Comments on: Rendering Technology, What’s Next? So, here's what I think, really blunt and in your face With all this sophistication, artists are taking waaaay to fucking long to do this stuff because they are trying to make every pixel and polygon perfect. They need to learn to go into blitz mode AT THE BEGINNING of the project and just hash a bunch of stuff out. Then once you see it in game, if something stands out to the average player, you fix it. I mean, if you are I are looking at it - I think we have pretty good eyes - and give it a pass, 97.5% of the population will give it a pass too. The artists however, sit in their fucking tower and poke sticks at each other and are like oh, I like that, I don't like that, Why did you change that? I don't like that color, oh make the hair look more like fucking Brad Pitts hair, that animation looks strange on frame 16 when you look at it in slow motion while you are high, lets spend 3 hours fixing that... etc. I HAVE SPOKEN! The other part is gameplay, I mean, it doesn't "flesh out". Its good at the beginning or it sucks forever. Period. But most people approach gameplay like a movie studio that says "hey, lets keep doing plastic surgery on this fat ugly bitch until she looks like what we want" instead of doing a casting call of several actresses, throwing out the fat ugly bitches, then having the 3 top hotties jello wrestle and giving the best one the role. I HAVE SPOKEN! As far as engineering goes, the rendering looks great when gameplay sucks WHY? Because: No one else has any idea how that stuff works and you have free reign to do whatever you want! Such as using proper engineering methods to find the solutions to things. Shocking strategy there. (Deep cartoon narrator voice) MEANWHILE ON THE GAMEPLAY TEAM: Producer: Hey guys, isn't this bitch sexy? Lets make some more changes to her. Team: uhhh yeah... I mean no... she's fucking ugly, so yeah lets do some plastic surgery on this ape monster. (Deep cartoon narrator voice) THREE MONTHS LATER: Producer: Hey guys, isn't this bitch sexy? Lets make some more changes to her. Team: ugh... she's still ugly... but there's been some improvement... I'm starting to have some faith in this horse-faced pit beast (Deep cartoon narrator voice) ANOTHER SIX MONTHS LATER: Producer: Hey guys, isn't this bitch sexy? Lets make some more changes to her. Team: Ummm.... Is she sexy? I can't remember... I've... been with her so long I can't tell anymore. But... I've gotten used to the way she bitches and how to get around it, and... I'm growing ever more fond of her spicey cheetos breath... Deep cartoon narrator voice) ANOTHER SIX MONTHS LATER: Producer: Hey guys, isn't this bitch sexy? I know its alpha, but... lets make some more changes to her. Team: I just don't care anymore. I just don't. This shit faced whale has consumed my soul. Let's just ship this fucking thing. I HAVE SPOKEN! So, here’s what I think, really blunt and in your face

With all this sophistication, artists are taking waaaay to fucking long to do this stuff because they are trying to make every pixel and polygon perfect. They need to learn to go into blitz mode AT THE BEGINNING of the project and just hash a bunch of stuff out. Then once you see it in game, if something stands out to the average player, you fix it.

I mean, if you are I are looking at it – I think we have pretty good eyes – and give it a pass, 97.5% of the population will give it a pass too. The artists however, sit in their fucking tower and poke sticks at each other and are like oh, I like that, I don’t like that, Why did you change that? I don’t like that color, oh make the hair look more like fucking Brad Pitts hair, that animation looks strange on frame 16 when you look at it in slow motion while you are high, lets spend 3 hours fixing that… etc. I HAVE SPOKEN!

The other part is gameplay, I mean, it doesn’t “flesh out”. Its good at the beginning or it sucks forever. Period. But most people approach gameplay like a movie studio that says “hey, lets keep doing plastic surgery on this fat ugly bitch until she looks like what we want” instead of doing a casting call of several actresses, throwing out the fat ugly bitches, then having the 3 top hotties jello wrestle and giving the best one the role. I HAVE SPOKEN!

As far as engineering goes, the rendering looks great when gameplay sucks WHY? Because: No one else has any idea how that stuff works and you have free reign to do whatever you want! Such as using proper engineering methods to find the solutions to things. Shocking strategy there.
(Deep cartoon narrator voice) MEANWHILE ON THE GAMEPLAY TEAM:
Producer: Hey guys, isn’t this bitch sexy? Lets make some more changes to her.
Team: uhhh yeah… I mean no… she’s fucking ugly, so yeah lets do some plastic surgery on this ape monster.
(Deep cartoon narrator voice) THREE MONTHS LATER:
Producer: Hey guys, isn’t this bitch sexy? Lets make some more changes to her.
Team: ugh… she’s still ugly… but there’s been some improvement… I’m starting to have some faith in this horse-faced pit beast
(Deep cartoon narrator voice) ANOTHER SIX MONTHS LATER:
Producer: Hey guys, isn’t this bitch sexy? Lets make some more changes to her.
Team: Ummm…. Is she sexy? I can’t remember… I’ve… been with her so long I can’t tell anymore. But… I’ve gotten used to the way she bitches and how to get around it, and… I’m growing ever more fond of her spicey cheetos breath…
Deep cartoon narrator voice) ANOTHER SIX MONTHS LATER:
Producer: Hey guys, isn’t this bitch sexy? I know its alpha, but… lets make some more changes to her.
Team: I just don’t care anymore. I just don’t. This shit faced whale has consumed my soul. Let’s just ship this fucking thing.

I HAVE SPOKEN!

]]>
By: graham madarasz/2011/03/30/rendering-technology-whats-next/#comment-2170 graham madarasz Thu, 31 Mar 2011 12:25:48 +0000 It's a constant battle and the reason why good tools programmers are worth their weight in precious metals. I think that giving artists feedback on what will or won't work well on a given platform, even in a bulk "good" / "not good" fashion in the form of offline preview tools can go a long way to bridging that gap, but you still come back around to the problem of how to determine what needs to be done differently in the art to get past a "not good". We want to give artists tools which can foster their expression without opening the Pandora's box of generality and bringing on the late night debug sessions for the runtime team to get back to target framerate / memory budget / visual quality. It’s a constant battle and the reason why good tools programmers are worth their weight in precious metals.

I think that giving artists feedback on what will or won’t work well on a given platform, even in a bulk “good” / “not good” fashion in the form of offline preview tools can go a long way to bridging that gap, but you still come back around to the problem of how to determine what needs to be done differently in the art to get past a “not good”.

We want to give artists tools which can foster their expression without opening the Pandora’s box of generality and bringing on the late night debug sessions for the runtime team to get back to target framerate / memory budget / visual quality.

]]>
By: graham madarasz/2011/03/30/rendering-technology-whats-next/#comment-2168 graham madarasz Thu, 31 Mar 2011 12:11:17 +0000 Our central game tech group here at PopCap deals with both runtime and tools, and considers them both equally important. We don't have hard walls between them; sometimes people doing front-end tools work need to make backing framework changes, other times people working in the framework runtime change a tool to utilize new features, etc. This has been the way it's worked at every good company I've been at. Senior people do both runtime and toolchain. Tools are tech, just like runtime is tech. They go hand in hand, and both deserve equal respect, because if you sell either of them short then your project does to crap. Our central game tech group here at PopCap deals with both runtime and tools, and considers them both equally important.

We don’t have hard walls between them; sometimes people doing front-end tools work need to make backing framework changes, other times people working in the framework runtime change a tool to utilize new features, etc.

This has been the way it’s worked at every good company I’ve been at. Senior people do both runtime and toolchain. Tools are tech, just like runtime is tech. They go hand in hand, and both deserve equal respect, because if you sell either of them short then your project does to crap.

]]>
By: Joseph Simons/2011/03/30/rendering-technology-whats-next/#comment-2160 Joseph Simons Wed, 30 Mar 2011 21:17:07 +0000 <blockquote>Instead of learning how to push the technology available to us by working within its constraints, we’ll spend more time figuring out where along the ever-deepening pipeline things went awry this time.</blockquote> If you're thinking of tools as being made by engine programmers, then yes, that is a problem. Engines serve hardware; tools serve people. You can only bury yourself when your technology becomes more important than the people using it. It can be very difficult to <em>not</em> expose hard limitations of the target platform through the pipeline to artists and designers, but wouldn't it be nicer for everyone if you could?

Instead of learning how to push the technology available to us by working within its constraints, we’ll spend more time figuring out where along the ever-deepening pipeline things went awry this time.

If you’re thinking of tools as being made by engine programmers, then yes, that is a problem. Engines serve hardware; tools serve people. You can only bury yourself when your technology becomes more important than the people using it.

It can be very difficult to not expose hard limitations of the target platform through the pipeline to artists and designers, but wouldn’t it be nicer for everyone if you could?

]]>
By: Szymon Swistun/2011/03/30/rendering-technology-whats-next/#comment-2158 Szymon Swistun Wed, 30 Mar 2011 19:20:26 +0000 This is a very good point. On my previous project, our 3 person rendering engineer group worked very closely together, when one was low level optimizing SPU code, the other would help take his artist support tasks to help keep him focused, then it would switch to the other on the team when another was in need etc... We looked out for each other. Our small group together did the designs (yes we design our algorithms first before we implement), talked with the artists to make sure it would fit, implemented, hookup up the tools, supported and low level optimized it. Even with optimizations, we found it was better when the individual owning the technique would handle all platform optimization work for it, because he has more context and flexibility to change the entire algorithm to work as needed for a different platform, rather then getting a PS3 SPU guru to do it for him without that context. In the end, I feel it worked out really well with this dynamic for us. This is a very good point. On my previous project, our 3 person rendering engineer group worked very closely together, when one was low level optimizing SPU code, the other would help take his artist support tasks to help keep him focused, then it would switch to the other on the team when another was in need etc… We looked out for each other. Our small group together did the designs (yes we design our algorithms first before we implement), talked with the artists to make sure it would fit, implemented, hookup up the tools, supported and low level optimized it. Even with optimizations, we found it was better when the individual owning the technique would handle all platform optimization work for it, because he has more context and flexibility to change the entire algorithm to work as needed for a different platform, rather then getting a PS3 SPU guru to do it for him without that context. In the end, I feel it worked out really well with this dynamic for us.

]]>
By: MarcT/2011/03/30/rendering-technology-whats-next/#comment-2150 MarcT Wed, 30 Mar 2011 18:01:48 +0000 The Toolsmiths.

]]>
By: Mike Nicolella/2011/03/30/rendering-technology-whats-next/#comment-2149 Mike Nicolella Wed, 30 Mar 2011 17:38:30 +0000 I found it surprising that you were worried about tools for debugging. If you're developing tools to enable content creators to create, then one of your goals should absolutely be making it impossible to create content that causes errors, or also developing tools to help artists to debug their assets. Hopefully some of the debugging tools grow out of the need to debug the tools themselves. I expect programmers are needed more in the earlier stages to help debugging, but hopefully you've put skilled (ie non-junior) programmers on the task that quickly reach that goal and mature the tools quickly so they can move on. If your programmers are constantly spread thin needing to debug problems in a large number of tools versus developing new / improving existing tools, I think there's something wrong. I found it surprising that you were worried about tools for debugging. If you’re developing tools to enable content creators to create, then one of your goals should absolutely be making it impossible to create content that causes errors, or also developing tools to help artists to debug their assets. Hopefully some of the debugging tools grow out of the need to debug the tools themselves. I expect programmers are needed more in the earlier stages to help debugging, but hopefully you’ve put skilled (ie non-junior) programmers on the task that quickly reach that goal and mature the tools quickly so they can move on. If your programmers are constantly spread thin needing to debug problems in a large number of tools versus developing new / improving existing tools, I think there’s something wrong.

]]>
By: graham madarasz/2011/03/30/rendering-technology-whats-next/#comment-2138 graham madarasz Wed, 30 Mar 2011 14:45:12 +0000 I think you raise a good point here: "the exponentially increasing cost of game development". Here's an old concept from from Economics: extensive growth is where you increase the output of a system by increasing the raw resource - put more people on it. While intensive growth is where you increase output by increasing your productivity. Essentially, without programmers who can (and WANT to) make tools that hide the inherent complexity of modern pipelines and increase productivity you end up being cost ineffective. But guess what, tools are not as fashionable as "real" game development. When was the last time you heard someone go "I'm sick of these fancy graphics features, I fancy myself some database programming, asset juggling code and maybe doing artists productivity tools"? And while we are on it, aren't some of those skills imeediately applicable to social games - databases and web development? Makes me think... you fear might be your saviour, who knows? I think you raise a good point here: “the exponentially increasing cost of game development”. Here’s an old concept from from Economics: extensive growth is where you increase the output of a system by increasing the raw resource – put more people on it. While intensive growth is where you increase output by increasing your productivity.

Essentially, without programmers who can (and WANT to) make tools that hide the inherent complexity of modern pipelines and increase productivity you end up being cost ineffective. But guess what, tools are not as fashionable as “real” game development. When was the last time you heard someone go “I’m sick of these fancy graphics features, I fancy myself some database programming, asset juggling code and maybe doing artists productivity tools”? And while we are on it, aren’t some of those skills imeediately applicable to social games – databases and web development? Makes me think… you fear might be your saviour, who knows?

]]>
By: Sam Martin/2011/03/30/rendering-technology-whats-next/#comment-2133 Sam Martin Wed, 30 Mar 2011 12:21:25 +0000 If I am hearing you correctly, a good solution would be to focus on what's important, mainly game-play. Skill for designing and executing that great game-play comes from really hard to find designers. So I guess people need to step up and become those great designers themselves instead of waiting to find them. That is good advice, thanks for sharing. If I am hearing you correctly, a good solution would be to focus on what’s important, mainly game-play. Skill for designing and executing that great game-play comes from really hard to find designers. So I guess people need to step up and become those great designers themselves instead of waiting to find them. That is good advice, thanks for sharing.

]]>
By: snake5/2011/03/30/rendering-technology-whats-next/#comment-2129 snake5 Wed, 30 Mar 2011 09:38:42 +0000