Maximum Strength Placebo(tm) may cause constipation, diarrea, hot flashes, cold flashes, drowsiness, and problems getting to sleep. Please consult your doctor before self-medicating.
I also believe that this is the explanation for why Opera is so popular in some areas of the world. Its just flat out more efficient where it really matters.
As far as the memory usage tests they did in the article.. thats always horseshit. If I have 1 gig of free memory before I run my browser, and my browser can better itself by using it all.. then it should! Opera has used a memory cache for web pages and so forth for a very long time and I suspect that most other browsers now do as well.
The real test, as you say, is how they deal with low memory situations. I remember how netscape completely and utterly sucked to the point of being unusable on a 4MB windows 3.1 system, while the same era Opera and even Internet Explorer were actually usable. This was back when 4MB systems were still quite common in the world.
Soon, the only drug we will need in Placebo(tm). This is to be expected since it has appeared in more clinical trials for more ailments than any other drug in history.
You are talking about 'cheaper' in regards to supercomputers? Their super computer has 68 of these, which are so expensive they wont even give you a price tag without calling them for a quote.
So we'll just have to make them boneless too. That way they wont walk off. Zap their muscles regularly to flex them (why not, they cant feel pain.) Stuff a tube down their throat to feed them, and BANG! you've got just about the perfect food source.
I've gotten sick of this "its just a design patent" garbage.
How many permutations of the same number of elements are there, and how many of those are actually reasonable?
Google now owns the idea of having a giant search box in the middle of the page, with two big buttons underneath and several small links nearby.
The search box can be large or small.
The buttons can be big or small, as well as above, below, to the left, or right.
Links can be big or small, nearby or far.
So pretty much only 64 permutations, assuming that none of them will be ruled as infringing another?
The problem is that design patents seem to be covering *functional* elements.
"The design of your coffee maker is infringing on Mr Coffee's Upright Container design patents."
I began using Opera back when I had Windows 3.1 on a 4MB 386/40 system with a 32MB hardcard.
Netscape was complete useless on this system. Unimaginably slow.
Opera was not only fast, it fit on a floppy.
I believe that these two things are the main reason that Opera has had such great momentum in the old soviet republic. It was the only thing that worked for many people for a very long time, and since its easily as good as any other browser out there.. no reason to switch.
..because Apple doesnt have a track record for selling devices that explode. They certainly didn't recall 1.8 million iBook and POwerBook batteries in 2006. Definately not.
Are you really skeptical that the NOAA is "scientific"?
After reading the GP's reply to this, yes. Thanks for prodding him into smacking your presumptive ignorance down.
Science is a methodology, not on organization acronym. When the NOAA actualy performs science, thats great! When they don't, but people like you still believe them hook line and sinker, that fucking sucks for everybody.
And how many MB can you address with a 32-bit pointer under the IEEE recommendations? I bet you can't work it out in your head.
I don't know if the person you replied to can, but I sure can. Thats easily calculated. A Mega is 2^20, and (32 - 20) = 12, and since I know 2^10 is a Kilo (1024), then we have (12 - 10) = 2 so its trivialy 2^2 KiloMegaBytes, or 4 KiloMegaBytes, or 4096 MegaBytes.
2^10 = Kilo
2^20 = Mega
2^30 = Giga
2^40 = Tera
It really is a simple system to those who have been initiated. If you know the powers of two from 1 to 10, and the decimal 2-power prefixes, you can always reduce.
I still think its stupid to force it on the general populace.
I am a determinist because I believe that everything that ever happened and everything that ever will happen can be explained by particles following one simple rule: The path of least resistance.
The fact is that in following that path, my brain insists on convincing you of my wisdom. It may very well be that in following that path, your brain begins to believe that I am in fact extremely wise.
Well it seems to me that they have already had more than a few years (stream format hasnt changed since 2004) to improve Theora (which itself was derived from a mature codec) so it is highly likely that they are already well on their way up their own diminishing returns curve.
Theora is lacking support for some techniques that other encoders take advantage of. I think B-frames is probably a significant one (looks to me like B-frames *double* the amount of context the compressor can use)
But due to the nature of floating point arithmetic, the last few digits of a floating point operation will vary according to processor architecture, so cross-platform play requires code and bandwidth overhead to synchronize all those game variables.
If the architecture follows the IEEE standard, then this is not true. I don't know of any architecture that doesnt follow it for 32-bit and 64-bit floats.
The problem is that programmers tell their compiler to screw standards, and optimize as much as possible. This causes two things to happen:
First is that on 32-bit Intel architectures, they use the 80-bit floating point stack exclusively without rounding down to 32 or 64 bits after every operation (something they would otherwise have to do to be considered following the standard.)
The second thing is that the compiler rearranges the order of operations and in some cases refactors them. For example, it might change (a + b) + c into a + (b + c), in addition it might change a * b + a * c into a * (b + c) or vise-versa. This is the real diverger. Compiler A and compiler B will not reorder or refactor them in the same way, even on the same platform. As you may or may not know, the order of operations is crucial in floating point work when trying to maximize accuracy.
But this is besides the point really.. (most, if not all) compilers can be instructed to NOT perform these optimizations and to rigidly adhere to IEEE standards.
Slashdot Mirror
Your "rights" over your own works should END the moment you view them as NOT WORTH PUBLISHING ANYMORE.
Who says I view them as not worth publishing 'anymore'?
Maybe I just view them as not worth publishing this year.
My book, my rights, my marketing strategy. Fuck off, thief.
Are you saying that my rights over my own works is directly linked to Googles financial statements?
Yes, but at least its a private version of copyright law that applies to what they themselves own.
Google is being given a private version of copyright law that applies to what they don't own.
Maximum Strength Placebo(tm) may cause constipation, diarrea, hot flashes, cold flashes, drowsiness, and problems getting to sleep. Please consult your doctor before self-medicating.
I also believe that this is the explanation for why Opera is so popular in some areas of the world. Its just flat out more efficient where it really matters.
.. thats always horseshit. If I have 1 gig of free memory before I run my browser, and my browser can better itself by using it all.. then it should! Opera has used a memory cache for web pages and so forth for a very long time and I suspect that most other browsers now do as well.
As far as the memory usage tests they did in the article
The real test, as you say, is how they deal with low memory situations. I remember how netscape completely and utterly sucked to the point of being unusable on a 4MB windows 3.1 system, while the same era Opera and even Internet Explorer were actually usable. This was back when 4MB systems were still quite common in the world.
Soon, the only drug we will need in Placebo(tm). This is to be expected since it has appeared in more clinical trials for more ailments than any other drug in history.
You are talking about 'cheaper' in regards to supercomputers? Their super computer has 68 of these, which are so expensive they wont even give you a price tag without calling them for a quote.
You've been out of the loop for awhile I see.
A 250GB SSD can have over 2.5PB (petabytes) written to it before it cannot be written to anymore.
Prude!
So we'll just have to make them boneless too. That way they wont walk off. Zap their muscles regularly to flex them (why not, they cant feel pain.) Stuff a tube down their throat to feed them, and BANG! you've got just about the perfect food source.
How many permutations of the same number of elements are there, and how many of those are actually reasonable?
Google now owns the idea of having a giant search box in the middle of the page, with two big buttons underneath and several small links nearby.
The search box can be large or small.
The buttons can be big or small, as well as above, below, to the left, or right.
Links can be big or small, nearby or far.
So pretty much only 64 permutations, assuming that none of them will be ruled as infringing another?
The problem is that design patents seem to be covering *functional* elements.
"The design of your coffee maker is infringing on Mr Coffee's Upright Container design patents."
I began using Opera back when I had Windows 3.1 on a 4MB 386/40 system with a 32MB hardcard.
Netscape was complete useless on this system. Unimaginably slow.
Opera was not only fast, it fit on a floppy.
I believe that these two things are the main reason that Opera has had such great momentum in the old soviet republic. It was the only thing that worked for many people for a very long time, and since its easily as good as any other browser out there.. no reason to switch.
...so if it renders at 30fps or more on my computer, that means its good to go on yours?
What browser doesnt have problems rendering slashdot?
Really... I'd like to know. Firefox, Opera, Safari, Chrome, and Internet Explorer all have issues.
..because Apple doesnt have a track record for selling devices that explode. They certainly didn't recall 1.8 million iBook and POwerBook batteries in 2006. Definately not.
Are you really skeptical that the NOAA is "scientific"?
After reading the GP's reply to this, yes. Thanks for prodding him into smacking your presumptive ignorance down.
Science is a methodology, not on organization acronym. When the NOAA actualy performs science, thats great! When they don't, but people like you still believe them hook line and sinker, that fucking sucks for everybody.
And how many MB can you address with a 32-bit pointer under the IEEE recommendations? I bet you can't work it out in your head.
I don't know if the person you replied to can, but I sure can. Thats easily calculated. A Mega is 2^20, and (32 - 20) = 12, and since I know 2^10 is a Kilo (1024), then we have (12 - 10) = 2 so its trivialy 2^2 KiloMegaBytes, or 4 KiloMegaBytes, or 4096 MegaBytes.
2^10 = Kilo
2^20 = Mega
2^30 = Giga
2^40 = Tera
It really is a simple system to those who have been initiated. If you know the powers of two from 1 to 10, and the decimal 2-power prefixes, you can always reduce.
I still think its stupid to force it on the general populace.
What the fuck is a ScrumMaster? What the fuck is this person asking? Seriously, how did this get in my RSS reader?
A ScrumMaster is a person that broadcasts RSS clips in such a way as to raise awareness for the stupidity of RSS feeds.
And wasn't it only two years ago the NVIDIA repeatedly stated that RTRT was stupid?
..and then a year ago that they began saying that they could do it just as well as Intel?
Don't get me wrong. I love nVidia cards. I just don't believe anything they say.
How could life ever have existed in such a desolate place?
Come on.. Denmark isnt THAT desolate!
No.
I am a determinist because I believe that everything that ever happened and everything that ever will happen can be explained by particles following one simple rule: The path of least resistance.
The fact is that in following that path, my brain insists on convincing you of my wisdom. It may very well be that in following that path, your brain begins to believe that I am in fact extremely wise.
That this is inevitable, I cannot say.
Just performed a statistical analysis, using the same above strategy..
:/
errors per million iteratios = 352566
Thats about 35% of the time
Yes you do. Not always, but yes.
I just wrote a quick VB.NET program to search for an example in 32-bit IEEE precision randomly... it found one in only 8 iterations!
The output:
iters = 8
a = 0.8480518
x = 0.9919022
y = 0.0326252
a*x+a*y = 0.8688523
a*(x+y) = 0.8688522
The code:
Sub Main()
Dim a, x, y, sum1, sum2 As Single
Dim rng As New Random(0)
Dim iters As Integer = 0
Do
a = rng.NextDouble()
x = rng.NextDouble()
y = rng.NextDouble()
sum1 = a * x + a * y
sum2 = a * (x + y)
iters += 1
Loop Until sum1 <> sum2
Console.WriteLine("iters = " & iters)
Console.WriteLine("a = " & a)
Console.WriteLine("x = " & x)
Console.WriteLine("y = " & y)
Console.WriteLine("a*x+a*y = " & sum1)
Console.WriteLine("a*(x+y) = " & sum2)
Console.ReadKey()
End Sub
Well it seems to me that they have already had more than a few years (stream format hasnt changed since 2004) to improve Theora (which itself was derived from a mature codec) so it is highly likely that they are already well on their way up their own diminishing returns curve.
Theora is lacking support for some techniques that other encoders take advantage of. I think B-frames is probably a significant one (looks to me like B-frames *double* the amount of context the compressor can use)
But due to the nature of floating point arithmetic, the last few digits of a floating point operation will vary according to processor architecture, so cross-platform play requires code and bandwidth overhead to synchronize all those game variables.
If the architecture follows the IEEE standard, then this is not true. I don't know of any architecture that doesnt follow it for 32-bit and 64-bit floats.
The problem is that programmers tell their compiler to screw standards, and optimize as much as possible. This causes two things to happen:
First is that on 32-bit Intel architectures, they use the 80-bit floating point stack exclusively without rounding down to 32 or 64 bits after every operation (something they would otherwise have to do to be considered following the standard.)
The second thing is that the compiler rearranges the order of operations and in some cases refactors them. For example, it might change (a + b) + c into a + (b + c), in addition it might change a * b + a * c into a * (b + c) or vise-versa. This is the real diverger. Compiler A and compiler B will not reorder or refactor them in the same way, even on the same platform. As you may or may not know, the order of operations is crucial in floating point work when trying to maximize accuracy.
But this is besides the point really.. (most, if not all) compilers can be instructed to NOT perform these optimizations and to rigidly adhere to IEEE standards.