It's a problem of economics. Its the businesses that aren't innovating for economical reasons.
1
Has the game industry reached a point of saturation?
Started by sankrant, Jul 06 2012 02:30 AM
47 replies to this topic
Ad:
#22 Moderators - Reputation: 3962
Posted 07 July 2012 - 06:59 AM
So to conclude myself, we will not be changing the basic tools of trade, but changing the methods and way 'how we do a problem'.... that sounds gold.
I would say to myself that the usage of C++ will not decline, but grow with new platforms..
We will see changes when the 'best tools for the job' change.
Right now C++ is the focus because it is the common factor between the consoles, however it's recently been reported that due to the popularity of the iOS platforms Objective-C is now the 'most popular' language out there.
Basically 'the best tool for the job' rules the roost; be it C++ for the consoles, C# & .Net for tools, Objective-C for iOS work or Python and Lua for game logic scripting. Certainly if you are working on a console or PC game you'll use at least 3 of those 4 during the development time.
#24 Moderators - Reputation: 3962
Posted 07 July 2012 - 07:55 AM
That depends on what you are talking about.
Most tools and pipelines these days are created using .Net based tools for Windows based companies.
Game logic is very often done in languages such as Lua or Python.
The core 'engine' might well be C++ but this depends on target platforms.
This is also the first time you've mentioned 'AAA' in any way; the games industry is far more than consoles and PC after all.
Most tools and pipelines these days are created using .Net based tools for Windows based companies.
Game logic is very often done in languages such as Lua or Python.
The core 'engine' might well be C++ but this depends on target platforms.
This is also the first time you've mentioned 'AAA' in any way; the games industry is far more than consoles and PC after all.
#26 Moderators - Reputation: 3962
Posted 07 July 2012 - 08:18 AM
Which are CURRENTLY console and PC based - but given Epic recently reported that their iOS based game was the most profitable they have had this might not remain the truth forever.
At one point PC games were the main area - now they play second place to console games due to profit - if companies feel they can make more money in the iOS space they will switch focus.
At one point PC games were the main area - now they play second place to console games due to profit - if companies feel they can make more money in the iOS space they will switch focus.
#27 Members - Reputation: 121
Posted 07 July 2012 - 09:33 AM
There are lots and lots of people like me who demand 'crysis', 'assasins creed', 'nfs'...... Whatever may be the platform.....
But if someday iphone hardware becomes a little more capable (which is likely), will the studios use the platform language obj-c....
Has anybody invented latency free access memory???
But if someday iphone hardware becomes a little more capable (which is likely), will the studios use the platform language obj-c....
Has anybody invented latency free access memory???
#29 Members - Reputation: 121
Posted 07 July 2012 - 06:16 PM
We went from 16 colours to 16 million which is what the human eye can distinguish. We went from 2D to 3D which is what a human mind can comprehend. I suppose a 4D game in 16 billion colours wouldn't make sense. Lol
Haha, I call that ''saturation''..... That's what I am talking about from the start.
The basic tools of trade wont' change, as per the most of the forum especially the clever answers from hodgman.
#30 Moderators - Reputation: 7472
Posted 07 July 2012 - 08:59 PM
I actually disagree.
I think the tools will change as we continue to see more cores available on all platforms. At this rate, given some of the rumblings from Intel, we may see manycore come to mobile/lightweight platforms before it becomes entrenched in the desktop space, but that's obviously just speculation.
C++98 is terrible for multicore programming. C++11 goes a ways towards making the language more robust, but IMHO there needs to be more to tap into truly heavily manycore computing. Go is an interesting attempt at moving that direction, as is Rust. I think it will be a very long time before C++ is dethroned, but it will be.
Even if we end up programming in C++22 instead of some other language, the flavor of C++ will have to change pretty radically to cater to that kind of computational demand. So even if it still bears the name C++, it's going to be a very different tool than what we use today - much like C++11 is a very different tool from C++98 in many regards.
As for this idea of "saturation" - I think that's a terribly narrow perception. There's all kinds of new opportunities opening up for all kinds of entertainment, especially games, and there will only be more in the future. Consider the potential for augmented-reality courtesy of ubiquitous high-speed, reliable wireless Internet access; or more "physically engaging" devices similar to what the Wii introduced and was carried forward by Kinect; imagine what could be done with the Google Goggles platform; and if you want to get really far-fetched, why not consider virtual reality via brain-injected nanotechnology that directly stimulates your senses?
I don't think we're saturated at all. I think we're barely getting started. The idea of the interactive simulation aka video game is only a handful of decades old; it will be a very, very long time before we run out of directions to take that.
I think the tools will change as we continue to see more cores available on all platforms. At this rate, given some of the rumblings from Intel, we may see manycore come to mobile/lightweight platforms before it becomes entrenched in the desktop space, but that's obviously just speculation.
C++98 is terrible for multicore programming. C++11 goes a ways towards making the language more robust, but IMHO there needs to be more to tap into truly heavily manycore computing. Go is an interesting attempt at moving that direction, as is Rust. I think it will be a very long time before C++ is dethroned, but it will be.
Even if we end up programming in C++22 instead of some other language, the flavor of C++ will have to change pretty radically to cater to that kind of computational demand. So even if it still bears the name C++, it's going to be a very different tool than what we use today - much like C++11 is a very different tool from C++98 in many regards.
As for this idea of "saturation" - I think that's a terribly narrow perception. There's all kinds of new opportunities opening up for all kinds of entertainment, especially games, and there will only be more in the future. Consider the potential for augmented-reality courtesy of ubiquitous high-speed, reliable wireless Internet access; or more "physically engaging" devices similar to what the Wii introduced and was carried forward by Kinect; imagine what could be done with the Google Goggles platform; and if you want to get really far-fetched, why not consider virtual reality via brain-injected nanotechnology that directly stimulates your senses?
I don't think we're saturated at all. I think we're barely getting started. The idea of the interactive simulation aka video game is only a handful of decades old; it will be a very, very long time before we run out of directions to take that.
Maker of Machinery
[Work - ArenaNet] [Epoch Language] [Scribblings] [Journal - peek into my shattered mind]
[Work - ArenaNet] [Epoch Language] [Scribblings] [Journal - peek into my shattered mind]
#31 Moderators - Reputation: 13471
Posted 07 July 2012 - 09:40 PM
The human eye can distinguish a lot more than 16 million colours (we wouldn't need HDR tone-mapping otherwise, 24-bit RGB is a very crude representation of real light...) -- as a basic test, use photoshop to paint a gradient from (0,0,0) to (0,255,0) on a high-res canvas; you should see obvious colour banding. A human rod-cell can be activated by a single photon, which I'm pretty sure your monitor can't emit reliably.We went from 16 colours to 16 million which is what the human eye can distinguish. We went from 2D to 3D which is what a human mind can comprehend. I suppose a 4D game in 16 billion colours wouldn't make sense. Lol
Yes we transitioned to real-time 3D a long time ago, which just means we've got the projection of the geometry correct for a pin-hole camera (not at all correct for a real camera or a human eye though!). We're still a long, long way from "photorealism" though, as indicated by the amount of computer graphics research being published each year.
Existing tools will continue to be used, but that doesn't mean that new tools won't appear. It may well be that John Carmack falls in love with Rust and decides to use it to make his next engine, or he might go back to using C, you'd have to ask himThe basic tools of trade wont' change, as per the most of the forum especially the clever answers from hodgman.
He might even invent his own language -- like how with the original Quake, he wasn't quite happy with C, so he first made up a new language "QuakeC" to write Quake with!
Edited by Hodgman, 07 July 2012 - 09:48 PM.
#32 Members - Reputation: 121
Posted 07 July 2012 - 11:00 PM
intel is itself investing a lot on its c++ compiler.... yeah to open a screw, we can use a knife, a spoon, a fork and a hammer too ;)... But the basic thing we have with us is a screwdriver.
If there's a doubt about saturation, then can I say that maybe we are (lim x tends to saturation = x )??
As far as changing the basic tools, will using go or rust, can change or mindset and idea??
If we are changing the basic tools, then the time period matters a lot....
Dont you believe that during the transition period the game industry would become more at rest???
If there's a doubt about saturation, then can I say that maybe we are (lim x tends to saturation = x )??
As far as changing the basic tools, will using go or rust, can change or mindset and idea??
If we are changing the basic tools, then the time period matters a lot....
Dont you believe that during the transition period the game industry would become more at rest???
#33 Moderators - Reputation: 1325
Posted 07 July 2012 - 11:18 PM
I see saturation in the sens that there are far more games released now than anyone will ever play.
In days of yore, we all pretty much played the same set of games.
There were a few thousand games in my time, and only a few hundred for the generation before me, so if you were a hard-core gamer, you played or saw almost all of them.
Now there's... millions of games?
0.o
Does this count for anything?
http://www.ford.com/electric/focuselectric/2012/
The value of assembly is so under-rated it was my ticket to six-figure salary.
The instruction set of CPU's and especially DPS's include operations that cannot be concisely expressed in C, so C compilers have no hope of selecting them.
It can take what would have been an O(n) algorithm and make it one op-code.
In days of yore, we all pretty much played the same set of games.
There were a few thousand games in my time, and only a few hundred for the generation before me, so if you were a hard-core gamer, you played or saw almost all of them.
Now there's... millions of games?
yes man, automobile industry is practically stagnent now....
0.o
Does this count for anything?
http://www.ford.com/electric/focuselectric/2012/
The value of assembly is so under-rated it was my ticket to six-figure salary.
The instruction set of CPU's and especially DPS's include operations that cannot be concisely expressed in C, so C compilers have no hope of selecting them.
It can take what would have been an O(n) algorithm and make it one op-code.
- The trade-off between price and quality does not exist in Japan. Rather, the idea that high quality brings on cost reduction is widely accepted.-- Tajima & Matsubara
#34 Moderators - Reputation: 3962
Posted 08 July 2012 - 03:29 AM
The human eye can distinguish a lot more than 16 million colours (we wouldn't need HDR tone-mapping otherwise, 24-bit RGB is a very crude representation of real light...) -- as a basic test, use photoshop to paint a gradient from (0,0,0) to (0,255,0) on a high-res canvas; you should see obvious colour banding. A human rod-cell can be activated by a single photon, which I'm pretty sure your monitor can't emit reliably.We went from 16 colours to 16 million which is what the human eye can distinguish. We went from 2D to 3D which is what a human mind can comprehend. I suppose a 4D game in 16 billion colours wouldn't make sense. Lol
Also there is the 'recent' push towards gamma correction for diffuse texutres something unheard of 10 years ago and now is practically a requirement (and a great source of arguements and bug reports as people fail to understand it, heh).
#35 Members - Reputation: 121
Posted 08 July 2012 - 04:32 AM
My primary example is warefare.
For thousands of years, warfare primarily depended on hand to hand weopons.... that was in some way 'saturation' of bladed combat weopons. Then sombody created gunpowder to fire bullets. We still use swords... in ceremonies.
Yes C++ and Company will always be there but at this day they are locking everything to saturation.
The thing is on which period we are living? *ancient?* *medieval?*????
For thousands of years, warfare primarily depended on hand to hand weopons.... that was in some way 'saturation' of bladed combat weopons. Then sombody created gunpowder to fire bullets. We still use swords... in ceremonies.
Yes C++ and Company will always be there but at this day they are locking everything to saturation.
The thing is on which period we are living? *ancient?* *medieval?*????
#36 Moderators - Reputation: 3962
Posted 08 July 2012 - 05:03 AM
You are focusing on the wrong thing; swords/bladed weapon technology continued to evolve during the time it was used even if the basic 'shape' wasn't the same.
Same thing can be applied to farming; the human race has been farming for thousands of years, but the technology and processes have improved while the 'shape' of farming has remained the same.
Same goes for software development (of which games are a subsection); while we use C++ it is not your father's C++ any more. The 'shape' might be the same but the idioms and practises have changed over the years. The industry itself has, arguably, had its 'gun powder' moment with the introduction of managed languages which caused a large shift in the way things are done - less about the details, more about the 'work'.
Even the processes we use in game development are changing; 10 years ago the idea of using more than a single thread was, by in large, unheard of. Today if you want to get performance you need to be thinking about how to efficently spread your workloads and all the comes with it (data layout and access, task granularity, batch sizes). The GPU becoming a general purpose co-processor has caused another shift in how people think about workloads and with CPUs adding GPU on-die this is becoming more important.
Things are not as 'static' as you seem to think; a great deal of thought is still being put into progressing the state of the art on various fronts. The fact that we might still use C++ is largely irrelevant in the grand scheme of things - it just happens to be the tool box which best fits the range of problems we have to solve right now taking into account platforms and performance requirements.
Put simply; the language isn't important, its what we do with it that counts.
Same thing can be applied to farming; the human race has been farming for thousands of years, but the technology and processes have improved while the 'shape' of farming has remained the same.
Same goes for software development (of which games are a subsection); while we use C++ it is not your father's C++ any more. The 'shape' might be the same but the idioms and practises have changed over the years. The industry itself has, arguably, had its 'gun powder' moment with the introduction of managed languages which caused a large shift in the way things are done - less about the details, more about the 'work'.
Even the processes we use in game development are changing; 10 years ago the idea of using more than a single thread was, by in large, unheard of. Today if you want to get performance you need to be thinking about how to efficently spread your workloads and all the comes with it (data layout and access, task granularity, batch sizes). The GPU becoming a general purpose co-processor has caused another shift in how people think about workloads and with CPUs adding GPU on-die this is becoming more important.
Things are not as 'static' as you seem to think; a great deal of thought is still being put into progressing the state of the art on various fronts. The fact that we might still use C++ is largely irrelevant in the grand scheme of things - it just happens to be the tool box which best fits the range of problems we have to solve right now taking into account platforms and performance requirements.
Put simply; the language isn't important, its what we do with it that counts.
#37 Members - Reputation: 121
Posted 08 July 2012 - 05:11 AM
yeah language isnt important.... Libraries are....
How are you going to make and succesfullu launch a rocket with an automatic(pun intended on GC) screwdriver withought any fuel???
Believe it or not.... we are still using hammer when we need a crane!!!!
How are you going to make and succesfullu launch a rocket with an automatic(pun intended on GC) screwdriver withought any fuel???
Believe it or not.... we are still using hammer when we need a crane!!!!
#38 Moderators - Reputation: 3962
Posted 08 July 2012 - 05:48 AM
Yeah, I'm done here - you are apprently so convinced that things are static and the same that you are refusing to listen to those of us working on new things which are challenging your view point.
Also your 'examples' are becoming simply stupid and I don't want to think down to your level any more.
Also your 'examples' are becoming simply stupid and I don't want to think down to your level any more.
#40 GDNet+ - Reputation: 1649
Posted 08 July 2012 - 06:47 AM
A screwdriver evolves to meet the screw. As the hardware changes so to will the software evolve to better interact with it. But the end of the day, the "idea" of what a screwdriver is remains constant regardless of it's physical evolution/differentiations, in similar fashion software will also remain similar in its "idea" whilst continuing to evolve/differentiate. But in both cases you will find that basic forms of the screwdriver as well software continue to exist and operate in today's world...simple reason why? The function they serve is still utilised. This does not mean that innovation does not occur...it simply means why spend money if we have something that is good enough for the intended purpose at this time.
