kingdorian

The 290 will be discontinued, in favor of the 390 all that's left is whatever stock of 290s that are left.

"Never" and "always" are so final. Integral values can be represented exactly in IEEE format (up to the precision of the mantissa), so you could store bank balances in cents (a.k.a "fixed point" arithmetic).  But in general, comparisons are fine as long as you understand what is going on. I've witnessed so many otherwise smart programmers say really stupid things about floating-point (examples: "there is no floating point compare instruction on the CPU" (it's called FCOM by the way), or "floating point numbers are not exact, so they can't be compared"). Is it just math anxiety? I think the problem is conflating computer arithmetic with real world arithmetic. For me, it helps to view computer arithmetic as just a bit-twiddling API independent of actual mathematics. In this API view you can compare values for equality, you can get deterministic results, and you can successfully write numerical code. You just have to understand what the functions do - as you would with any other API. This is definitely an area needing more attention.  David Eberly was working on a book covering all aspects of computer arithmetic a few years ago, but his publisher canceled the project fearing there wouldn't be enough demand.  How sad.

Ik moest lachen.
Gekke mensen op Marktplaats met hun lage boden

OK. Good looking out. It's a start. Here's a quick rundown of how a floating point value is represented with IEEE 754 (let's see single precision, double is similar). I'll try not to be confusing.
A single precision floating point is 4 bytes, which makes it a 32-bit representation. The general representation of a floating point is scientific notation in the form of (-1)signal * 1.mantissa * 2exp.
Remember that a bit may hold 0 or 1.
 
The bit distribution is as follows:
[s][eeeeeeee][mmmmmmmmmmmmmmmmmmmmmmm]
 
[s] - Represents the signal of the number (0 is positive, 1 is negative). 1 bit
[eeeeeeee] - Exponent portion of the number. This is calculated with a -127 excess (we'll see what this means below). 8 bits (can represent a number as high as 28-1 = 255)
[mmmmmmmmmmmmmmmmmmmmmmm] - The mantissa portion of the number. This is the fractional part of your real number. 23 bits
 
Let's take a look at an example:
[0] [01101000] [10101010100001101000010]
 
First of all, it is clear that this is a positive number (the signal bit is 0).
Next up, the exponent has a value of 01101000, which means that it represents 104 in decimal1. The reason for the -127 excess, is so that you can represent numbers in [-127,128]. This way it is possible to represent negative and positive exponents (for really big and really small numbers).
So this exponent is in fact 104-127 = -23.
 
Finally, the mantissa (the troublemaker). Conversion is (naturally) done this way. I assume you know your math and understand why we're using negative exponents. :P
Right to left, -1 to -n. So, in this case:
1*2-1 + 0*2-2 + .... + 0*2-22 = 0.666115.
 
As I mentioned previously, the general form of this standard is (-1)signal * 1.mantissa * 2exp. So my example number effectively represents 1.666115 * 2-23, which is roughly 1.986 * 10-7.
That is a simple explanation of how you would convert FP to decimal. Now, consider the following 2 representations:
 
0 01111010 01001111110111110011110 = 0.041 (this is the result of assigning this value to a variable)
+       -5                      0.312
 
0 01111010 01001111110111110011101 = 1.3118778 * 2-5 = 0.04099618125 ~= 0.041 (this is the result of adding 0.001 to a variable with 0.04)
+       -5                     0.3118778
 
The reason why this approximation is done, is pretty well explained in point 3 of that link you found. The final mantissa bits are approximations, rather than precise numerics (the only numbers that you can represent precisely are numbers which can be represented by fractions with a 2n denominator).
Now, if you are curious and want to understand why that's the result of this sum, go check out how to calculate floating point arithmetic. Also, a quick Google search brought this up.
 
Good luck! 
 
TL;DR: Never use floating point values for arithmetic when you'll need to compare them. This is why any (decent) programming teacher will tell you never to use floating point variables to handle things like bank balances. Most of them don't know why, but it's still good advice. :)
 
________________________________________________________________________________
1 - If you don't know how to convert from binary to decimal, a simple way is to sum powers of 2 from left to right, counting only the ones with a 1. So the first bit is worth 1, the next is 2, then 4, etc. 
So in this case 01101000 is 8+32+64 = 104. The general form is, in this case, 011010002 = 0*20 + 0*21 + 0*22 + 1*23 + 0*24 + 1*25 + 1*26 + 0*27 = 0+0+0+8+0+32+64+0 = 10410

10 PRINT "Penus "20 GOTO 10

I'll just leave this here, berkel is even more awesome then I thought he already was...
 
https://www.youtube.com/watch?v=TJDKwPYD4XA

I'll just leave this here, berkel is even more awesome then I thought he already was...
 
https://www.youtube.com/watch?v=TJDKwPYD4XA

I'll just leave this here, berkel is even more awesome then I thought he already was...
 
https://www.youtube.com/watch?v=TJDKwPYD4XA

People were hating for the sake of hating, further reinforcing my belief that gamers spend more time bitching than they do gaming. 
 
Also, lol at the "this is not a repost". Whoever calls repost on you is full of it. Quality posts are quality, keep up the good work. 

Basically, with heat exchangers (i.e. radiators), the efficiency is improved if you make both fluids (in this case air and water) flow in opposite ways. In this particular case, that would mean that you connect your rads in series and make the side where the air exhausts the water input and make the side where the air goes into the radiator the water output. This will make you achieve the most efficient use of your radiators.

there gos three and a half months worth of my data cap

r9 290 fits your budget perfectly, the 970 is also an option if you don't mind going a little bit over your budget. But may I ask you why "What nvidia gpu?" Fanboyism?

People mention AMD because a lot of people come on here loaded with misinformation because of Nvidias superior marketing machine. Google any AMD card vs Nvidia card and whats the first thing that pops up? GPUBOSS, which is an Nvidia marketing site disguised as a review site.
 
That's not to say Nvidia wouldnt better suit his needs if say he has a micro ITX case and requires a blower style cooler, but more often than not when people insist on Nvidia its because they are loaded with misinformation and marketing bullshit with little basis in reality.
 
Its painful to watch someone buy a gtx 960 would they could have had an r9 290x for almost the same price.

the msi r9 290 is like 300
 
this one the sapphire trix
 
http://www.newegg.com/Product/Product.aspx?Item=N82E16814202143&cm_re=r9_290-_-14-202-143-_-Product

They should do a Channelsuperfun video, my boss/employee does my makeup challenge!

They should do a Channelsuperfun video, my boss/employee does my makeup challenge!

6 foot 2, 73 kilos, I'd give me a 6/10 for looks
 
Here's a pic of me and my vacuum
 

He's probably interested in dressing up as a girl/woman.Kappa

I'll just leave this here, berkel is even more awesome then I thought he already was...
 
https://www.youtube.com/watch?v=TJDKwPYD4XA

I'll just leave this here, berkel is even more awesome then I thought he already was...
 
https://www.youtube.com/watch?v=TJDKwPYD4XA

berkl scored. 

Ok... I'm gonna ask it either way.
HOW DID YOU FIND THIS?!

So you tought this project was dead, huh? Jokes on you!
In 2 weeks I'm gonna have a marvelous week :3
Also: First Post updated

i would save my data on a usb key or HDD external or not. its always better to start with a fresh installation. imho.

that's cool for.. basically nothing