game deployment options easier. Dedicated servers for multiplayer obviously being the big win.

#29 The major things that I'm right now missing is the since beta 1 announced rewrite of the advanced lighting related shaders and especially their performance (thats potentially done now as my B1 bug report on the flicker fun seems to be cleared out) and the fixing of the terrain engine + lighting issues (the remove capability would be a welcome thing for 1.0 thought ).


If the other things get added by 1.1 I'm fine :)
As long as T3D doesn't follow TGEA which had to wait till 1.7.1 to be somewhere around its original advertisement, with 1.7.1 taking an eternity, I guess I can life with it :D
T3D so far is aside from TGB the most "release ready" Torque tech I've seen :)


Bryan: Thats the fault of the Linux developers themself I fear. Many decided against proprietary drivers for "crap hack pride" and for "their own dream" they still are incapable to agree on a single standardized linux core with "non standard things" only on top of the whole OS layer (OS layer includes a clear line on using proprietary drivers where available and basic things along that line, things which sadly are normally not granted).
I personally, when I release stuff for Linux, only support Mandriva, as they are the only ones so far with the interest to really make Linux a major end user platform by integrating those things that are just standard if you want endusers and not just Wiki Geeks to use your OS.
That are two major problems if you create anything thats advanced 3D, which on OpenGL already can be trouble some enough, without troublesome drivers.

#30 @Marc: The proprietary source debate is an unfortunate issue with Linux however it is no reason not to develop for it. Ubuntu has become a very big distro and has full support for the proprietary Nvidia driver(ATI unfortunately have never kept their drivers updated enough to be worth it on Linux, though they do work with some effort, the same could be said about ATI on Windows as well), as well as many other distros out there, and Linux as a whole (except in the distros that take special pains not to use any non-GNU code) has fairly good support for at least graphics drivers as the kernel now includes a wrapper system to allow a proprietary binary driver to be plugged into the wrapper which does all the real work with the kernel and works just fine. Regardless of the distro used, as long as something is compiled for the desired system architecture into a standard ELF binary it will function, and there is no need for any commercial developer to support any of the different package managers that these linux distributions use, so other then the nonsense about proprietary code and non-GNU code it's not the fault of any Linux developers.

@Brett: My efforts towards a Linux port depend mostly on how far the OSX build goes since my GLSL shader knowledge is a bit weak and would otherwise need to duplicate all the HLSL shaders to work with OpenGL. I have considerable experience with the X11 API and Linux development in general so I could easily port over any of the system code once I've taken the time to read through and understand all of the enormous T3D codebase. What exactly is involved in "official help?" As I don't have a Mac to test the OSX build on, a complete, current feature comparison or roadmap between the two systems would be an ideal way to decide if I should begin the port now or wait until a later build.

#31 Things I want to see fixed before 1.0 comes out:

1) No light leaks to other rooms/zones, please!
2) Fix Auto-Update. I have to close/open the Editor to see the updated mesh.
3) SSAO improvements.
4) Get the Set Empty working (I know it's coming.)
5) Material Picker improvements, such as the Eye-Drop tool in Photoshop.
(Right now it's difficult to select a texture, out of the many available on ONE mesh.)
6) *.kmz/*.zip files support.
7) Full Lightmapping support.
8) Wetness: Nice to see you guys are working on a better implementation. two thumbs up!

After 1.0: Voxel Terrain system.
Shaders library.
Direct support for nVidia FxComposer *.fx and Render Monkey shaders.

#32 Personally I would rather have a feature clamp after beta 5 with a heavy focus on optimizations. I get better performance in many games with a lot more features turned on than I get with less turned on in T3D.

#33 It is probably a good idea to make sure as many of the features that you have posted being available publicly are in before release even if it drives the timeline on a bit further. Also, Advanced Lighting needs more optimization before it is released. It is targetted for high end cards, but you will get blasted on this if it is released as it is now. Basic Lighting is pretty solid compared to TGEA performance wise, but AL needs a bit of love in this area. More documentation wouldn't hurt either, but a good start has been made in this area already so that is good to see.

The rest of the stuff I have personally ran into are small things that will be picked off through the RC process or Beta5.

#34 @Joshua: We will do more optimization for Advanced Lighting, but it is a high-end feature and we won't be compromising quality for older hardware compatibility there. The Basic Lighting option is getting much, much better though. Almost a shame to call it "Basic." It's essentially an high-quality forward rendering solution. It's better and faster than TGEA's lighting.

#35 B4 basic lighting is a HUGE improvement over previous betas, looks nearly identical to AL with shadows turned off. I've a testmap with 1000 trees and it gives 25fps AL and 100fps BL (1280x800).

#36

Quote:Personally I would rather have a feature clamp after beta 5 with a heavy focus on optimizations. I get better performance in many games with a lot more features turned on than I get with less turned on in T3D.

Ditto!

Are the changes tot he camera system gonna be in B5? Right now this is the single biggest headache that we are having. the code for camera interaction is spread all over the engine. This is something that needs to be way easier to alter as it is pivotal for doing things other than FPS.

I say do B5 and then prolly 2 almost back to back RC releases and then break out the champagne and party hats!

#37

Quote:Almost a shame to call it "Basic."

Yes, but for the term basic I fear.
Barebone would be more appropriate as no shadow at all in a 2009 technology is by no meaning basic anymore, as it is "lighting capability wise" en par with DirectX7 at best.
Don't see which part of it makes it "better than TGEA" for you under the current conditions other than that it potentially might run better (thats something I would need to test first in a fair and realistic comparision, don't think there was a head to head comparision available to self test on this claim so far, its just a subjective feeling of many)


If we are fair, then we have to realize that even TGE has shadows and that roots back to a soon a decade old technology.

BL currently is a no-end hardware fallback that works on a large amount of systems, which surely is a good thing.
But the problem is it leaves a pretty large gap between the trash systems and the high end systems that are either forced to play on low performance or on drastically reduced atmosphere, a lose - lose situation.

In the current conditions BL offered, one would expect that some serious work went into raising the capabilities and quality of the static lighting model, so BL games don't lose all their atmosphere.

#38 @Marc: I'm cheating a little bit because there's more coming with Basic that we haven't shown yet. There are shadows, and lots of lights, and it's very fast.

#39 Ok

Thank you for that clarification


I didn't know that.
I asked once about the BL and its plans a longer while ago, but as far as I remember, there wasn't an answer on it, at least none that would have clearly indicated that BL would go that far within the forseable (2009) future or how far it is meant to go at all.

#40

Quote:shadows, and lots of lights, and it's very fast

Onoes!