TGEA description under products. Physics?
by Jaimi McEntire · in Torque Game Engine Advanced · 01/13/2009 (7:19 pm) · 29 replies
Quote:Torque includes efficiently networked physics functionality, and is built with an abstracted C++ Physics API for use with other major 3rd party physics libraries as well. Sample implementations include:
PhysX
Bullet
ODE
Using these libraries, Torque supports:
Rigid body
Vehicles
Ragdoll
Destructible objects
Dynamic fluid
Particle system physics
Destroyable joints
Various vehicles
Fluid buoyancy
This would be really cool, but I'm guessing this is for the next version?
#22
http://www.garagegames.com/community/blogs/view/15989
01/14/2009 (6:33 pm)
wow, looks like this comment got some action:http://www.garagegames.com/community/blogs/view/15989
#23
I couldn't agree more. This isn't "marketing liberties" when you know the physics resources you mentioned are broken. I would call this dishonest. That may sound harsh but I don't know what else to call it when you knowingly say an engine has a certain feature that doesn't actually work. When a customer buys the engine they expect to have physis working out of the box unless otherwise noted. You don't mention anywhere that the physics feature is based off of broken community supplied resources. I think someone purchasing the engine has a right to know that physics is broken before making a purchase.
Furthermore, I think its cheap to list Physics as a feature when it's not a part of the core engine and especially if it doesn't work. It's not about being "petty" it's about being honest so that customers know the full story when they buy a product. If I bought TGEA and found out physics wasn't a part of the core engine and then found it was an old broken community supplied resource, I would be really angry. Possibly angry enouph to seek legal counsel.
If this sounds harsh I apologize as I'm not trying demonize anyone personally. I just feel very strongly that this is very unethical and the right thing to do would be to list only features that actually work and are a part of the core engine.
01/15/2009 (3:33 pm)
Quote:
January 14, 2009 @Brett:
My original post was assuming that these were coming for Torque3D, and someone had jumped the gun putting them up in the project description - That would be a great feature for T3D. I'd assumed someone would post "Oops, we got caught with egg on our face..." and delete these features.
But please don't jump on here and say that is supported now when it's not. Half written, buggy, and incompatible resources from older engines/versions do not constitute an "efficiently networked abstracted C++ Physics API".
Quote:
--------------------------------------------------------------------------------
Seems bizarre to begrudge GG this kind of thing in any case
--------------------------------------------------------------------------------
I'm not begrudging anything. This was more of a wakeup call, thinking maybe you prematurely posted something. But if you want to leave it up there, it is your business not mine.
I couldn't agree more. This isn't "marketing liberties" when you know the physics resources you mentioned are broken. I would call this dishonest. That may sound harsh but I don't know what else to call it when you knowingly say an engine has a certain feature that doesn't actually work. When a customer buys the engine they expect to have physis working out of the box unless otherwise noted. You don't mention anywhere that the physics feature is based off of broken community supplied resources. I think someone purchasing the engine has a right to know that physics is broken before making a purchase.
Furthermore, I think its cheap to list Physics as a feature when it's not a part of the core engine and especially if it doesn't work. It's not about being "petty" it's about being honest so that customers know the full story when they buy a product. If I bought TGEA and found out physics wasn't a part of the core engine and then found it was an old broken community supplied resource, I would be really angry. Possibly angry enouph to seek legal counsel.
If this sounds harsh I apologize as I'm not trying demonize anyone personally. I just feel very strongly that this is very unethical and the right thing to do would be to list only features that actually work and are a part of the core engine.
#24
Rigid body physics is part of TGEA - via the internal RigidShape class. It just didn't have all the features mentioned, that is all.
01/15/2009 (3:53 pm)
@Jason - Rigid body physics is part of TGEA - via the internal RigidShape class. It just didn't have all the features mentioned, that is all.
#25
If the engine supports rigid body physics then list that and the features that go along with it as long as they all actually work.
Again, I'm not trying to be a jerk about this and if I'm coming across too harsh i apologize, however, I've been burned by this kind of thing in the past so it's somewhat of a sore spot for me. I would hate to have other people spend their hard earned money on a product when they don't have an honest picture of what they're buying.
That's all i'm trying to say for what it's worth.
01/15/2009 (4:05 pm)
I understand that rigid body physics are there. My issue is listing the PhysX, ODE, and Bullet items. Then it says using these libraries torque supports the following features... However, if all those resources are broken then how can those features be supported. Only working out of the box features should be listed or it should be noted which features aren't currently working so that a customer knows what they're getting into.If the engine supports rigid body physics then list that and the features that go along with it as long as they all actually work.
Again, I'm not trying to be a jerk about this and if I'm coming across too harsh i apologize, however, I've been burned by this kind of thing in the past so it's somewhat of a sore spot for me. I would hate to have other people spend their hard earned money on a product when they don't have an honest picture of what they're buying.
That's all i'm trying to say for what it's worth.
#26
That said, the existing PhysX resource is also severely limited (and the ODE even more so). It hasn't been significantly updated for a long time, doesn't handle large collision meshes well, and has no functional network interpolation (and so far none of us have figured out how to fix that bit).
That line is included there because it's simply no longer acceptable these days to sell a modern engine without a physics solution (Rigid Class doesn't cut it).
That said, I'm happy to skip entirely over whether or not it's fair to use this in marketing because I know how unpleasant it is to market a product, especially to the mostly dysfunctional entertainment industry. I'm happy to skip that if we can talk about this:
T3D needs to have PhysX already built in. Not as a community resource, not as a suggested do-it-yourself addon. It needs to be supported out of the box. It doesn't necessarily need to include ragdoll corpses or PhysX vehicles/players out of the box, but you should be able to drop in some PhysX actors and let them roll around. If you have a better physics solution lined up, that's fine too, I'm just suggesting PhysX because it's easy, it's free, and nVidia has been sneaking the hardware into our PC's for over a year now.
This sounds like a whine (and it's not exactly the right thread for it), but I (and many others) have been trying to get a comprehensive physics solution working in TGE for years now. I for one will not upgrade to a new engine without out of the box physics support. I'm fine with TGEA's level of functionality for graphics and networking, so unless T3D has a major physics upgrade, I'm uninterested at best.
01/15/2009 (4:32 pm)
That's really not especially ethical in my opinion. I agree that, in the hands of a skilled dev team, you can implement PhysX support without it being especially difficult, but you can also implement any number of other 3rd party packages if you know what you're doing, so why don't we advertise those as well?That said, the existing PhysX resource is also severely limited (and the ODE even more so). It hasn't been significantly updated for a long time, doesn't handle large collision meshes well, and has no functional network interpolation (and so far none of us have figured out how to fix that bit).
That line is included there because it's simply no longer acceptable these days to sell a modern engine without a physics solution (Rigid Class doesn't cut it).
That said, I'm happy to skip entirely over whether or not it's fair to use this in marketing because I know how unpleasant it is to market a product, especially to the mostly dysfunctional entertainment industry. I'm happy to skip that if we can talk about this:
T3D needs to have PhysX already built in. Not as a community resource, not as a suggested do-it-yourself addon. It needs to be supported out of the box. It doesn't necessarily need to include ragdoll corpses or PhysX vehicles/players out of the box, but you should be able to drop in some PhysX actors and let them roll around. If you have a better physics solution lined up, that's fine too, I'm just suggesting PhysX because it's easy, it's free, and nVidia has been sneaking the hardware into our PC's for over a year now.
This sounds like a whine (and it's not exactly the right thread for it), but I (and many others) have been trying to get a comprehensive physics solution working in TGE for years now. I for one will not upgrade to a new engine without out of the box physics support. I'm fine with TGEA's level of functionality for graphics and networking, so unless T3D has a major physics upgrade, I'm uninterested at best.
#27
01/15/2009 (4:52 pm)
I don't know where US law stands on the issue, but where I come from, advertising something is considered to be an implicit promise to deliver it. What I find sad is that we often hear from GG staff that they don't wish to discuss the new features they are working on, because they don't want to be raked over the coals for not delivering them, even when they clearly state that there are no promises. These advertised features effectively are promises, so don't be surprised if you start smelling coals.
#28
It's at version 1.1.1. It pre-dates 90% of the features in TGB. So no, TGB as advertised does not support Linux in any reasonable way shape or form.
And you know what? That's OK. GG has repeatedly stated that they have no Linux-experienced developers in house and no intention of supporting Linux. That's also OK. What's not ok is then claiming on the software sales page you support it, since you've repeatedly said you can't and won't!
re: As for 360, Wii and iPhone, those are obviously legit and it seems pretty petty to complain about listing those even though they're not included in the price listed. It's a pretty major feature of TGB to have available paths to major consoles and devices.
No, they are not obviously legit. If I go to the page that lists those as supported, http://www.garagegames.com/products/torque-2D, and click the buy button, I do not receive a product that contains support for those platforms. No matter which way you bend it, this is incorrect.
I have a TGB Professional license, and when I download the package it contains no indication of any kind of way to deploy it to any consoles, Linux, or the web.
re: There do exist sample implementations for Bullet, PhysX and ODE as resources
Oh, your advertising is talking about the code that *I* wrote! I didn't realise that!
In that case I call bollocks on the line "Torque includes efficiently networked physics functionality, and is built with an abstracted C++ Physics API".
"Bollocks" in the sense of "ODEScript makes use of nothing you could conceivably describe as an abstracted physics API, because such a thing does not exist".
To be fair, I aired both these gripes with mPerry in IRC an hour ago, and he had valid-ish answers to both. Something along the lines of "I'm ok with the wording about the physics stuff", which upon re-reading is sufficiently obtuse that I guess a good lawyer would get away with it.
He also mentioned that I can't see console deployment paths because I don't have a wii or 360 development kit. I guess that's almost valid, but I still maintain that if I go to one of those webpages and cliky "buy", I get something different from what's advertised.
On another note, it would be wise to avoid describing your paying customers as "petty" because they gave you [or are about to give you] money expecting one thing and got something else
Gary (-;
01/15/2009 (6:28 pm)
re: TGB still does have Linux support IIRC. Is it up to 1.7.4?It's at version 1.1.1. It pre-dates 90% of the features in TGB. So no, TGB as advertised does not support Linux in any reasonable way shape or form.
And you know what? That's OK. GG has repeatedly stated that they have no Linux-experienced developers in house and no intention of supporting Linux. That's also OK. What's not ok is then claiming on the software sales page you support it, since you've repeatedly said you can't and won't!
re: As for 360, Wii and iPhone, those are obviously legit and it seems pretty petty to complain about listing those even though they're not included in the price listed. It's a pretty major feature of TGB to have available paths to major consoles and devices.
No, they are not obviously legit. If I go to the page that lists those as supported, http://www.garagegames.com/products/torque-2D, and click the buy button, I do not receive a product that contains support for those platforms. No matter which way you bend it, this is incorrect.
I have a TGB Professional license, and when I download the package it contains no indication of any kind of way to deploy it to any consoles, Linux, or the web.
re: There do exist sample implementations for Bullet, PhysX and ODE as resources
Oh, your advertising is talking about the code that *I* wrote! I didn't realise that!
In that case I call bollocks on the line "Torque includes efficiently networked physics functionality, and is built with an abstracted C++ Physics API".
"Bollocks" in the sense of "ODEScript makes use of nothing you could conceivably describe as an abstracted physics API, because such a thing does not exist".
To be fair, I aired both these gripes with mPerry in IRC an hour ago, and he had valid-ish answers to both. Something along the lines of "I'm ok with the wording about the physics stuff", which upon re-reading is sufficiently obtuse that I guess a good lawyer would get away with it.
He also mentioned that I can't see console deployment paths because I don't have a wii or 360 development kit. I guess that's almost valid, but I still maintain that if I go to one of those webpages and cliky "buy", I get something different from what's advertised.
On another note, it would be wise to avoid describing your paying customers as "petty" because they gave you [or are about to give you] money expecting one thing and got something else
Gary (-;
#29
I am beginning to think that the person writing the advertising has no clue of the products they are selling.
01/16/2009 (5:58 am)
I found it odd that they used ChunkyKs's ODE implementation in the advertising.I am beginning to think that the person writing the advertising has no clue of the products they are selling.
Torque Owner Jeremy Easoz
So you tell me you know how big the team was that made Silverlight at Microsoft but you don't know if your own product supports linux or not.
I've been holding my military sign on bonus for awhile now waiting on Torque 2.0 but you know what.
Fuck it, i'm taking my money else where.
Giving us the run around like were morons or something.
You know damn well if something isn't built into the engine and works out of the box, you can't put it on your specifications list and sell it like it does.
Someone posted what 2 days ago "Depth of View shader for sale" for TGEA.
Why, because tgea doesn't ship with a depth of view shader?