What is $FFFF ?
Blitz3D Forums/Blitz3D Beginners Area/What is $FFFF ?
| ||
| Does anyone know where I can find more information about the keys $A through $F. Are these used for binary math? I can't figure out how they multiply by adding a letter. example... $F = 15 , $FF = 255 , $FFF = 4095, but how $F*$F <> 255 $F*$F*$F <> 4095 So how do they step? Thanks for any help. |
| ||
| This is not binary math, it's hexadecimal math ;-) This means that we use 16 "digits". Instead of counting from 0 to 9, we count from 0 to 15, using letters for digits greater than 9. So: A = 10, B = 11, C = 12, D = 13, E = 14, F = 15 Going from right to left, multiply the second digit by 16, the third by 256 (16 * 16), the fourth by 4096 (16 * 16 * 16) and so on... So for example: $ABCD = ($A * 4096) + ($B * 256) + ($C * 16) + $D = (10 * 4096) + (11 * 256) + (12 * 16) + 13 = 43981 Well, at least I can understand myself ;-). Hope this helps. Or you can google for "hex maths" ;-) |
| ||
| Google for "hexadecimal notation" |
| ||
| This is not binary math, it's hexadecimal math ;-) This means that we use 16 "digits". Hooray, someone who knows Hexadecimal! I thought I was the only person! On the subject of Binary, I love this quote: "There are only 10 types of people in this world. Those who understand Binary, and those who don't." |
| ||
| As soon as you start messing with colors or using the readpixel writepixel functions you will learn everything there is to know about hexadecimals very quickly! :O) |
| ||
| Hooray, someone who knows Hexadecimal! I thought I was the only person! ??? You need to get out more. Or at least out with more geeks. |
| ||
| Hey, I understand hex too! LOL! Adam, I love that quote! "There are only 10 types of people in this world. Those who understand Binary, and those who don't." |
| ||
| "There are only 10 types of people in this world. Those who understand Binary, and those who don't." There's also its counterpart: "There are only 11 types of people in this world. Those who understand Binary, and those who don't." |
| ||
| There's also "There are only 13 types of people in this world. Those who understand Binary, and those who don't. Er what's Binary? IPete2. |
| ||
| Yes, but remember if you like bees: DEAD+BEE = 1+BAD+DEED Bees make HEX-agonal nests! :O) |
| ||
| IPete2: Binary is a number system based around two digits: 0 and 1. These digits, on their own, are integral to computers: it's how they "think." 1 represents on, true, yes, etc, and 0 represents off, false, no, and so on. Believe it or not, these two numbers are what allow computers to perform complex mathematical functions such as rendering 3D graphics. Long live technology! |
| ||
| Er, I think IPete2 may just have been joking! ;) |
| ||
| I know hex. I understand There are only 10 types of people... but whats behind There are only 11 types of people... |
| ||
| But what's behind 'There are only 11 types of people... that would be those who don't understand binary. :-? |
| ||
| There are only two types of people: Those who understand jokes about binary, and those who do not. :o) |
| ||
| There is only 2 people in the world, Those who know what $FFFF means and those who don't give a $FFFF. |
| ||
| Well I am both, but I still don't get how you can just plug the sequince rgb=ReadPixelFast(x,y) And $FFFFFF. Not to mention the part about putting And after a statement, I thought that it was for conditional checks you know like If $FFFF and (binary <> Question) then ect you know the rest. I know binary. I saw it on TV once. You got eggs that fill containers (shaped like eggs) and each container holds a byte (and salt and pepper). |
| ||
| There's two versions of And in Blitz. The logical one that is typically used in If..Then statements (which you talk about), and also the bitwise version that is used in maths (such as in ReadPixel, etc). The problem with Blitz3D is that they both look the same and Blitz3D decides which to use depending on the context in which it is used. However, the thing to remember is that even in those If..Then statements is that each condition is evaluated to True or False anyway (1 or 0) so a bitwise check is ultimately done to determine whether the condition has been met. Most other languages actually make a distinction between the two to make it clearer what you want. For example, BlitzMax has And for logical operations and & for the bitwise version. |
| ||
| Don't people do truth tables for logic gates at school anymore? How you can not understand a bitwise AND operation and still program I don't know. . . :O) |
| ||
| JazzieB: I'm pretty sure Blitz3D only supports And as a bitwise operator. To use it as a logical operator, you have to something like: If (val1 <> 0) And (val2 <> 0) Then ... Haven't used B3D in ages so I may be off, here - bit rusty. :) |
| ||
| Blitz3D only supports And as a bitwise operator. ... If (val1 <> 0) And (val2 <> 0) Then ... Hmmm. interesting. . . but it makes sense because there's no way for a compiler to tell if you want bitwise or normal boolean based on context (because there is no boolean data type) I also assumed Blitz3d evaluates any non-zero result as true. If both of these are true then a lot of my code may be iffy. I guess I may have been lucky that my code tends not to generate (except when I intend to perform a bitwise) occasions where 1010 (10) is being anded with 0101 (5) etc, because I would have gotten false when I would have expected True! I'm not at a PC with blitz installed right now, but I will check into this tonight to be sure. |
| ||
| I'm not at a PC with blitz installed right now Reminds me, I wish Blitz would publish the Blitz3d language reference as well as the command reference section in the online manual. |
| ||
| If both of these are true then a lot of my code may be iffy. I guess I may have been lucky that my code tends not to generate (except when I intend to perform a bitwise) occasions where 1010 (10) is being anded with 0101 (5) etc, because I would have gotten false when I would have expected True! Yep. If you're not aware of this, it can lead to some seemingly invisible bugs in your code logic. Nasty. :) |
| ||
| Darn. . . now I need to check a few million lines of code! Though, so far I don't think I've experienced this because whenever I'm intending to use a boolean I always use just values of 1 and 0, or the keywords True and False (which are the same thing.) I guess it's REALLY important to know this though - everyone take heed! :O) |
| ||
| JazzieB: I'm pretty sure Blitz3D only supports And as a bitwise operator. To use it as a logical operator, you have to something like: Possibly, but then all conditions are evaluated to True (1) or False (0) anyway, so ultimately it's all bitwise. I've not found brackets necessary unless you have some complicated conditions to be met. In your example, if both val1 and val2 were non zero, Blitz would be doing a bitwise 1 and 1, which is 1 (True). If val1 was 0, then it would be 0 and 1, which is 0 (False).If (val1 <> 0) And (val2 <> 0) Then ... Haven't used B3D in ages so I may be off, here - bit rusty. :) I also assumed Blitz3d evaluates any non-zero result as true. True. And this also includes negative numbers. Only 0 is False. |
| ||
| example... $F = 15 , $FF = 255 , $FFF = 4095, but how $F*$F <> 255 Last time I checked, 15 * 15 == 225, not 255. Kinda like how 9 * 9 == 81, not 99. rgb=ReadPixelFast(x,y) And $FFFFFF When you And, you get only the bits back that you have selected. The above code gets rid of anything not in the lowest 24 bits - presumably the alpha channel in $FF000000. |
| ||
| In your example, if both val1 and val2 were non zero, Blitz would be doing a bitwise 1 and 1, which is 1 (True) Try this: val1% = 10 val2% = 5 Print ((val1) And (val2)) Print ((val1<>0) And (val2<>0)) WaitKey End and you'll see the difference! |
| ||
| I was just wondering which of ( <>, And ) were evaluated first by default. The following expression evaluates to 0: (1<>2 And 2) Therefore <> must evaluate before And because (2 And 2) = 2 so the expression would be equivalent to 1<>2 if and evaluated first and would give the value True (1). This means that Jazzieb is right in that you don't need the brackets in If (val1 <> 0) And (val2 <> 0) Then But you do need to still include the <>0 parts to be sure to get a single bit boolean And between val1 and val2. |
| ||
| I included the brackets for clarity. ;) |
| ||
| oh, btw I have probably steered this way off topic - sorry! If you are familiar with binary but not hex, the easiest way to think of hex is as a simple encoding of a binary number where: Hex = Binary 0 = 0000 1 = 0001 2 = 0010 3 = 0011 4 = 0100 5 = 0101 6 = 0110 7 = 0111 8 = 1000 9 = 1001 A = 1010 B = 1011 C = 1100 D = 1101 E = 1110 F = 1111 you can therefore easily convert from one to the other. For example, A7 in hex is the same as 1010 0111 in binary. |
| ||
| I included the brackets for clarity. Good idea big10p! It's generally best to include brackets if there is any doubt because then the code is easier to understand for people that aren't familiar with the conventions of the specific language. I try to use brackets often because I program in quite a few different languages and I can never remember (nor do I want to even know or think about) the priority of operations in a given language. Having said that, generally most languages use the same or similar priorities but there will always be that exception that kills you! |
| ||
| Try this: val1% = 10 val2% = 5 Print ((val1) And (val2)) Print ((val1<>0) And (val2<>0)) WaitKey End and you'll see the difference! I already know the difference, I'm no newbie! The point I was trying to get across was that conditions are evaluated to True (1) or False (0), so bitwsie ANDing, ORing and whatnot are happening in the background to determine if a condition is true or not. The first Print statement only has the values of those variables, so a bitwise And is done. The second you are comparing them with other values first, giving an outcome of either 1 or 0 for each, but a bitwise And is still done after getting those results - it just looks logical. |
| ||
| Base 10(decimal) is 0123456789 Base 16(hexadecimal) is 0123456789ABCDF In hex, one character= the less herd of NYBBLE. Ther are 4 bits in a nibble and 2 nibbles in a BYTE(or 8 bits) a collection of bytes is called a WORD FF is an 8 bit word (2 nibbles or 1 byte) which is (16x16=256)-1(for zero)=255 (256 possible numbers inc. 0) FFFF is a 16 bit standard word made up of 2 bytes therefor (16x16x16x16 or 256x256=65536)-1(for zero)=65535 |
| ||
| 2 bytes = word 4 bytes = longword At least it was back in the day when I was programming in 68000. Terminology may have changed over the years, so fair enough if that's the case. Just a quick note, Blitz doesn't refer to multiple byte values as words or longwords, instead they are... 2 bytes = short 4 bytes = int(eger) |
| ||
| correct, word, longword, quadword, octaword.... but I think most people nowadays just say 8/12/16/24/32/64 etc. 'bit word'. an integer in blitz is (i think!) a signed 32bit word . (or 4 bytes as u say with the first bit(or is it last?) specifying +/-) LOL, its been way to long since archetecture 101 :) |
| ||
| but I think most people nowadays just say 8/12/16/24/32/64 etc. 'bit word'. Twelve isn't usually part of that sequence... Is it? |
| ||
| xlsior, There are 10 [2] types of people in this world. Those who understand binary, and those who don't ........ There are 11 [3] types of people in this world. Those who understand binary, and those who don't One joke seems to repeat the other joke! |
| ||
| One joke seems to repeat the other joke! Yes, but the first joke is from the perspective of someone who clearly understands binary - and thus reads '10' as two instead of ten. The second is from the perspective of someone who does not understand binary (to whatever degree), since the statement makes no sense using the binary number they did. Two sides of the same coin. Not counting the edge, of course. :o) |
| ||
| "twelve isn't usually part of that sequence... Is it?" Not on conventional computer hardware, no. Some other machines like cameras, DACs, sound devices etc use so called '12 bit' words for whatever reason |
| ||
| rgb=ReadPixelFast(x,y) And $FFFFFF I'm sorry but I am still not sure what's going on in this statement. How you can asign rgb a value of ReadPixelFast(x,y) And 16777215, or is that $FFFFFF, read as 255,255,255. In JFK's function here there seams to me no reason for And $FFFFFF, acept that it won't work without it. Function SetMaskChannel(tex) ;this function will set the alpha channel of black pixels to be fully transparent ;(as required by masked textures using flag 4). Local w,h,rgb,r,g,b,x,y,x2,y2,count,average_rgb w=TextureWidth(tex) h=TextureHeight(tex) Dim tex_memory(w,h) SetBuffer TextureBuffer(tex) LockBuffer() For y=0 To h-1 For x=0 To w-1 rgb=ReadPixelFast(x,y) And $FFFFFF If rgb=0 Then WritePixelFast x,y,0 Else EndIf Next Next UnlockBuffer() SetBuffer BackBuffer() End Function |
| ||
| Readpixel will return a 32 bit value. When you mask it, using "AND $FFFFFF"or "AND $00FFFFFF" you will set the most upper 8 Bits to zero. This is the alpha channel. What's left is the 24 bit RGB value of the pixel or texel. You have to see the 32 bits as this: $AARRGGBB (alpha red green blue) where each channel has 8 bits or one byte, a value from 0 to 255 aka from $00 to $FF. In this function I didn't care about the loaded alpha channel data, but used to create a new one based on RGB. |
| ||
| rgb=ReadPixelFast(x,y) And $FFFFFF I'm sorry but I am still not sure what's going on in this statement. How you can asign rgb a value of ReadPixelFast(x,y) And 16777215, or is that $FFFFFF, read as 255,255,255. In JFK's function here there seams to me no reason for And $FFFFFF, except that it won't work without it. $FFFFFF is the hex equivilant for 255,255,255 as you would use in a Color statement. 255 is the same as $FF Colors are stored using 1 byte for red, 1 byte for green, and 1 byte for blue. But ReadPixel returns 4 bytes. The first byte is the alpha (not used in images but it is used in textures), the second is red, the third is green, and the fourth blue. Because JFK didn't want the alpha channel, he threw it away: "And $FFFFFF" "And" performs a bitwise AND operation on the two numbers (the number returned by ReadPixel and $FFFFFF) Boolean math - true or false - math is used on each corresponding bit. The three Boolean operations are: And - if both bits are true then the result is true, otherwise the result is false Or - if any of the bits is true then the result is true, otherwise the result is false Xor - if the bits are different then the result is true, otherwise the result is false So, for JFK's program, "xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx" is the binary form of the number returned by ReadPixel and "0000 0000 1111 1111 1111 1111 1111 1111" is the binary form of $FFFFFF (it's stored as $00FFFFFF but the zeros at the beginning are dropped for the sake of being concise). That said, here is what happens: Alpha Red Green Blue
xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx
And 0000 0000 1111 1111 1111 1111 1111 1111
=================================================
0000 0000 xxxx xxxx xxxx xxxx xxxx xxxxThe alpha channel has been removed but left the RGB channels intact. P.S. I noticed that people like my signature:) but that a number of you don't understand it. :( "10" is binary for the decimal number "2", therefore when read by someone who understands binary, they read it correctly: There are two kinds of people in the world - those that understand binary, and those that don't! But because people learn decimal in school, the first time it is read, people think it says: There are ten kinds of people in the world - those that understand binary, and those that don't! P.P.S Did you know that you can count to 1023 (that's in decimal) on your fingers? (binary "0011 1111 1111", notice, ten ones - one for each finger. Just be careful who sees you show the number four or one-twenty-eight!) |
| ||
| Thanks. I Think I get it now. That meens there are 3 kinds of people in the world Those that know binary, those who don't and those who will. |
| ||
| Now I'm just being argumentative because my kids have annoyed me, but the people 'who will' will be in one of the other groups (presumably the "don't", but not necessarily) before they do. :o) |
| ||
| If you flip a coin you can see my point. Heads on, tails off, while still in the air flipping, void, null, not. Ney? |
| ||
| Flipping coins, sure. But in that instance it can be in one of three states - being heads or tails when at rest doesn't preclude it from being neither (or both) whilst in motion. But people (in the context of knowledge as we're discussing) are less like one-state-or-the-other as much as they're zero or non-zero. And if you accept they're either zero or non-zero, that doesn't leave a third alternative. Of course, this all depends on one's definition of 'know'. For the sake of this argument, I'm using "to have no knowledge of, compared to having some knowledge of". |
| ||
| D-Grafix: Base 10(decimal) is 0123456789 Base 16(hexadecimal) is 0123456789ABCDF In hex, one character= the less herd of NYBBLE. Ther are 4 bits in a nibble and 2 nibbles in a BYTE(or 8 bits) a collection of bytes is called a WORD FF is an 8 bit word (2 nibbles) which is (16x16=256)-1(for zero)=255 (256 possible numbers inc. 0) FFFF is a 16 bit word made up of 2 bytes (or 4 nibbles sometimes called long) therefor (16x16x16x16 or 256x256=65536)-1(for zero)=65535 Ummm, two nybbles is a byte, not a word. A word is two bytes, not one byte. |
| ||
| FF is an 8 bit word (2 nibbles) which is (16x16=256)-1(for zero)=255 (256 possible numbers inc. 0) FFFF is a 16 bit word made up of 2 bytes (or 4 nibbles sometimes called long) therefore (16x16x16x16 or 256x256=65536)-1(for zero)=65535 While it is true that 16x16=256 ($Fx$F=$100), HEX numbers, like binary and decimal, add multiples of the base for each digit. Two examples: 123 is (1x<base squared>)+(2x<base>)+(3x1) FB2A is (Fx<base cubed>)+(Bx<base squared>)+(2x<base>)+(Ax1) Therefore $FFFF is: $Fx($10x$10x$10)+$Fx($10x$10)+$Fx$10+$Fx$1 which in decimal is: 15x(16x16x16)+15x(16x16)+15x16+15x1 (If you haven't figured it out yet, the dollar sign in front of a number - or the letters A through F - means that it is a hexadecimal number in Blitz) |
| ||
| In World of Warcraft, the punchcards in Gnomeragon have binary on them. the binary says exactly what the text above says - something like "SUPER SECURE TRIPLE-ENCODED PUNCHCARD" - I don't remember for sure quite what they say. |
| ||
| While it is true that 16x16=256 ($Fx$F=$100), HEX numbers, like binary and decimal, add multiples of the base for each digit. 16 x 16 = 256 - yes, but. $F x $F is $E1 (15 x 15 = 225). Tut-tut. 16 = $10, so $10 x $10 = $100 (16 x 16 = 256). And just for fun, 10 x 10 is ALWAYS 100, no matter what base you're talking about. Ahem. Sorry. |
| ||
| Ah, the wonderful & numerous powers of ten. :) |
| ||
| "Ummm, two nybbles is a byte, not a word. A word is two bytes, not one byte." i said '8 bit word' not 'WORD' theres a difference: a 2 nyble structure is a BYTE or 8bitword (TECHNICALLY a word can be any length of bits but normally in the PC world, a word refers to collection of bytes not bits.) Orherwise how is it possible to have a so called '12 bit word'? (im not talking about intel based) FOr example (randfomly found with google): http://www.keyless.com/PDF/8BitWord.pdf A normal 'bog standard' WORD is 16 bits unless otherwise specified. I did make another mistake tho, what i meant to say 32bit word is sometimes called long not 16 bit LOL is this all helping riveratt or are you going mad yet (i think i am :) |
| ||
| I have to agree, common knowledge is: 8 Bit= 1 Byte 16 Bit = 1 word 32 Bit = 1 Long 64 Bit = 1 very long - no wait a minute..well however 4 Bit = 1 Nibble tho this is rarely seen anymore. There may be of course other lengths. I think what finally counts is how you gonna store it in RAM, as 8, 16 or 32 Bit value. This may be no issue in Blitz anyway. I wold consider anything other than 4,8,16,32,64 simply as a value, eg. an "unsigned 15bit value". BTW 1 Bit is the smallest unit they say. But actually there are some modem handshake modes that are using 1.5 stopbits. So what we have here is splitting a bit in 2 pieces! Probably we could use it to store 2 Bits in two halfbits. This could become a pretty cool utility - "magic ram multiplyer". :o) |
| ||
| LOL, what do we call that then(1.5), 1 bit and a sniff :) |
| ||
| LOL, what do we call that then(1.5), 1 bit and a sniff :) How about a very simple "going-to-bite-you-but-haven't-bit-yet", or "gy" for short, using the first letters of the first-and-last words to form a word (not a WORD!) LOL! |
| ||
| This thread has gone stupid. |
| ||
| No my fault. I just mentioned the "1.5 Stopbits" paradoxon: http://www.eecs.umich.edu/dco/faq/matlab-6.5/help/toolbox/instrument/stopbits.html |
| ||
| Ah, well, we're game programmers; what can you expect, Gfk? If I didn't crack up at least once a week, I'd probably crack up! ^_^ |
   |