to xna or tge?
by Rad · 09/12/2008 (7:34 am) · 12 comments
i was just getting into tge, and the whole dts business...then a little birdie told me about torque x and to go straight there as that is the future and the rest is soon to be dead history
now i'm looking at just diving straight in to xna
i already am familiar with c#, .net and all the rest, having done a ton of code for non-game projects - did download the updates to visual studio 2008 - a whole freaking dvd iso lol
plus the need for a long, at least freebie, import/export toolchain just to create dts's and get it into tge got me a bit frustrated - anim8or (thanks to whoever mentioned it on the forums) to 3ds, into customized blender to export it as a dts for torque - that's a bit whacked imo - and blender's interface sucks pingpong balls
i was thinking of developing a simplistic dts generator to bone mesh up real-world images straight from tlrb pngs - sort of a halfass digitizer to bone up and wrap the mesh round the inside of the image instead of the other way around of trying to create the boned mesh first, then covering it with the png as a texture
maybe even a rotatable perspective view of the ortho png's and a nice flat stretchy mesh to wrap it with, and some bones for inserting - sort of like making a clay model than making a dts - who knows, it might catch on - it would be a nice addition and would be fairly easy to produce, and i may still do it, although i'll have to see what xna needs first
the 360 sata drive works great
i just found out mass effect was xna
time to start downloading the tutorials - x 1.0.51 or 2?
waiting on the vs2008 dvd iso to download - a long time between %'s
got both installers for 1.0.51 and 2 just in case
--rad
now i'm looking at just diving straight in to xna
i already am familiar with c#, .net and all the rest, having done a ton of code for non-game projects - did download the updates to visual studio 2008 - a whole freaking dvd iso lol
plus the need for a long, at least freebie, import/export toolchain just to create dts's and get it into tge got me a bit frustrated - anim8or (thanks to whoever mentioned it on the forums) to 3ds, into customized blender to export it as a dts for torque - that's a bit whacked imo - and blender's interface sucks pingpong balls
i was thinking of developing a simplistic dts generator to bone mesh up real-world images straight from tlrb pngs - sort of a halfass digitizer to bone up and wrap the mesh round the inside of the image instead of the other way around of trying to create the boned mesh first, then covering it with the png as a texture
maybe even a rotatable perspective view of the ortho png's and a nice flat stretchy mesh to wrap it with, and some bones for inserting - sort of like making a clay model than making a dts - who knows, it might catch on - it would be a nice addition and would be fairly easy to produce, and i may still do it, although i'll have to see what xna needs first
the 360 sata drive works great
i just found out mass effect was xna
time to start downloading the tutorials - x 1.0.51 or 2?
waiting on the vs2008 dvd iso to download - a long time between %'s
got both installers for 1.0.51 and 2 just in case
--rad
About the author
#2
09/12/2008 (8:33 am)
I have been looking over theTGEA and the XNA versions trying to decide which one would be better to build design to. I am not convinced the XNA versions are going to dominate or take over anytime soon. I still haven't seen speed comparisons of a large game done in TGEA vs one done in Torque X. I would like to think they would be the same.
#3
My toolchain consists of - model in MS3D (35$ program), do UV mapping in UU3D (30ish dollar program), animate in MS3D and export my DSQs, drop right into TGE...
Is that really a whacked toolchain?
09/12/2008 (8:55 am)
Quote:
plus the need for a long, at least freebie, import/export toolchain just to create dts's and get it into tge got me a bit frustrated
My toolchain consists of - model in MS3D (35$ program), do UV mapping in UU3D (30ish dollar program), animate in MS3D and export my DSQs, drop right into TGE...
Is that really a whacked toolchain?
#4
09/12/2008 (10:07 am)
Mass Effect on XNA is priceless. 
#5
09/12/2008 (11:05 am)
why would you go from anim8tor to 3ds to blender....thats just stupid...3ds can convert to .dts already. 
#7
09/12/2008 (3:04 pm)
Toolchain = Blender 3D and Constructor. $0. All good. 
#8
I definitely do not see Torque X as a competitor to TGE/A or even XNA. Each engine has its own place. TGE is the best option for games targeting multiple platforms, including Windows, MacOS, and Linux. TGEA is the better Windows-only choice for sexy shader support. Torque X is the best solution for Xbox 360 games or fast prototyping of Windows games.
TGE & TGEA both are based on C++ whereas Torque X is absed on C#. What I really like about the C# language is so easy to work with, that I can put together new game concepts really quickly - in a matter of hours in fact. Instead of fighting with the C++ language structure and pointer management and class hierarchies, I can simply focus game-related problems. (this is an over-simplification, of course).
Torque X is also a value-adder to XNA. You can make great games in XNA, but they'll just take longer to create them than with Torque X. Torque X provides services for level loading, spawning, input mapping, materials, GUI, and so on, which all help you save time.
In any case, if you are already familiar with C#, then looking at Torque X is really worth the time. Also, be sure to pull down 2.0 (the latest release) since it is compatible with XNA 2.0
John K.
09/12/2008 (4:53 pm)
Unfortunately, Torque X doesn't support DIF interior formats, so Constructor won't really help. I definitely do not see Torque X as a competitor to TGE/A or even XNA. Each engine has its own place. TGE is the best option for games targeting multiple platforms, including Windows, MacOS, and Linux. TGEA is the better Windows-only choice for sexy shader support. Torque X is the best solution for Xbox 360 games or fast prototyping of Windows games.
TGE & TGEA both are based on C++ whereas Torque X is absed on C#. What I really like about the C# language is so easy to work with, that I can put together new game concepts really quickly - in a matter of hours in fact. Instead of fighting with the C++ language structure and pointer management and class hierarchies, I can simply focus game-related problems. (this is an over-simplification, of course).
Torque X is also a value-adder to XNA. You can make great games in XNA, but they'll just take longer to create them than with Torque X. Torque X provides services for level loading, spawning, input mapping, materials, GUI, and so on, which all help you save time.
In any case, if you are already familiar with C#, then looking at Torque X is really worth the time. Also, be sure to pull down 2.0 (the latest release) since it is compatible with XNA 2.0
John K.
#9
Ive been working with TGB but XNA keeps pulling at me, just ordered Johns Torque X book earlier today. We will see. PC AND Xbox has potential.
09/12/2008 (11:00 pm)
Since you seam to be quite familiar with C# I would think XNA would be the way to go, big head start there. And as it appears you have TGB you should have TXB as well, cheap entry.Ive been working with TGB but XNA keeps pulling at me, just ordered Johns Torque X book earlier today. We will see. PC AND Xbox has potential.
#10
Only problem with ordering the book is none of the uk bookstores seem to be stocking it until October and delivery to here could take 7 days, both of which is no good for the dream build play compo. Shame there isn't a purchasable e book version. :(
Any heads-up on stockist/providers that do quick overseas delivery would be welcomed!
Anyway, just got my xbox yesterday, so it's time (God Willing) for that moment when you get to see your prototype / creation for the first time running on your console on your TV, sweet! Hopefully, all goes well :P
09/13/2008 (3:54 am)
I think I'm due to get John's book too as the components thing is complicating things for me rather then help in some aspects at the mo. I'm also trying to get my head around what commands to call/use for when, getting there though.Only problem with ordering the book is none of the uk bookstores seem to be stocking it until October and delivery to here could take 7 days, both of which is no good for the dream build play compo. Shame there isn't a purchasable e book version. :(
Any heads-up on stockist/providers that do quick overseas delivery would be welcomed!
Anyway, just got my xbox yesterday, so it's time (God Willing) for that moment when you get to see your prototype / creation for the first time running on your console on your TV, sweet! Hopefully, all goes well :P
#11
I also agree with jydog. I know a lot of people most comfortable with C++ and like getting into the details of the engine. I steer them to TGE/A. I also have a bunch of friends that know C# and only have "weekend" time to make games. Performance is less important to them, but high productivity with little time is crucial. I steer them to Torque X. But just to be clear, XNA/Torque X performance is just fine - simple scenes run at abot 400fps and very dense 3D scenes with full bloom run at 70fps (according to Fraps). Of course, you can really bash it and bring it down to 2-3fps if you wanted to - but overall, performance is pretty competitive with TGE.
John K.
09/13/2008 (9:02 am)
Leroy, sorry to hear about the component complication. The book makes a real effort to shape how you think about components as problem solvers. I really like TGB's use of the word "behaviors" since it's a nice way to think about components. For example - How do you want a specific scene object (like a sprite, scroller, particle effect, etc) to behave? Also, all the code samples from the book are going up on the torquexbook.com website tomorrow (Sunday). So everyone can start looking there for code samples (even without the book - wink)I also agree with jydog. I know a lot of people most comfortable with C++ and like getting into the details of the engine. I steer them to TGE/A. I also have a bunch of friends that know C# and only have "weekend" time to make games. Performance is less important to them, but high productivity with little time is crucial. I steer them to Torque X. But just to be clear, XNA/Torque X performance is just fine - simple scenes run at abot 400fps and very dense 3D scenes with full bloom run at 70fps (according to Fraps). Of course, you can really bash it and bring it down to 2-3fps if you wanted to - but overall, performance is pretty competitive with TGE.
John K.
#12
I should be getting the book, it's more a question of when, maybe I'll chance it next week and see if it arrives before the weekend, which would give me a few days to work through some of it.
As far as the components are concern, I think I understand their benefit, I just don't really see where some of the choices of code commands/structures are used in the tutorials, I suppose I need to grid through it a bit, might be easier once I figure out had to check for collisions etc. Anyhow, thanks again :)
09/13/2008 (11:10 am)
Thanks for the tip John ;)I should be getting the book, it's more a question of when, maybe I'll chance it next week and see if it arrives before the weekend, which would give me a few days to work through some of it.
As far as the components are concern, I think I understand their benefit, I just don't really see where some of the choices of code commands/structures are used in the tutorials, I suppose I need to grid through it a bit, might be easier once I figure out had to check for collisions etc. Anyhow, thanks again :)
Associate David Montgomery-Blake
David MontgomeryBlake
Mass Effect was not an XNA game. It was developed with a full XBox 360 DevKit and a ton of dedicated console programmers working in C++.
Unless you are planning on using XNA 3.0's pre-release, I would not recommend installing VS C# 2008 Express. XNA 2.0 (and TX 2.0, which is built on XNA 2.0) does not support it. They support 2005 Express. 3.0 will add 2008 support, but it is still in pre-release mode right now.
Also, you will want TX 2.0 and XNA 2.0 rather than previous versions. They are the stable versions which are recommended currently.
Your 2D image > 3D model > bones model system is interesting, but extremely complex in terms of development. The tech behind creating 3D models from 2D images is out there and has been presented quite often at numerous conferences and is available in tons of whitepapers, but it is still emerging tech in terms of realistic and useful utilities. You might want to look at some of the whitepapers on procedural bone and mesh generation systems if you choose to continue with the idea.
But I wouldn't consider it "fairly easy" by any stretch of the imagination. Well, okay, perhaps in terms of raycasting a 2D map into 3D ala Wolf3D or Doom, but not in terms of taking a picture of a building and attempting to reproduce the visual aspects (and approximating the unseen aspects).