I've never witnessed this happen with any popular emulator or see why it would. Especially open source emulators like this one.
x86 and windows aren't going anywhere anytime soon. Instances of older windows applications breaking in newer windows OS are pretty rare. Even rarer when you use compatibility mode. Microsoft is pretty dumb but they're not dumb enough to break backwards compatibility with all existing windows apps for no good reason. And then we have linux builds which are even more bullet proof. And even if somehow a miracle happens and something does break compatibility with dolphin completely in all major OS it's open source so fixing it should be easy.
The odds of quantum computers being in your living room in 10 years are probably lower than you winning the lottery every single day for the rest of your life.
(03-11-2014, 07:03 AM)NaturalViolence Wrote: [ -> ]The odds of quantum computers being in your living room in 10 years are probably lower than you winning the lottery every single day for the rest of your life.
Actually, I wonder, if you treat the emergence of widespread quantum computing as a precise and short (as opposed to gradual, over the course of years or decades) event whose likelihood rises the further you go in time, and you play the lottery every day, at which point will the likelihood of you winning the lottery every single day up to that point be equal to the likelihood of quantum computers becoming widespread at that point in time?
Honestly, because IRL quantum computing isn't a chance-based event, but rather depends on how much money and time is put into research, the answer would probably be pretty absurd. Especially, provided they become widespread at all, they will do so within a reasonable amount of time - possibly 60-70 years; and if more money is put into it to accelerate research, once the initial breakthroughs are made, it could be done much faster. While the odds of you having a winning streak of lottery are astronomically low and drop exponentially. So I think that a couple of years or less would be a reasonable answer to my question - probably not 10.
Anyway, back on topic, if we extend the prediction from 10 years on from today to 20+, I think that at some point, x86 is gonna go. It just can't stay forever. The change, however, will probably be gradual, and so there will be plenty of time for Dolphin to get ported to the new system.
teh_speleegn_polease Wrote:Actually, I wonder, if you treat the emergence of widespread quantum computing as a precise and short (as opposed to gradual, over the course of years or decades) event whose likelihood rises the further you go in time, and you play the lottery every day, at which point will the likelihood of you winning the lottery every single day up to that point be equal to the likelihood of quantum computers becoming widespread at that point in time?
You're making a lot of dangerous assumption here. I don't treat it as short and precise. Any new technology is implemented gradually. Quantum computing would definitely have to be implemented gradually as well. You're also assuming that both of those things increase in probability over time when they would likely do the opposite. If on a single day you don't win the lottery the probability instantly drops to 0%. And as more and more time goes on and cost effective quantum computing continues to not show any significant progress the probability of it ever happening will go down.
teh_speleegn_polease Wrote:Honestly, because IRL quantum computing isn't a chance-based event, but rather depends on how much money and time is put into research,
This assumes that anything is possible with enough time and money. Including breaking the most fundamental laws of physics. And worse this assumes that anything no matter how impractical and expensive (and even you have to admit how impractical quantum computing is at the moment) will eventually become cheap and practical if given enough time and money. Despite countless examples of the exact opposite happening. The majority of "great ideas" like this in any industry end up getting nowhere even when billions of dollars and decades of time are thrown at it. Then eventually it's dropped when it's deemed to be impractical beyond a shadow of a doubt and some better alternative comes around. People make bad predictions about revolutionary technology all the time. It happens. People make mistakes. Even smart people. Often times this is the result of an assumption or educated guess that seemed reasonable on paper but turned out to be false once testing revealed new evidence. These happen far more often than actual breakthroughs. Science just doesn't always turn out the way people would hope. This is why big hardware companies fund lots of big projects simultaneously that attempt to fix the same problems with different solutions. Most of them end up hitting a dead end but as long as at least one succeeds it doesn't matter. Right now there are half a dozen alternatives to quantum computing that are just as promising.
No matter how much time and money they put in it could fail. And no matter how much time and money they put in it will likely never become common.
teh_speleegn_polease Wrote:the answer would probably be pretty absurd.
Your "absurd answer" is a miracle. That's what you're asking for here. You are stating that even though all current evidence points to no a miracle is going to happen at some point that completely changes everything. And you somehow know that it's inevitable that this miracle is going to happen yet you can't say what it will be or when. You have to admit that if anyone else used this same logic on you to argue that any other prototype technology was going to become common you would definitely consider them nuts. I don't know about you but when I ask someone how they're going to solve a problem and their response is "I'm sure they'll figure it out somehow in the future" that invites a lot of skepticism from me.
teh_speleegn_polease Wrote:once the initial breakthroughs are made
I think you mean "
if the initial breakthroughs are made".
teh_speleegn_polease Wrote:While the odds of you having a winning streak of lottery are astronomically low and drop exponentially.
Each day you win they increase linearly (since the number of remaining days decreases). But once you lose it drops to 0%. Either way you look at it that's not dropping exponentially.
If you'll recall we talked this through in another thread. And in that thread MaJoR brought up the portable nuclear reactors that everyone knew where going to be everywhere in the future despite nobody having ever successfully completed a working prototype or demonstrating in any way that it would be possible. People thought it was possible for the same reason. Science does such amazing things for us everyday that people assume that with enough time and money almost anything is possible or even practical. This was a good example of this since miniature nuclear reactors have many of the same problems on paper as the cooling systems for quantum computers. Problems that everyone assumed would be washed away by some miracle breakthrough in the future. Yet here we are 6 decades later with no progress in sight. And they certainly threw plenty of money at it.
teh_speleegn_polease Wrote:Anyway, back on topic, if we extend the prediction from 10 years on from today to 20+, I think that at some point, x86 is gonna go. It just can't stay forever. The change, however, will probably be gradual, and so there will be plenty of time for Dolphin to get ported to the new system.
x86 is a language. As such no matter how dramatically you change the underlying hardware it will always be possible to implement as long as binary is still used as the numeral system. There is very little reason to ever do away with it. While it might eventually be removed that would have to be in the far far future and by then emulation of x86 would be easy enough to make porting dolphin unnecessary (yes I do realize the irony of making an assumption here).
(03-11-2014, 08:09 AM)teh_speleegn_polease Wrote: [ -> ]Anyway, back on topic, if we extend the prediction from 10 years on from today to 20+, I think that at some point, x86 is gonna go. It just can't stay forever. The change, however, will probably be gradual, and so there will be plenty of time for Dolphin to get ported to the new system.
And it will be replaced with... what? A gradual change to... what? If you imagined a better version with backwards compatibility, then conceivably yes, it could happen. By why does x86 have to go? Why can't it stay forever, or at least as long as it needs to be around (which is most likely more than 20 years)? If it just
has to go away because change just
has to happen eventually, that's quite short on specifics. Some things, even in the fast-paced world of computer science, don't change, either not at all or not very often.
(03-11-2014, 12:00 PM)Shonumi Wrote: [ -> ] (03-11-2014, 08:09 AM)teh_speleegn_polease Wrote: [ -> ]Anyway, back on topic, if we extend the prediction from 10 years on from today to 20+, I think that at some point, x86 is gonna go. It just can't stay forever. The change, however, will probably be gradual, and so there will be plenty of time for Dolphin to get ported to the new system.
And it will be replaced with... what? A gradual change to... what? If you imagined a better version with backwards compatibility, then conceivably yes, it could happen. By why does x86 have to go? Why can't it stay forever, or at least as long as it needs to be around (which is most likely more than 20 years)? If it just has to go away because change just has to happen eventually, that's quite short on specifics. Some things, even in the fast-paced world of computer science, don't change, either not at all or not very often.
I'm just thinking that it would be weird if 100 years from now, all computers ran almost the same way they do today. And no, I don't have a plausible prediction about what will replace it.
NaturalViolence Wrote:You're making a lot of dangerous assumption here.
I know, and this isn't meant to be a realistic way of predicting when quantum computing will emerge (especially since, as you said, it will be a gradual event).
NaturalViolence Wrote:No matter how much time and money they put in it could fail. And no matter how much time and money they put in it will likely never become common.
Well, I guess I should have put this clearer:
teh_speleegn_polease Wrote:provided they become widespread at all,
I am not assuming that quantum computing WILL inevitably happen no matter what. I am only assuming that for the purpose of the discussion of what will happen IF it does emerge, not to have to litter every sentence with phrases like the one I just quoted. I guess I should have made that clearer.
NaturalViolence Wrote:Your "absurd answer" is a miracle.
About as much of a miracle as winning the lottery every day for 2 years. Except that in theory, winning the lottery for 2 years is possible, while quantum computing emerging in 2 years is virtually impossible, but in practice I do believe they are both about as likely.
NaturalViolence Wrote:I think you mean "if the initial breakthroughs are made".
Yeah.
NaturalViolence Wrote:Each day you win they increase linearly (since the number of remaining days decreases). But once you lose it drops to 0%. Either way you look at it that's not dropping exponentially.
But look at it this way: say you have a 1/1,000,000 chance of winning. The chance of winning for two days in a row is 1/1,000,000,000,000. The chance of winning three days in a row is 1/1*10^-18. And so on. That's an exponential drop.
NaturalViolence Wrote:x86 is a language. As such no matter how dramatically you change the underlying hardware it will always be possible to implement as long as binary is still used as the numeral system.
Good point. I wonder if computers working in tertiary will ever be more than a strange curiosity though. Probably not... =(
A lot of SSD NAND works in ternary, and some is being developed which uses base 4. It is converted to binary before any work is done on it, though.
(03-11-2014, 09:53 PM)teh_speleegn_polease Wrote: [ -> ]I'm just thinking that it would be weird if 100 years from now, all computers ran almost the same way they do today.
But why? If you assume that things always change no matter what, then sure it seems weird, but change is not always a given. It depends. Perhaps we simply found something (x86) that can effectively be used for centuries before it quits being useful to us. Perhaps it won't suit our needs past 2050. We don't know precisely when change like that will happen though, or if it will happen at all.
(03-12-2014, 01:45 AM)AnyOldName3 Wrote: [ -> ]A lot of SSD NAND works in ternary, and some is being developed which uses base 4. It is converted to binary before any work is done on it, though.
Of course such ssds are really awful
(03-10-2014, 06:52 AM)RachelB Wrote: [ -> ]Dolphin has a generic build that should work on any hardware, although with only interpreter and software renderer. The JIT, and hardware graphics backends would need to be remade if future hardware broke compatibility, but otherwise, i think everything else in dolphin is safe.
I don't think we will fall back to the software renderer at some time. Both d3d and ogl did a huge cleanup in the last 5 years and we only use the new API. So as long as GPU are build for polygon based rendering (eg with in order framebuffer writes), dolphin should be supported. And still then, llvmpipe is much faster then our software renderer.
But x86 is doomed to fail, at least I hope so. For performance, we would have to rewrite the JIT. Good news, dolphin already supports ARM7. There was also a try to use LLVM for code generation, but this was by far too much overhead.