Experience counts when Comparing Engines
by Ron Kapaun · 11/28/2016 (4:49 pm) · 18 comments
First off, a note to everyone! This is not a 'T3D is Terrible' thread! I know there are not that many T3D users that stop in here anymore. I also understand that the newest version of T3D uses DX11. This is all great. In fact, I did develop a quick test level in the latest DX 11 version of T3D and it did outperform all the past versions (at least on the stats I can track in engine).
This is NOT a 'I hate T3D' thread because well, T3D does a HUGE number of things right! Many of which I TOOK FOR GRANTED! Congrats to the Steering Committee! Very nice work.
So, let's talk about this touchy topic, T3D v. UnReal 4 (an ACTUAL comparison from someone who has produced content for both!)
As you all likely know, my last few blogs here were about photogrammetry and using those techniques to produce very accurate 3D models within a game engine environment. T3D started off very promising and I dropped of the face the interwebs in order to conduct some MAJOR research. Well, that was way back in late 2014. Over the last year+ I devoted my time to learning Blender (like a pro), photogrammety (again... like a pro) and shifting from T3D to UnReal 4... (bet you can guess how I thought of this approach).
I figured it would take me may be, 2 months to do all of this. Wow, was I ever WRONG. Blender took forever to figure out but, I am pretty good at it now. Photogrammetry... well it's STILL a work in progress. I learn new techniques and new ways to shoot things with every object I decide to capture. Then there was UnReal. (latest results below :-);
.
.
Now, I figured that I understood the basics of game development. When it comes to T3D, I was confident...heck I was actually beyond cocky about making stuff for Torque. Then I decided to switch over. WOW, was I ever OVER estimating my skills! UnReal 4, sent my ass BACK to school. UnReal 4 appears simple and powerful on the surface and it is until you dig a bit. There are some things that T3D does WAY better and far easier than UnReal. Let's just list a few;
1. Ever try creating terrain textures in UnReal? OMG. In T3D, you add a few textures, get fancy and create a displacement map and you can paint a terrain texture for days! In UnReal 4.... well NOT exactly simple. (See the screen shot);
.
.
See UnReal 4 uses something called blueprints (think of these as 'visual scripting'.) They start off REALLY simple and then you get to a point where you just want to paint grass and rocks and crap on a landscape. This is a sample of a very 'basic' set of terrain materials that I made. It is a grand total of 4 freaking terrain materials.
2. Let's talk water... OMG water in T3D... add a water plane or river object and play with the settings and BAM... you can make a pretty decent game version of water... UnReal 4... DAMN.... This is just a simple water SURFACE (Yes, you MUST create a separate UNDERWATER material using something called a Post Process Volume... good lord.);
.
.
Looks freakin impressive doesn't it? Not really, in fact, I think it looks about the same as a stock T3D water plane. You know, the one where all you do it click it and adjust it's Z-axis to set the level. Well, this is a BASE water material. Here is a quick Screenshot from in-game;
.
.
NOTE: NO TRANSPARENCY! I have to add that yet... also, this does NOT shade 'underwater'... Like I said earlier, that is a separate material attached to a separate 'gizmo' that has to be added to your scene and then scaled and adjusted to 'match' your surface. It is a real pain, trust me.
There are a number of things that UnReal 4 does well, PBR (Physically based Rendering) of textures and screen space reflections, etc. are cool. It's lighting system is pretty much unparalleled at this point in time. In fact, you could spend WEEKS just learning all the settings that can be tweeked in that system alone.
Alas, I miss the days of T3D. I could produce a scene, object or lighting system without laying out hundreds of little squares and linking them all up. If I wanted a basic AI, hell in T3D I could have a base AI that would hunt you and shoot you with just a bit of script. UnReal 4... hell most of the TUTORIALS for doing that are hours and hours long.
I could go on and on and compare the pros and cons for both engines but, I think these simple examples show that T3D may be just as, if not more powerful than UnReal 4. Depending on what you want to do. I am happy with all the work I put in. In the end, the ONLY real positive I have found for swapping to UnReal is the lighting model, the ability to use MANY more Anti-Aliasing solutions and it's Global lighting and light bounce solutions. (See Interior Render below):
.
.
All in all, given the time and effort, I think I could pull off nearly the same level of work in T3D as I have in UnReal 4. All of the latest improvements in the T3D pipeline do no less than bring it even closer to a 'modern generation' game engine. I will continue to watch the project.
However, this will be my final blog concerning T3D. I will follow the engine until the very end but, I will not likely comment or add anything new. The T3D engine and all of you have been great to me over the years. I sincerely wish you all the best and I want to personally say thank you to each and every one of you that has helped me, worked with me, encouraged me and criticized me over these years. Good Luck, God (or your choice of higher power) Bless!
Signing out,
Ron
OH YEAH... P.S. My web site no longer has all the free T3D content but you can find it at CGTrader.com, just search under 3tdstudiosron or click this link :)
www.cgtrader.com/3tdstudiosron
I will miss you all!
This is NOT a 'I hate T3D' thread because well, T3D does a HUGE number of things right! Many of which I TOOK FOR GRANTED! Congrats to the Steering Committee! Very nice work.
So, let's talk about this touchy topic, T3D v. UnReal 4 (an ACTUAL comparison from someone who has produced content for both!)
As you all likely know, my last few blogs here were about photogrammetry and using those techniques to produce very accurate 3D models within a game engine environment. T3D started off very promising and I dropped of the face the interwebs in order to conduct some MAJOR research. Well, that was way back in late 2014. Over the last year+ I devoted my time to learning Blender (like a pro), photogrammety (again... like a pro) and shifting from T3D to UnReal 4... (bet you can guess how I thought of this approach).
I figured it would take me may be, 2 months to do all of this. Wow, was I ever WRONG. Blender took forever to figure out but, I am pretty good at it now. Photogrammetry... well it's STILL a work in progress. I learn new techniques and new ways to shoot things with every object I decide to capture. Then there was UnReal. (latest results below :-);
.
. Now, I figured that I understood the basics of game development. When it comes to T3D, I was confident...heck I was actually beyond cocky about making stuff for Torque. Then I decided to switch over. WOW, was I ever OVER estimating my skills! UnReal 4, sent my ass BACK to school. UnReal 4 appears simple and powerful on the surface and it is until you dig a bit. There are some things that T3D does WAY better and far easier than UnReal. Let's just list a few;
1. Ever try creating terrain textures in UnReal? OMG. In T3D, you add a few textures, get fancy and create a displacement map and you can paint a terrain texture for days! In UnReal 4.... well NOT exactly simple. (See the screen shot);
.
.
See UnReal 4 uses something called blueprints (think of these as 'visual scripting'.) They start off REALLY simple and then you get to a point where you just want to paint grass and rocks and crap on a landscape. This is a sample of a very 'basic' set of terrain materials that I made. It is a grand total of 4 freaking terrain materials.
2. Let's talk water... OMG water in T3D... add a water plane or river object and play with the settings and BAM... you can make a pretty decent game version of water... UnReal 4... DAMN.... This is just a simple water SURFACE (Yes, you MUST create a separate UNDERWATER material using something called a Post Process Volume... good lord.);
.
.Looks freakin impressive doesn't it? Not really, in fact, I think it looks about the same as a stock T3D water plane. You know, the one where all you do it click it and adjust it's Z-axis to set the level. Well, this is a BASE water material. Here is a quick Screenshot from in-game;
.
.
NOTE: NO TRANSPARENCY! I have to add that yet... also, this does NOT shade 'underwater'... Like I said earlier, that is a separate material attached to a separate 'gizmo' that has to be added to your scene and then scaled and adjusted to 'match' your surface. It is a real pain, trust me.
There are a number of things that UnReal 4 does well, PBR (Physically based Rendering) of textures and screen space reflections, etc. are cool. It's lighting system is pretty much unparalleled at this point in time. In fact, you could spend WEEKS just learning all the settings that can be tweeked in that system alone.
Alas, I miss the days of T3D. I could produce a scene, object or lighting system without laying out hundreds of little squares and linking them all up. If I wanted a basic AI, hell in T3D I could have a base AI that would hunt you and shoot you with just a bit of script. UnReal 4... hell most of the TUTORIALS for doing that are hours and hours long.
I could go on and on and compare the pros and cons for both engines but, I think these simple examples show that T3D may be just as, if not more powerful than UnReal 4. Depending on what you want to do. I am happy with all the work I put in. In the end, the ONLY real positive I have found for swapping to UnReal is the lighting model, the ability to use MANY more Anti-Aliasing solutions and it's Global lighting and light bounce solutions. (See Interior Render below):
.
.
All in all, given the time and effort, I think I could pull off nearly the same level of work in T3D as I have in UnReal 4. All of the latest improvements in the T3D pipeline do no less than bring it even closer to a 'modern generation' game engine. I will continue to watch the project.
However, this will be my final blog concerning T3D. I will follow the engine until the very end but, I will not likely comment or add anything new. The T3D engine and all of you have been great to me over the years. I sincerely wish you all the best and I want to personally say thank you to each and every one of you that has helped me, worked with me, encouraged me and criticized me over these years. Good Luck, God (or your choice of higher power) Bless!
Signing out,
Ron
OH YEAH... P.S. My web site no longer has all the free T3D content but you can find it at CGTrader.com, just search under 3tdstudiosron or click this link :)
www.cgtrader.com/3tdstudiosron
I will miss you all!
About the author
Guy that has been with Torque Game Engines since... well, since here was a Torque Game Engine. (V12 and beyond).
#2
Yeah, I know. I may post over there but, I figured I owed one last post here. Nostalgia... Glad to see you are still active. Hope all is well.
Ron
11/28/2016 (6:19 pm)
Hey Steve, Yeah, I know. I may post over there but, I figured I owed one last post here. Nostalgia... Glad to see you are still active. Hope all is well.
Ron
#3
11/29/2016 (2:25 am)
Good luck on your new endeavor, those shots looks great btw.
#4
11/29/2016 (4:31 pm)
Good to hear from you again, Ron, and thank you again for all the contributions you'd made to T3D over the years!
#5
And yeah, UE4 does do some stuff that's really nice(namely the rendering side) but it seems pretty consistent that you have to do a lot of work to set up the GAME side(or, you know, drop cash on the store to get hopefully-this-is-good premade assets).
Us guys in the build/steering group definitely keep an eye on the other big engines like UE4, Unity and Cryengine for things they do well, and things people say they don't as we try to plot out what's actually NEEDED, rather than shiney bullet points that no one will actually use in the end.
So yeah, by all means, if you've got details about stuff you think worked or didn't in relation to your experience with UE4 please, please don't hesitate to drop those bits. Either here, or heck feel free to thrown an email my way.
Can't make things better if we don't have data, after all. ;)
Speaking of pipeline, for T3D 4.0, we're moving over to using PBR as well(it's most of the way implemented in R&D branches, mostly just sorting out oddball outlier cases and the like), I'm curious if you had any particular opinions on PBR workflows at this point in time?
We'd been largely using the Allegorithmic Substance tools because they're extensive and cheap, but there's a lot of other tools out there like Quixel, and I hadn't tried doing PBR photogrammetry yet. So I'm just curious how you felt about the pipeline and if there was stuff you'd like seem improved we could take notes on for T3D's PBR implement.
11/30/2016 (7:01 am)
Hey man, good to hear you're still doing well and trucking along!And yeah, UE4 does do some stuff that's really nice(namely the rendering side) but it seems pretty consistent that you have to do a lot of work to set up the GAME side(or, you know, drop cash on the store to get hopefully-this-is-good premade assets).
Us guys in the build/steering group definitely keep an eye on the other big engines like UE4, Unity and Cryengine for things they do well, and things people say they don't as we try to plot out what's actually NEEDED, rather than shiney bullet points that no one will actually use in the end.
So yeah, by all means, if you've got details about stuff you think worked or didn't in relation to your experience with UE4 please, please don't hesitate to drop those bits. Either here, or heck feel free to thrown an email my way.
Can't make things better if we don't have data, after all. ;)
Speaking of pipeline, for T3D 4.0, we're moving over to using PBR as well(it's most of the way implemented in R&D branches, mostly just sorting out oddball outlier cases and the like), I'm curious if you had any particular opinions on PBR workflows at this point in time?
We'd been largely using the Allegorithmic Substance tools because they're extensive and cheap, but there's a lot of other tools out there like Quixel, and I hadn't tried doing PBR photogrammetry yet. So I'm just curious how you felt about the pipeline and if there was stuff you'd like seem improved we could take notes on for T3D's PBR implement.
#6
Thanks for the info and the RFI (request for information). Yes, there are a large number of things in UR4 that need some simplification from the art side. (I am working on the blueprint side as well but, honestly I do no know enough there to make any suggestions vesus cold scripting (really its the same either way)) As for the art, all the things I stated in the blog apply. As an additional note; I honestly LOVE the T3D camera work. Keep all of that as it is or even improve it. I just today released a new 'UnReal ready' video on YouTube (it's UnReal specific but here is the addy for anyone interested: www.youtube.com/watch?v=c1WmYqD1fwo and I have to say, it was an GINORMOUS pain in the butt. I really TRUELY missed T3D's Smooth Camera and the intuitive way T3D allowed me to 'shoot' video and drop it into the editor of my choice. In contrast, I had to spend almost 3 hours creating a 'camera rig' in UnReal and (granted) it simulated things like 'real world cranes and dolly tracks' but, it lacked T3D's 'natural' feel. So, keep those features and ONLY improve on them.
As for the PBR stuff, Let me know when it is ready for testing. So much of that crap is based on 'set' properties and such. Also, if you use anything like CrazyBump or something like that you have to add a ton of crap to the UnReal blueprint in order to make it work in game. I get that however, ages ago I made a blog concerning 'marble, brass' and some other materials and they LOOKED correct in T3D (minus accurate reflections.) Although, I could have created a quick 'dirty' cube map in T3D to make it seem correct.... anyway. Let me know when you are ready for an 'art' test and I will help. Also, ya all need a 'real' demo level or two.... Just sayin. It's been YEARS since something NOT programmer 'art' has been put out..... I might be able to help there as well.
Oh Yeah... If you watch my video, you will see something that I tried and tried to do in T3D. The subsurface leaf work. Note that in the video, you see the 'veins' in the leaves and stuff. This was a 'simple' task in UnReal, I just added a multiplied mask to the subsurface color main track. (if that makes sense.) Just adding the ability to properly mask textures in T3D would be a HUGE help.
There is far more of course. I may just send an Artist 'base' email to you since this is probably not the best forum.
Ron
12/01/2016 (5:22 pm)
Hey Jeff,Thanks for the info and the RFI (request for information). Yes, there are a large number of things in UR4 that need some simplification from the art side. (I am working on the blueprint side as well but, honestly I do no know enough there to make any suggestions vesus cold scripting (really its the same either way)) As for the art, all the things I stated in the blog apply. As an additional note; I honestly LOVE the T3D camera work. Keep all of that as it is or even improve it. I just today released a new 'UnReal ready' video on YouTube (it's UnReal specific but here is the addy for anyone interested: www.youtube.com/watch?v=c1WmYqD1fwo and I have to say, it was an GINORMOUS pain in the butt. I really TRUELY missed T3D's Smooth Camera and the intuitive way T3D allowed me to 'shoot' video and drop it into the editor of my choice. In contrast, I had to spend almost 3 hours creating a 'camera rig' in UnReal and (granted) it simulated things like 'real world cranes and dolly tracks' but, it lacked T3D's 'natural' feel. So, keep those features and ONLY improve on them.
As for the PBR stuff, Let me know when it is ready for testing. So much of that crap is based on 'set' properties and such. Also, if you use anything like CrazyBump or something like that you have to add a ton of crap to the UnReal blueprint in order to make it work in game. I get that however, ages ago I made a blog concerning 'marble, brass' and some other materials and they LOOKED correct in T3D (minus accurate reflections.) Although, I could have created a quick 'dirty' cube map in T3D to make it seem correct.... anyway. Let me know when you are ready for an 'art' test and I will help. Also, ya all need a 'real' demo level or two.... Just sayin. It's been YEARS since something NOT programmer 'art' has been put out..... I might be able to help there as well.
Oh Yeah... If you watch my video, you will see something that I tried and tried to do in T3D. The subsurface leaf work. Note that in the video, you see the 'veins' in the leaves and stuff. This was a 'simple' task in UnReal, I just added a multiplied mask to the subsurface color main track. (if that makes sense.) Just adding the ability to properly mask textures in T3D would be a HUGE help.
There is far more of course. I may just send an Artist 'base' email to you since this is probably not the best forum.
Ron
#7
Oh, yeah. Feel free to throw me an email. Areloch at gmail dotcom I'm all up for talking shop and ideas :)
And yeah, I'll probably take you up on your offer for poking and prodding on PBR when she's ready. It's like, 90% there, currently, mostly just gotta wrap up reflection probes and then nail down the weird, outlier situations that crop up with oddball values or different texture formats. Hadn't tried using Crazybump, I remember it from a while back though, seemed like a good tool.
In regards to the camera stuff, yeah definitely. The plan is, even with the shift to using entity/component stuffs, to have all the gameplay class functionality ported over so there's the huge swath of starting stuff to get someone rolling, camera included.
I'd planned to port all the modes for the camera as a matter of course, but until you highlighted how much it matters, I hadn't really considered the impact of how important from a user perspective just having something like a built in smooth move mode could really be. Sure, it's not hard to add, but the entire idea is having it in there already to avoid needing to, and just letting people get on with what they want to do.
Definitely want to allow more functionality, not less, for cameras though. I'd already been experimenting with making it easy to do stuff like the classic RE fixed camera perspectives by going into triggers and the like, so the hope is to make the camera stuff powerful enough that there's not really any question of 'how do I get the camera to do X'.
For the subsurface materials stuff, yeah, one of my personal itches is properly opening up the material/shader system. The Shadergen/material editor we have now is actually really nice for getting a huge range of materials made really fast. But if you have to add a new feature, or do custom mat/shaders, it becomes a LOT more work real quick.
Opening all that up will go a LONG way to letting art peoples like yourself just get on with making the art, rather than needing to fuss around with the backend.
Anywho, yeah, my all means shoot me a mail if you want. I'm totally game for discussion and feedback.
12/04/2016 (2:46 pm)
@RonOh, yeah. Feel free to throw me an email. Areloch at gmail dotcom I'm all up for talking shop and ideas :)
And yeah, I'll probably take you up on your offer for poking and prodding on PBR when she's ready. It's like, 90% there, currently, mostly just gotta wrap up reflection probes and then nail down the weird, outlier situations that crop up with oddball values or different texture formats. Hadn't tried using Crazybump, I remember it from a while back though, seemed like a good tool.
In regards to the camera stuff, yeah definitely. The plan is, even with the shift to using entity/component stuffs, to have all the gameplay class functionality ported over so there's the huge swath of starting stuff to get someone rolling, camera included.
I'd planned to port all the modes for the camera as a matter of course, but until you highlighted how much it matters, I hadn't really considered the impact of how important from a user perspective just having something like a built in smooth move mode could really be. Sure, it's not hard to add, but the entire idea is having it in there already to avoid needing to, and just letting people get on with what they want to do.
Definitely want to allow more functionality, not less, for cameras though. I'd already been experimenting with making it easy to do stuff like the classic RE fixed camera perspectives by going into triggers and the like, so the hope is to make the camera stuff powerful enough that there's not really any question of 'how do I get the camera to do X'.
For the subsurface materials stuff, yeah, one of my personal itches is properly opening up the material/shader system. The Shadergen/material editor we have now is actually really nice for getting a huge range of materials made really fast. But if you have to add a new feature, or do custom mat/shaders, it becomes a LOT more work real quick.
Opening all that up will go a LONG way to letting art peoples like yourself just get on with making the art, rather than needing to fuss around with the backend.
Anywho, yeah, my all means shoot me a mail if you want. I'm totally game for discussion and feedback.
#8
First off nice write up. I will have to disagree though, T3D is dead unless it comes out with mobile dev for Android, IOS and Xbox One. Sure, the engine is good for Windows, I agree, but today's gaming is moving to varying platforms, not just windows. Torque can not keep up with that, and I feel sad bc it can't. I have been hoping to see a mobile compiler for android, and yeah, one once was promised, but I never seen it come to light. Torque NEEDS Android, IOS and Xbox exporters to survive against Unity and Unreal Engine 4. Plus, like Torque 3D, Unreal just made their engine 4 FREE. Also, we have Torque X CE which was nice for the time, but $99 for something that is dead for Xbox 360? XB360 is dead too, everything has now moved to the new One instead.
I'm still using Torque for small projects since I've switched to Unreal Engine 4. Yeah the learning curve is rough but once you grab it, your good. I honestly think Unreal has come a long way to where they are. Torque is moving on with people keeping it alive, which is great, but IMHO it needs an Android, IOS and XB One exporter to survive. I would HOPE that is in the plans down the road for Torque, I would like to see it continue to grow and become something to keep free for us Indie gamer designers.
I've been around since about version 1.4 of Torque Game Engine, bought into the $505 fee for the Torque 3D and then when they made it free, my heart broke cause I felt ripped off since I put that money into the engine to get a pro engine that ended up going free anyways. If I knew GG was going to ever do that in the beginning, I would have put that money into Unity at the time or another engine where I knew they at least had kept the PRO version closed to just anyone. Oh and too, we now have Game Studio even being able to export to Android, IOS and Windows too, so I know if all these engines can export to other formats, I'm sure Torque can get it put in there to keep it's head above the water level.
And before anyone goes off the deep end saying I'm bashing Torque, I'm not. I've been a long time user and will continue to do so. Just for tiny projects, unless they come out with Android and XB One kits. I mainly use Torque in the haunt industry making visual images, moody scenes and things like that to help people take away from the real world and relax with our Ambient Sounds and such.
I hope to see Torque grow to add these features in coming releases. I think it would prove successful for the engines life span and to keep it in check with the other guys who already provide these great features.
Thanks for reading!
12/07/2016 (12:09 am)
Hi Ron,First off nice write up. I will have to disagree though, T3D is dead unless it comes out with mobile dev for Android, IOS and Xbox One. Sure, the engine is good for Windows, I agree, but today's gaming is moving to varying platforms, not just windows. Torque can not keep up with that, and I feel sad bc it can't. I have been hoping to see a mobile compiler for android, and yeah, one once was promised, but I never seen it come to light. Torque NEEDS Android, IOS and Xbox exporters to survive against Unity and Unreal Engine 4. Plus, like Torque 3D, Unreal just made their engine 4 FREE. Also, we have Torque X CE which was nice for the time, but $99 for something that is dead for Xbox 360? XB360 is dead too, everything has now moved to the new One instead.
I'm still using Torque for small projects since I've switched to Unreal Engine 4. Yeah the learning curve is rough but once you grab it, your good. I honestly think Unreal has come a long way to where they are. Torque is moving on with people keeping it alive, which is great, but IMHO it needs an Android, IOS and XB One exporter to survive. I would HOPE that is in the plans down the road for Torque, I would like to see it continue to grow and become something to keep free for us Indie gamer designers.
I've been around since about version 1.4 of Torque Game Engine, bought into the $505 fee for the Torque 3D and then when they made it free, my heart broke cause I felt ripped off since I put that money into the engine to get a pro engine that ended up going free anyways. If I knew GG was going to ever do that in the beginning, I would have put that money into Unity at the time or another engine where I knew they at least had kept the PRO version closed to just anyone. Oh and too, we now have Game Studio even being able to export to Android, IOS and Windows too, so I know if all these engines can export to other formats, I'm sure Torque can get it put in there to keep it's head above the water level.
And before anyone goes off the deep end saying I'm bashing Torque, I'm not. I've been a long time user and will continue to do so. Just for tiny projects, unless they come out with Android and XB One kits. I mainly use Torque in the haunt industry making visual images, moody scenes and things like that to help people take away from the real world and relax with our Ambient Sounds and such.
I hope to see Torque grow to add these features in coming releases. I think it would prove successful for the engines life span and to keep it in check with the other guys who already provide these great features.
Thanks for reading!
#9
12/07/2016 (10:07 am)
First, nice to see you are still around. What you describe is exactly what I am experiencing, Unreal has a pretty steep learning curve and the only reason I am still using T3D is that my game The Master's Eye is near preAlpha release, so it takes a hell of a lot time to bring it over to Unreal, but my new projects are all in Unreal. The biggest issue in my humble opinion, is that with T3D I loose a lot of time developing the engine and not my game (I think this is what you have when using an open-source engine, allthough I date back from the time it was still around $1000). You learn a lot (at least I did), but it is somewhat counter-productive when you have to earn a living with gamedev. Nevertheless it was most of the time enjoyable, and the community is of good will when having problems or issues.
#10
Why do you feel that mobile support is a make or break feature? Sure, the mobile dev space is gigantic, which is always a good idea to attempt to capitalize on, but do you feel that doing one set of things well - which is currently where we're pushing T3D, to be a very solid platform on desktops, isn't as viable as doing kinda-ok in a whole bunch of them?
Just kinda curious on how you feel about division of effort, there. Because like said, while mobile is a huge space, it's also really volatile, and supporting everything for it is hard, even for major studios. So it strikes me as a 'do one thing well or a dozen things meh' sort of situation.
That said, yes, I would like to see console support happen, especially with the fact that as the console generations advance, getting SDKs and porting is easier and easier. It also helps that they're almost just specialized PCs at this point, haha.
So yeah, fully agree it'd be nice to see console support get in there in the future (mostly just a matter of getting ahold of dev units so the work can be done) but I'm not as sold that mobile is a must-have space currently.
Any particular reasons you feel that'd be the case? I'm really interested to hear your thoughts
@Richard
Yeah, a big push with the design on the 4.0 front for T3D is cutting back how much time you have to spend in the engine itself, and instead just spending time in the script and tools to get stuff done. Fully agree with you on that front.
Source should be a bonus and let you do well optimized or weird stuff. Probably shouldn't be the mainstay modus of development.
Are there any spots in particular that gave you what you felt an undue amount of work to deal with?
Issues like this - workflow, where effort is applied when building a game - are really high priority for 4.0, so I'd really like to hear your feedback on that.
If you hadn't already, I'd suggest taking a look through my workblog on the new forums about the work being done. A lot of it is tools/pipeline improvements, so as said, if you've got any particular input on that sort of thing, I'd really like to hear it.
12/07/2016 (11:18 am)
@WillWhy do you feel that mobile support is a make or break feature? Sure, the mobile dev space is gigantic, which is always a good idea to attempt to capitalize on, but do you feel that doing one set of things well - which is currently where we're pushing T3D, to be a very solid platform on desktops, isn't as viable as doing kinda-ok in a whole bunch of them?
Just kinda curious on how you feel about division of effort, there. Because like said, while mobile is a huge space, it's also really volatile, and supporting everything for it is hard, even for major studios. So it strikes me as a 'do one thing well or a dozen things meh' sort of situation.
That said, yes, I would like to see console support happen, especially with the fact that as the console generations advance, getting SDKs and porting is easier and easier. It also helps that they're almost just specialized PCs at this point, haha.
So yeah, fully agree it'd be nice to see console support get in there in the future (mostly just a matter of getting ahold of dev units so the work can be done) but I'm not as sold that mobile is a must-have space currently.
Any particular reasons you feel that'd be the case? I'm really interested to hear your thoughts
@Richard
Yeah, a big push with the design on the 4.0 front for T3D is cutting back how much time you have to spend in the engine itself, and instead just spending time in the script and tools to get stuff done. Fully agree with you on that front.
Source should be a bonus and let you do well optimized or weird stuff. Probably shouldn't be the mainstay modus of development.
Are there any spots in particular that gave you what you felt an undue amount of work to deal with?
Issues like this - workflow, where effort is applied when building a game - are really high priority for 4.0, so I'd really like to hear your feedback on that.
If you hadn't already, I'd suggest taking a look through my workblog on the new forums about the work being done. A lot of it is tools/pipeline improvements, so as said, if you've got any particular input on that sort of thing, I'd really like to hear it.
#11
I will see what I can do. When is 4.0 scheduled, because at the moment I am really busy getting the preAlpha done (it is scheduled december 16)?
In any case I believe, that before trying to get all those fancy consoles and mobiles integrated, you must have an engine that is 99.9% bug free and stable (this is what we in Dutch language call: kick in an open door).
Also take a look at the options other "big" engines have. For example, Unreal has a cable object, which simulates a rope attached to one point, or between 2 points, including physics and particles mount. This is one thing one my list I was planning to do in T3D.
Another issue is the use of compiled shaders in T3D. You can precompile shaders in VC (not the procedural shaders), but then you have to modify T3D and the shaders to use them (shadermacros and defines in TS does not seem to work). Then I want T3D to save and use the compiled shaders (both the common and the procedural), but all I get is a purple screen (I do not know if this problem is on my side or in the engine yet).
My biggest problem is something I created myself. I started developing The Master's Eye in T3D release 1.0! including modification and adding to the source. When 1.2 came out I integrated all my stuff in that build, but from then on I did some WinMerge until 3.0 and now a WinMerge to 3.9 . You can imagine that it is sometimes difficult to distinct if something was caused by the older source or by the additions, allthough I must say that the build I am running now is more or less 3.9 with my stuff integrated.
Anyway, if something more comes into mind I will drop it to you.
Edit:
I almost forgot: the implementation of DirectX11 was a huge step forward and I say a big thank you to all the people who made that possible.
12/07/2016 (12:02 pm)
@JeffI will see what I can do. When is 4.0 scheduled, because at the moment I am really busy getting the preAlpha done (it is scheduled december 16)?
In any case I believe, that before trying to get all those fancy consoles and mobiles integrated, you must have an engine that is 99.9% bug free and stable (this is what we in Dutch language call: kick in an open door).
Also take a look at the options other "big" engines have. For example, Unreal has a cable object, which simulates a rope attached to one point, or between 2 points, including physics and particles mount. This is one thing one my list I was planning to do in T3D.
Another issue is the use of compiled shaders in T3D. You can precompile shaders in VC (not the procedural shaders), but then you have to modify T3D and the shaders to use them (shadermacros and defines in TS does not seem to work). Then I want T3D to save and use the compiled shaders (both the common and the procedural), but all I get is a purple screen (I do not know if this problem is on my side or in the engine yet).
My biggest problem is something I created myself. I started developing The Master's Eye in T3D release 1.0! including modification and adding to the source. When 1.2 came out I integrated all my stuff in that build, but from then on I did some WinMerge until 3.0 and now a WinMerge to 3.9 . You can imagine that it is sometimes difficult to distinct if something was caused by the older source or by the additions, allthough I must say that the build I am running now is more or less 3.9 with my stuff integrated.
Anyway, if something more comes into mind I will drop it to you.
Edit:
I almost forgot: the implementation of DirectX11 was a huge step forward and I say a big thank you to all the people who made that possible.
#12
Yeah, by all means, feedback is really important. And alas, 4.0's not scheduled for Dec 16th ;)
3.10 should be wrapped in a few days when we PR a vsync bugfix and a fix for the collada export, then an RC period, then full steam on 4.0. Lots of the stuff for it is done or pretty far along for it already, but it'd still be several months before it'd really be ready.
But hey, if we get 'er locked in by the time you do beta or release, I wouldn't mind helping you port over if you wanted :)
Anywho, honestly, while there's definitely a good number of bugs yet to squash, it's always struck me that T3D's been actually rather stable. Maybe we in the devgroup haven't done the "weird stuff", but the engine's seemed stable, we just gotta crush those bugs.
I feel that the move to the entity/component stuff should help with that quite a bit. Because all the gameplay classes(player, vehicle, etc) are going to be ported to it so we have the same default object offering, but componentized, means that everything's gunna get some eyeballing time, which is a really good time to fix a lot of those leftover floating bugs.
So yeah, I think that'll help a good deal there.
For the shaders stuff, I totally get you. Compiled shader discussion's come up a good bit in the devgroup. The current plan is a refactor of the shadergen to be much more open and flexible, which should help a good bit in general.
But Timmy's brought attention on precompiling the shaders and stuff(I think he's messed with it before already) so I'll keep that in mind as well.
And yeah, I completely get you on the porting/upgrading. I mean, it's a fairly well know thing for ANY engine that 'At some point you have to stick with a version and just make your game', but that does NOT mean it should be HARD to upgrade.
With the push to script or components for gameplay stuff, and a lot less 'smorgasboard of dozens of files' for anything from players to static shapes, I think it should be a lot easier to keep up with updates of the engine.
Do you feel it was a 'The changes are so expansive it's hard to deal with', or was it more just a 'I'm not sure how to do this' - like no documentation on suggestions for how to update your codebase? 'Cause the stuff above should help with the former, but if it's more of a knowledge/documentation issue, that would need specific handling.
But the main point being that, the less time you have to spend in the engine means less stuff that may run aground on updates. My main goal on that front - which I've been testing a lot in my R&D build - is that outside of rendering or custom physics - frame intensive stuff -, everything else should be possible to prototype purely in script and the editors.
That'd go a long way to avoiding 'Update-itosis' from setting in. If all your stuff is in script during the prototyping/design phase, and will only go into the engine for the optimization/finalization phase, it means that updates on the engine side should rarely, if ever, collide with your work.
Heck, in those cases, you should be able to just drop in the new precompiled binary and maybe update a few things script-side and carry on your way.
If you're curious, I can toss you a third person 'screw around' build I've been testing snippets and stuff in to see what I mean, where all the actual gameplay stuff, from player controls and simple physics rules(not like, Bullet/PhysX physics, but player movement), projectile simulation, interaction, etc are pretty much all done with a stock of default engine-side components and script.
12/07/2016 (12:19 pm)
@RichardYeah, by all means, feedback is really important. And alas, 4.0's not scheduled for Dec 16th ;)
3.10 should be wrapped in a few days when we PR a vsync bugfix and a fix for the collada export, then an RC period, then full steam on 4.0. Lots of the stuff for it is done or pretty far along for it already, but it'd still be several months before it'd really be ready.
But hey, if we get 'er locked in by the time you do beta or release, I wouldn't mind helping you port over if you wanted :)
Anywho, honestly, while there's definitely a good number of bugs yet to squash, it's always struck me that T3D's been actually rather stable. Maybe we in the devgroup haven't done the "weird stuff", but the engine's seemed stable, we just gotta crush those bugs.
I feel that the move to the entity/component stuff should help with that quite a bit. Because all the gameplay classes(player, vehicle, etc) are going to be ported to it so we have the same default object offering, but componentized, means that everything's gunna get some eyeballing time, which is a really good time to fix a lot of those leftover floating bugs.
So yeah, I think that'll help a good deal there.
For the shaders stuff, I totally get you. Compiled shader discussion's come up a good bit in the devgroup. The current plan is a refactor of the shadergen to be much more open and flexible, which should help a good bit in general.
But Timmy's brought attention on precompiling the shaders and stuff(I think he's messed with it before already) so I'll keep that in mind as well.
And yeah, I completely get you on the porting/upgrading. I mean, it's a fairly well know thing for ANY engine that 'At some point you have to stick with a version and just make your game', but that does NOT mean it should be HARD to upgrade.
With the push to script or components for gameplay stuff, and a lot less 'smorgasboard of dozens of files' for anything from players to static shapes, I think it should be a lot easier to keep up with updates of the engine.
Do you feel it was a 'The changes are so expansive it's hard to deal with', or was it more just a 'I'm not sure how to do this' - like no documentation on suggestions for how to update your codebase? 'Cause the stuff above should help with the former, but if it's more of a knowledge/documentation issue, that would need specific handling.
But the main point being that, the less time you have to spend in the engine means less stuff that may run aground on updates. My main goal on that front - which I've been testing a lot in my R&D build - is that outside of rendering or custom physics - frame intensive stuff -, everything else should be possible to prototype purely in script and the editors.
That'd go a long way to avoiding 'Update-itosis' from setting in. If all your stuff is in script during the prototyping/design phase, and will only go into the engine for the optimization/finalization phase, it means that updates on the engine side should rarely, if ever, collide with your work.
Heck, in those cases, you should be able to just drop in the new precompiled binary and maybe update a few things script-side and carry on your way.
If you're curious, I can toss you a third person 'screw around' build I've been testing snippets and stuff in to see what I mean, where all the actual gameplay stuff, from player controls and simple physics rules(not like, Bullet/PhysX physics, but player movement), projectile simulation, interaction, etc are pretty much all done with a stock of default engine-side components and script.
#13
@Will: Can't and won't make promises on stuff that isn't being actively looked into with any seriousness, but I will note in the last couple years we've tacked on both Linux and (as of the latest development fork) Mac Desktop support. Both of which rely on opengl. For phones, that'd entail at a minimum openglES, which is a variant. (assuming folks wouldn't want to skip straight to Vulcan) If one or more of the community wished to spend the time and effort to assist with porting, I highly doubt the effort would be rejected out of hand.
12/07/2016 (4:07 pm)
@Ron: thanks. Always good to get a list of stuff to avoid breaking that folks actually like. Easy to overlook when you're in the middle of ripping up the floorboards.@Will: Can't and won't make promises on stuff that isn't being actively looked into with any seriousness, but I will note in the last couple years we've tacked on both Linux and (as of the latest development fork) Mac Desktop support. Both of which rely on opengl. For phones, that'd entail at a minimum openglES, which is a variant. (assuming folks wouldn't want to skip straight to Vulcan) If one or more of the community wished to spend the time and effort to assist with porting, I highly doubt the effort would be rejected out of hand.
#14
Simple, unless companies are willing to continue to stay in the main stream loop, engines like Torque will eventually get sucked under by other open source engines like Unreal Engine 4. GG wants people to make games with their tech, not run from the engine cause it isn't competing with others like UE4, Game Maker 1.4/2.x and Unity.
Maybe I misunderstood GG thinking but I was under the impression over the years they wanted to make Torque 3D and Torque 2D the best engine available for game design. Maybe I'm wrong, if that is the case then I am sorry. But I honestly thought Torque 3D was driving towards a bigger better future then just a Windows, Mac and Linux platform to make games at. I was hoping it would branch and grow to a powerhouse that could compete with UE4/3, GM2 and Unity 3D. It is not looking that the engine wants to move that direction.
T2D seems to want to at least move to IOS and branch out some, why not Torque 3D? T3D was GG best achievement (at least I think so), then it went free which I think hurt the engine as there is a lack of users from what it looks like on the board from the days I used to sell my content packs here. Again, maybe it is just me for being away for so long. But one thing I know, the industry is moving in a new direction with VR (Rift) which Torque 3D has thankfully and now Android games for tablets and phones are the next big thing (again, at least I think it is). I just would like to see the T3D engine grow and get better, not sink cause it is only available for Laptops/Desktop OSes.
Sure, I could be wrong on that way of thinking but it seems everyone... including UE4, GM2 and Unity 3D are making progress with new customers, new people buying into to their tech. I honestly think Android game design and maybe VR Android game design, should be one of the future plans of of Torque 3D. Think about how Android gaming would be if T3D was there to compete with the others? I think T3D would be right on top of the game myself, due to ease of use, great fan base and an open source model.
Just something to think about. Thanks for listening guys and it is nice to be back. BTW, Using Torque for 3d haunt scenes, Halloween window displays and soon a VR haunt (with the Rift) for our coming haunted attraction in 2017 or 18. Depending when we can afford the launch with a more stable overhead. Talk soon!
Will
12/09/2016 (6:43 am)
@ Jeff:Quote:Why do you feel that mobile support is a make or break feature?
Simple, unless companies are willing to continue to stay in the main stream loop, engines like Torque will eventually get sucked under by other open source engines like Unreal Engine 4. GG wants people to make games with their tech, not run from the engine cause it isn't competing with others like UE4, Game Maker 1.4/2.x and Unity.
Maybe I misunderstood GG thinking but I was under the impression over the years they wanted to make Torque 3D and Torque 2D the best engine available for game design. Maybe I'm wrong, if that is the case then I am sorry. But I honestly thought Torque 3D was driving towards a bigger better future then just a Windows, Mac and Linux platform to make games at. I was hoping it would branch and grow to a powerhouse that could compete with UE4/3, GM2 and Unity 3D. It is not looking that the engine wants to move that direction.
T2D seems to want to at least move to IOS and branch out some, why not Torque 3D? T3D was GG best achievement (at least I think so), then it went free which I think hurt the engine as there is a lack of users from what it looks like on the board from the days I used to sell my content packs here. Again, maybe it is just me for being away for so long. But one thing I know, the industry is moving in a new direction with VR (Rift) which Torque 3D has thankfully and now Android games for tablets and phones are the next big thing (again, at least I think it is). I just would like to see the T3D engine grow and get better, not sink cause it is only available for Laptops/Desktop OSes.
Sure, I could be wrong on that way of thinking but it seems everyone... including UE4, GM2 and Unity 3D are making progress with new customers, new people buying into to their tech. I honestly think Android game design and maybe VR Android game design, should be one of the future plans of of Torque 3D. Think about how Android gaming would be if T3D was there to compete with the others? I think T3D would be right on top of the game myself, due to ease of use, great fan base and an open source model.
Just something to think about. Thanks for listening guys and it is nice to be back. BTW, Using Torque for 3d haunt scenes, Halloween window displays and soon a VR haunt (with the Rift) for our coming haunted attraction in 2017 or 18. Depending when we can afford the launch with a more stable overhead. Talk soon!
Will
#15
Thanks for your view.
I'll note up front that the engine is maintained by the community. I'm the lead steering committee guy and we have a core group of dudes that do some big work, as well as the various people using the engine's excellent contributions which has lead to stellar progress so far. But GG's not involved at this point. They passed the torch to us in the community.
That's why I was asking your opinion on why you felt like it was critical to pursue, as I'd be one of the people to talk to about pushing in that direction.
That said, while a lot of engines are tackling mobile, I don't fully agree that it's the 'new direction for the industry'.
Mobile games are very different from your classic PC and console games. Due to hardware limitations, screen size/UI size limitations, and control/input limitations, there's a lot of types of games and genres that are hard if not impossible to properly pursue on a mobile device.
It's a huge industry on it's own, not a doubt about that, but it's not even really targeting the same types of games that PC and consoles are. That's why I feel that it's not 'the new direction' the industry is headed, so much as an additional one.
All that said, you're not wrong in that offering more options can be a good thing, if possible. It's why VR support with the rift/vive was implemented. But it's a community effort, not GG, which means people need to have the know-how and drive to work on that aspect. Currently, most of the people that are utilizing the engine, or want to/have looked into utilizing T3D haven't been asking questions about mobile support, so it hasn't been considered a driving force compared to updated rendering, expanded tools and more flexibility.
If you've got some knowledge in the field though, or know people that do, feel free to push them in this direction, so they can at least provide some ideas, concepts and feedback that may get the ball rolling.
As az said, if people wanna help with getting mobile support to take off, we're definitely not going to say no. It's just that currently, compared to the desktop stuff(and probably consoles in the future), there hasn't been enough interest to push working on support of that to the forefront, but I'm also not going to say it won't happen, because I have no problems with it happening if someone wants to jumpstart that side of production.
That sounds pretty sweet. If you hadn't seen it, Vive support was also added, so you can capitalize on both platforms now :)
Most of the community traffic has shuffled over to the newer forums forums.torque3d.org, so I'd suggest tossing up some posts there about the work you're doing as well as here. I've always liked haunted houses, so I'm really curious to see a VR one in action.
12/09/2016 (7:44 am)
@WillThanks for your view.
I'll note up front that the engine is maintained by the community. I'm the lead steering committee guy and we have a core group of dudes that do some big work, as well as the various people using the engine's excellent contributions which has lead to stellar progress so far. But GG's not involved at this point. They passed the torch to us in the community.
That's why I was asking your opinion on why you felt like it was critical to pursue, as I'd be one of the people to talk to about pushing in that direction.
That said, while a lot of engines are tackling mobile, I don't fully agree that it's the 'new direction for the industry'.
Mobile games are very different from your classic PC and console games. Due to hardware limitations, screen size/UI size limitations, and control/input limitations, there's a lot of types of games and genres that are hard if not impossible to properly pursue on a mobile device.
It's a huge industry on it's own, not a doubt about that, but it's not even really targeting the same types of games that PC and consoles are. That's why I feel that it's not 'the new direction' the industry is headed, so much as an additional one.
All that said, you're not wrong in that offering more options can be a good thing, if possible. It's why VR support with the rift/vive was implemented. But it's a community effort, not GG, which means people need to have the know-how and drive to work on that aspect. Currently, most of the people that are utilizing the engine, or want to/have looked into utilizing T3D haven't been asking questions about mobile support, so it hasn't been considered a driving force compared to updated rendering, expanded tools and more flexibility.
If you've got some knowledge in the field though, or know people that do, feel free to push them in this direction, so they can at least provide some ideas, concepts and feedback that may get the ball rolling.
As az said, if people wanna help with getting mobile support to take off, we're definitely not going to say no. It's just that currently, compared to the desktop stuff(and probably consoles in the future), there hasn't been enough interest to push working on support of that to the forefront, but I'm also not going to say it won't happen, because I have no problems with it happening if someone wants to jumpstart that side of production.
Quote:BTW, Using Torque for 3d haunt scenes, Halloween window displays and soon a VR haunt (with the Rift) for our coming haunted attraction in 2017 or 18.
That sounds pretty sweet. If you hadn't seen it, Vive support was also added, so you can capitalize on both platforms now :)
Most of the community traffic has shuffled over to the newer forums forums.torque3d.org, so I'd suggest tossing up some posts there about the work you're doing as well as here. I've always liked haunted houses, so I'm really curious to see a VR one in action.
#16
T3D even in it's current state is doing MANY MANY things right. As for your mobile comment, sure, I can see that as important to a small and select set of developers.
Keep in mind that the mobile market is...well, in my opinion a virtual junk shop that really has less and less relevance over time. All you have to do is check out the total abundance of articles out on the interwebs concerning 'pay to win' or 'pay to play' mobile games. (seriously, I just read about a guy that spent one million bucks on a mobile game...of which he stole 5 million in order to support his habit.) Not to mention unless you have a pretty large team and a truly unique idea... you will probably not even make a 'blip' on the marketing radar for mobile.
For serious games though. I am keeping an eye on T3D. This team has done some good stuff and yeah... I have used UnReal 4, and CryEngine (the plug-ins and BS software requirements suck for CryTek's beast by the way. We wont even get into the BLOAT of their code base... seriously? Have they ever 'cleaned anything out" over the years?)
Anyway, my point is this... we are in a totally new generation of development. T3D will benefit from performing well as a proper PC/Linux game engine if these fine developers pull off what they are doing. Mobile and all that... heck, that is 'extra' as far as I and I assume, a large number of developers are concerned. Oh and by the way.... from experience... UnReal Mobile is GIMPed as all hell. Just making a point.
As for VR... uhm... yeah... nobody is using it. Develop it sure, but really?? The sales are TERRIBLE. Getting a real '3D TV' vibe here at the moment. Support sure, focus...'F' no.
Either way, thanks for starting the conversation and making some good points. I agree, eventually T3D will need a mobile version. However, let's get a solid AAA game engine set in stone first ok.
Ron
12/14/2016 (1:35 pm)
@Will, Thanks for the comment and unfortunately I have to disagree with you on a few points. T3D even in it's current state is doing MANY MANY things right. As for your mobile comment, sure, I can see that as important to a small and select set of developers.
Keep in mind that the mobile market is...well, in my opinion a virtual junk shop that really has less and less relevance over time. All you have to do is check out the total abundance of articles out on the interwebs concerning 'pay to win' or 'pay to play' mobile games. (seriously, I just read about a guy that spent one million bucks on a mobile game...of which he stole 5 million in order to support his habit.) Not to mention unless you have a pretty large team and a truly unique idea... you will probably not even make a 'blip' on the marketing radar for mobile.
For serious games though. I am keeping an eye on T3D. This team has done some good stuff and yeah... I have used UnReal 4, and CryEngine (the plug-ins and BS software requirements suck for CryTek's beast by the way. We wont even get into the BLOAT of their code base... seriously? Have they ever 'cleaned anything out" over the years?)
Anyway, my point is this... we are in a totally new generation of development. T3D will benefit from performing well as a proper PC/Linux game engine if these fine developers pull off what they are doing. Mobile and all that... heck, that is 'extra' as far as I and I assume, a large number of developers are concerned. Oh and by the way.... from experience... UnReal Mobile is GIMPed as all hell. Just making a point.
As for VR... uhm... yeah... nobody is using it. Develop it sure, but really?? The sales are TERRIBLE. Getting a real '3D TV' vibe here at the moment. Support sure, focus...'F' no.
Either way, thanks for starting the conversation and making some good points. I agree, eventually T3D will need a mobile version. However, let's get a solid AAA game engine set in stone first ok.
Ron
#17
Ron
01/04/2017 (1:23 pm)
So, bad spot to do this but, I do have a number of 'samples' of my photo scanned 'VituaScan' Products available for anyone interested at 3tdstudios.com. THEY ARE ALL IN .FBX format and will NOT work in T3D in it's current state. They will work in any other game engine or 3d renderer that imports and works with that format. They are free to use for commercial and personal projects with no need of credit or anything else. If there is enough interest, I may work on T3D compatible version of these as well. Let me know.Ron
#18
And I wouldn't fret TOO hard on the format business. for 4.0 we'll be getting assimp rolled in, so T3D'll support FBX as well :)
01/04/2017 (4:26 pm)
Sweet. I'll be giving these a look for sure.And I wouldn't fret TOO hard on the format business. for 4.0 we'll be getting assimp rolled in, so T3D'll support FBX as well :)
Associate Steve Acaster
[YorkshireRifles.com]
The brave/foolish* souls still using Torque tend to be over at forums.torque3d.org/ these days.
(* delete as applicable)