Hello Jet 3d folks
Posted: Tue Aug 09, 2005 1:54 am
First off, hello to all the old-timers in the crowd - good to see everyone actually talking and conversing, even if it is a bit of a pitched battle ;}
I really don't think it's worth fighting about who's renderer is the better one - it's like jocks competing over who has the most inadequate body part...
The biggest issue I think is the fact that most people simply do not realize how much work has been put into the Genesis side of the code by the RF developers (QoD & Wxb1 primarily) in the past few years.
These two have taken Genesis from the antique renderer that it was when eclipse left the building to a fairly modern, up to date renderer.
Some things that have been added recently take RF beyond most of the 'free' engines on the market, namely:
- normal mapping (dot3)
- embm bumpmapping (for geometry)
- stencil shadows (up to 8 lights with dynamic shadowing)
- projected shadows (optimized detail shadows for actors if you want)
RF's sheer flexibility is what makes it appealing to most people - Jet 3d is and will be stuck in a rut as far as users goes until it has a gameshell.
As a few of the Jet3d developers have mentioned, they spend most of their time in the editor flying around in 'cheat mode' without a true gameshell to run programs in.
The entire RF community has been built around the fact that there IS a gameshell to play their games in - an editor is fine - the Jet editor is truly more advanced and a much better way to approach things than the Genesis compile/text/compile/test version, but still doesn't solve the issue of actually 'doing' anything with Jet as an engine.
It's like having a car engine that is super powerful but no car. it might be cool, it might make alot of noise, but it won't get you anywhere.
RF is as popular as it is simply because ANYONE can get in, make a level and get their own 'game' up and running in a period of a few days.
I have been teaching a summer camp for 12 & 13 year olds this past week and they had their own RF games up and running the first day of the camp.
This is the power of RF - having people that have never made games before able to get their hands dirty and 'take ownership' of something that they created in the engine.
Without this crucial step, Jet is a programmers tool and not much else.
It's like the other big open source engines like ogre etc - a dump of source code is useless to 99% of the community without a running exe that can be configured and customized.
Destiny had the same problem - and over emphasis on the renderer and no emphasis on the gameshell component of the engine.
With RF we did the exact opposite approach - Ed Averill & Ralph Deane did wonders by simply focusing on the features that the most people were likely to need to get started using the system, namely:
1) customizable cameras
2) customizable player config (gravity, movement speed etc)
3) menu's
4) hud
and so on...
If Jet could provide a simple gameshell that lets the users create their own applications from start to finish, then alot more people would be willing to overlook the things in Jet that may not be 100% up to date or complete.
It's the difference between Linux hackers and Windows users. To windows users, linux is incomprehensible and useless, because it just doesn't let them 'do stuff'
To linux hackers however, it's a dream OS because you have to get under the hood to do anything.
I'm overexaggerating of course, but that's the basic point.
Joel Spolsky wrote a good article about this difference here:
http://www.joelonsoftware.com/articles/ ... alism.html
It really sums up the differences which I believe are the core clash of cultures that we are seeing in this thread.
The difference between code that is cool for coding sake, and code that actually 'does stuff'.
I don't mean to be creating more of a storm here, I fully believe that Jet has the potential to do some very cool things, but I think the project needs a different approach to attract users.
It's the same problem that the GarageGames folks are having - they keep asking their community why more games aren't made with the engine, and the core reason is that probably 99% of the people that use Torque are not programmers, and they are not able to do much besides what the original code provides.
If you want people to use the engine, then provide some simply way for them to configure and modify the engine without programming
This is our main motivation that we are using to enhance Beyond Virtual - the Renderer, while it is a goal, is not what will make people use the engine.
Useability, Portability and Compatibility are and should be the Top Goals of any game engine.
Anyone can write a renderer - to create a truly useable game ENGINE takes alot more pieces than just the renderer.
Cheers
Mike W
http://www.beyondvirtual.com
I really don't think it's worth fighting about who's renderer is the better one - it's like jocks competing over who has the most inadequate body part...
The biggest issue I think is the fact that most people simply do not realize how much work has been put into the Genesis side of the code by the RF developers (QoD & Wxb1 primarily) in the past few years.
These two have taken Genesis from the antique renderer that it was when eclipse left the building to a fairly modern, up to date renderer.
Some things that have been added recently take RF beyond most of the 'free' engines on the market, namely:
- normal mapping (dot3)
- embm bumpmapping (for geometry)
- stencil shadows (up to 8 lights with dynamic shadowing)
- projected shadows (optimized detail shadows for actors if you want)
RF's sheer flexibility is what makes it appealing to most people - Jet 3d is and will be stuck in a rut as far as users goes until it has a gameshell.
As a few of the Jet3d developers have mentioned, they spend most of their time in the editor flying around in 'cheat mode' without a true gameshell to run programs in.
The entire RF community has been built around the fact that there IS a gameshell to play their games in - an editor is fine - the Jet editor is truly more advanced and a much better way to approach things than the Genesis compile/text/compile/test version, but still doesn't solve the issue of actually 'doing' anything with Jet as an engine.
It's like having a car engine that is super powerful but no car. it might be cool, it might make alot of noise, but it won't get you anywhere.
RF is as popular as it is simply because ANYONE can get in, make a level and get their own 'game' up and running in a period of a few days.
I have been teaching a summer camp for 12 & 13 year olds this past week and they had their own RF games up and running the first day of the camp.
This is the power of RF - having people that have never made games before able to get their hands dirty and 'take ownership' of something that they created in the engine.
Without this crucial step, Jet is a programmers tool and not much else.
It's like the other big open source engines like ogre etc - a dump of source code is useless to 99% of the community without a running exe that can be configured and customized.
Destiny had the same problem - and over emphasis on the renderer and no emphasis on the gameshell component of the engine.
With RF we did the exact opposite approach - Ed Averill & Ralph Deane did wonders by simply focusing on the features that the most people were likely to need to get started using the system, namely:
1) customizable cameras
2) customizable player config (gravity, movement speed etc)
3) menu's
4) hud
and so on...
If Jet could provide a simple gameshell that lets the users create their own applications from start to finish, then alot more people would be willing to overlook the things in Jet that may not be 100% up to date or complete.
It's the difference between Linux hackers and Windows users. To windows users, linux is incomprehensible and useless, because it just doesn't let them 'do stuff'
To linux hackers however, it's a dream OS because you have to get under the hood to do anything.
I'm overexaggerating of course, but that's the basic point.
Joel Spolsky wrote a good article about this difference here:
http://www.joelonsoftware.com/articles/ ... alism.html
It really sums up the differences which I believe are the core clash of cultures that we are seeing in this thread.
The difference between code that is cool for coding sake, and code that actually 'does stuff'.
I don't mean to be creating more of a storm here, I fully believe that Jet has the potential to do some very cool things, but I think the project needs a different approach to attract users.
It's the same problem that the GarageGames folks are having - they keep asking their community why more games aren't made with the engine, and the core reason is that probably 99% of the people that use Torque are not programmers, and they are not able to do much besides what the original code provides.
If you want people to use the engine, then provide some simply way for them to configure and modify the engine without programming
This is our main motivation that we are using to enhance Beyond Virtual - the Renderer, while it is a goal, is not what will make people use the engine.
Useability, Portability and Compatibility are and should be the Top Goals of any game engine.
Anyone can write a renderer - to create a truly useable game ENGINE takes alot more pieces than just the renderer.
Cheers
Mike W
http://www.beyondvirtual.com