Meshes ? OMG data structures to manage lines. My brain just froze. So what is you're API that lets you turn pixels on and off? Its not that hard to go from a cloud of points in openGL to a cloud of points in another api.
Contour lines can be harder. you need to connect points on the same elevation without drawing a line through higher/lower elevation. I'd say don't bother connecting lines. just draw a lot of points (of certain colors for certain heights) with a fine enough granularity that they bleed into each other.
Well 0% of the people with playstations know how to program them. 1% may not seem like a lot, (and its a high estimate) but 1% of millions of calculators is still a lot of programmers. I doubt that they know how to program on the PCs. Computers no longer ship with an easy to use basic that gives instant results.
Yes there won't be any formal instruction. Is that a problem? Would any self respecting slashdotter posting at midnight on a friday admit that they needed to be taught programming by a teacher? How much formal teaching did you need to learn the Apple II's built in language?
These slashdot "oh no, the kids no longer have the opportunity to program simple machines like xyz that I did when I was a kid" articles are starting to get monotonous.
The original question mentioned mixing 16 and 32 bit, not explicitely variables. When I code c I'll use 32 bit integers for variables without worrying. But for arrays, data structures, and arrays of data structures I'll put some thought into whether I need a field to be 32 or 16 bits.
you are right that mixing 32 bit and 16 bit variables is a recipe for slowness.
The m68k is a good counter example. 8 bit and 16 bit adds take 4 clocks, but 32 bit math takes 8 clocks. But if you aren't working with embedded systems you don't need to worry about things like this.
Also Most processors have a carry/extend flag (there are exceptions) so a 64 bit add with a 32 bit registers can be done with 2 adds.
>>In a 32-bit RISC processor, most of the instruction bits are reserved to allow large immediate operands for memory offsets, jump targets, and arithmetic/logic operations
In the procs I've studied only half the instruction is used for imediate values, and thats only for instructions that need them. Instructions that don't have an imediate value put these bits to use encoding other things.
There are a fair number of c/asm games written by novices for ti calcs that use empty loops to waste time. But the properly made games use the timer interupts available.
During normal idling/input there shouldn't be a significant drain.
I don't know how the z80 calcs manage thier power but this is how the ti89 does it.
There is a low power port, writing to it disables the cpu till one of several specified interupts occur. If you are actively typing then the cpu will write 0b11111 to the port, if the calc has been sitting idle for a few seconds it will write 0b11110. The difference in the last bit means that the fast timer (256hz/350hz depending on hw version) interupt will not be able to wake the cpu. (If you are typing you want a fast response so the cpu should not be disabled for too long, but if the calc is sitting idle then it will try to save more power.) In both cases the 20hz timer can enable the cpu to take care of things like cursor blinking. (bit #4 is the 20 hz timer, #3 is the link IO, #2 is the 1hz real time clock on a hw2/USB on HW3, #1 is a keypress indicator #0 is the the 256hz/350hz timer)
First a correction: 299! will result in 613 digits, but 2^(255*8-1) is 614 digits.
The ti89's and 49g+'s software is a Computer Algebra System, and one of the simplest rules of a CAS (Even on Mathematica) is that if you give an integer input it should return an integer output. If you want an approximate answer on the ti89 you can do: A) '299.!'//note the decimal B) approx(299!) C) set the calc to approximate mode D) Evaluate the entry line by pressing diamond+ enter, instead of just enter. (diamond is a modifier key like control or alt)
For integers they do display every digit. On the ti89 calc an integer object can be up to 255 bytes (613 decimal digits). For numbers greater than that the calc will switch over to floating point. So the ti89 can do 299! but not 300!
On the hp49g+ the emulated CAS keeps big integers in a variable length binary coded decimal format. And it can support integers much larger than the ti89s 613 digits. The factorial function written in ARM assembly could be computationally faster EXCEPT that it needs to be able to easily turn its results into the variable length BCD format of the 49g+.
Slashdot Mirror
lol, I love making mistakes on \.
Well here's a 57 inch lcd tvX XAA.asp?page=Features
http://www.samsung.com/Products/TV/LCDTV/LNS5797D
I have a 32' lcd TV from Samsung and I've been really impressed.
Meshes ? OMG data structures to manage lines. My brain just froze.
So what is you're API that lets you turn pixels on and off? Its not that hard to go from a cloud of points in openGL to a cloud of points in another api.
Contour lines can be harder. you need to connect points on the same elevation without drawing a line through higher/lower elevation. I'd say don't bother connecting lines. just draw a lot of points (of certain colors for certain heights) with a fine enough granularity that they bleed into each other.
so now what comes first the dinosaur or the egg ?
Well 0% of the people with playstations know how to program them. 1% may not seem like a lot, (and its a high estimate) but 1% of millions of calculators is still a lot of programmers. I doubt that they know how to program on the PCs. Computers no longer ship with an easy to use basic that gives instant results.
Yes there won't be any formal instruction. Is that a problem? Would any self respecting slashdotter posting at midnight on a friday admit that they needed to be taught programming by a teacher? How much formal teaching did you need to learn the Apple II's built in language?
You beat me to this reply.
These slashdot "oh no, the kids no longer have the opportunity to program simple machines like xyz that I did when I was a kid" articles are starting to get monotonous.
-Samuel S
The original question mentioned mixing 16 and 32 bit, not explicitely variables. When I code c I'll use 32 bit integers for variables without worrying. But for arrays, data structures, and arrays of data structures I'll put some thought into whether I need a field to be 32 or 16 bits.
you are right that mixing 32 bit and 16 bit variables is a recipe for slowness.
The m68k is a good counter example. 8 bit and 16 bit adds take 4 clocks, but 32 bit math takes 8 clocks. But if you aren't working with embedded systems you don't need to worry about things like this.
Also Most processors have a carry/extend flag (there are exceptions) so a 64 bit add with a 32 bit registers can be done with 2 adds.
>>In a 32-bit RISC processor, most of the instruction bits are reserved to allow large immediate operands for memory offsets, jump targets, and arithmetic/logic operations
In the procs I've studied only half the instruction is used for imediate values, and thats only for instructions that need them. Instructions that don't have an imediate value put these bits to use encoding other things.
Ok fine I'll go as high as a dollar. But compared to channel surfing during normal tv thats plenty.
And I don't want 'production'
All I want are subtitiles. And Quality wise I would like 180 megs per episode encode.
Price wise for a DVD I'll buy if its $5 per episode. So I'm willing to pay 20%. That isn't as much as the DVD but they don't have to pay voice actors.
Its money they could be making.
Uploading helps cover distribution costs, (for which I don't have a head for at all.) But if its cheap lets bump that encode up to 210 megs.
I think the industry is shooting itself in the foot.
I would buy much more anime if most disks had 5 or more episodes per DVD.
I do download a naruto and bleach, and I could see myself paying $0.50 an episode (and gladly uploading till I got to a 1.00 share rating)
But to pay $25 for a disk with 3 episodes. Give me a break, after I skip the intro and endings thats 60 minutes of content. I expect a 'movies worth'
There are a fair number of c/asm games written by novices for ti calcs that use empty loops to waste time. But the properly made games use the timer interupts available.
During normal idling/input there shouldn't be a significant drain.
I don't know how the z80 calcs manage thier power but this is how the ti89 does it.
There is a low power port, writing to it disables the cpu till one of several specified interupts occur. If you are actively typing then the cpu will write 0b11111 to the port, if the calc has been sitting idle for a few seconds it will write 0b11110. The difference in the last bit means that the fast timer (256hz/350hz depending on hw version) interupt will not be able to wake the cpu. (If you are typing you want a fast response so the cpu should not be disabled for too long, but if the calc is sitting idle then it will try to save more power.) In both cases the 20hz timer can enable the cpu to take care of things like cursor blinking. (bit #4 is the 20 hz timer, #3 is the link IO, #2 is the 1hz real time clock on a hw2/USB on HW3, #1 is a keypress indicator #0 is the the 256hz/350hz timer)
To save power the clock speed of the arm is lowered when it is idle, when calculations are running the 49g+ will raise the cpu speed to 75Mhz.
First a correction: 299! will result in 613 digits, but 2^(255*8-1) is 614 digits.
//note the decimal
The ti89's and 49g+'s software is a Computer Algebra System, and one of the simplest rules of a CAS (Even on Mathematica) is that if you give an integer input it should return an integer output. If you want an approximate answer on the ti89 you can do:
A) '299.!'
B) approx(299!)
C) set the calc to approximate mode
D) Evaluate the entry line by pressing diamond+ enter, instead of just enter. (diamond is a modifier key like control or alt)
For integers they do display every digit. On the ti89 calc an integer object can be up to 255 bytes (613 decimal digits). For numbers greater than that the calc will switch over to floating point. So the ti89 can do 299! but not 300! On the hp49g+ the emulated CAS keeps big integers in a variable length binary coded decimal format. And it can support integers much larger than the ti89s 613 digits. The factorial function written in ARM assembly could be computationally faster EXCEPT that it needs to be able to easily turn its results into the variable length BCD format of the 49g+.
Displaying the results is a larger problem than doing the factorial.