Want to learn Blitz, but I have a problem.

[diablothe2nd]

my algrebra sucks.

I'm trying to create a variable that counts from 255 down, in 1 step increments.

n=n-1 seems to take care of the negative step... but its getting the starting number of 255 in there somewhere that's screwing with my head

closest i've got is:

x=255
n=n-1+x

which is obviously wrong. I know there's a simple answer but I've been thinking about this too hard i think and got my head stuck in it's own loop

[diablothe2nd]

n=n-1+255?


ugh maybe i'm doing this wrong and simply need to make N a number that goes from 255-0 in another way.

multiple for and next loops dont seem to all work at once which i why i'm declaring them as variables first.

[diablothe2nd]

here's the code, perhaps it'll explain more:

y=255-(n=n+1)
For z=255 to 0 Step -1:Line 0,y,320,y,z,1:Next

which *should* simply draw a gradient of white to black from bottom to top.

[P-J]

Originally Posted by diablothe2nd

here's the code, perhaps it'll explain more:

y=255-(n=n+1)
For z=255 to 0 Step -1:Line 0,y,320,y,z,1:Next

which *should* simply draw a gradient of white to black from bottom to top.

This makes no sense dude. You want to draw from Y-line 255 to Y-Line 0 one by one, have the colour value also go from 255 to 0 as you come back up?

Why not just :

For y=255 to 0 Step -1:Line 0,y,320,y,y,1:Next

Also, you said earlier that multiple FOR/NEXT loops wouldn't do what you wanted. Were you using nested loops?

[diablothe2nd]

Originally Posted by P-J

Why not just :

For y=255 to 0 Step -1:Line 0,y,320,y,y,1:Next

Also, you said earlier that multiple FOR/NEXT loops wouldn't do what you wanted. Were you using nested loops?

Gah! so simple to me now. i was totally overthinking it and got myself in a tizzy.

thanks

RE:loops, no i think i was simply misunderstanding how they work.

say though that the Z and Y values were opposite. one climbing to 255, one reducing. how would i go about doing that?

much obliged!

[P-J]

Originally Posted by diablothe2nd

say though that the Z and Y values were opposite. one climbing to 255, one reducing. how would i go about doing that?

Plain version

For y=255 to 0 Step -1:Line 0,y,320,y,(255 - y),1:Next

Better version for learning

for y=255 to 0 step -1
    z = (255 - y)   // So if y = 255, z = 0. If y = 254, z = 1 (etc...)
    line 0, y, 320, y, z, 1
next

A good way to do this sort of stuff is with a pen and paper. Write down the values you want side by side, and then try and work out a formula that links them. In this sort of case it's not necessary, but as you do game programming you could come up against some real sticky situations that need a lot of maths. That's when I turn to the trusty pen and paper and just draw it out.

In this case you could also use some kind of 8-bit NOT if your language supports in. NOT(255) would be 0, NOT (254) would be 1 and so on. But it's worth getting an understanding of binary to understand how that works. Not really worth it if you're coding in basic since that's all taken care of with built-in stuff.

Hope this helps.

[diablothe2nd]

effin' AWESOME! just what the doctor ordered

algebra was never my strong point. and i've never come across For and Next functions before so wasn't sure what i could get away with.

[P-J]

Originally Posted by diablothe2nd

algebra was never my strong point.

Me neither. When I left school and spent 23 hours a day coding on the Amiga (and then PC), I learnt more algebra in a few days than I was ever able to learn at school.

Have you used debuggers before? I think Blitz has one, and possibly a better replacement somewhere. It's a useful learning tool because you can (for example) add a watch on a variable, step through your code, and then see how that variable is affected by each operation. It's great to get into the habit of using one, especially if you plan to progress to coding in assembly.

[diablothe2nd]

just to add to your edit

in this case you could also use some kind of 8-bit NOT if your language supports in. NOT(255) would be 0, NOT (254) would be 1 and so on. But it's worth getting an understanding of binary to understand how that works. Not really worth it if you're coding in basic since that's all taken care of with built-in stuff.

ironically i'm quite knowledgable in both binary and hex. The difficulty for me though is i've never done any programming more advanced than editing configs, so dont know what function to choose, and more specifically when it should be used.

the pdf of functions is helping though. I guess it's like CAD, or any graphical program that i'm completely fluent in. There are many ways of doing it, but figuring out the most efficient way is only gained from experience. If i wanted to draw say a hexagon on screen i wouldn't know whether or not to draw all 6 lines using 6 line parameters... or use 1 and somehow copy and manipulate it! in CAD it'd be 2 key presses, 1 click and typing a single dimension... total time taken, 3 seconds :P

its so frustrating for me, cos i've got a near photographic/videographic image my head of everything i want the end user to see and do... from launching the game to finishing/completing it. putting that down into images is a piece of piss for me. coding it though.... phew! I've noticed you programmers start out with a vague idea and just go with the flow. I think my line of thinking of "this is what it will be", even though i'm considering limitations of hardware, is my problem. i simply have no knowledge of the tools infront of me to back trace.

[Thorham]

Going from dark to light:

for y=0 to 255 step 1
    line 0, y, 320, y, y, 1
next

Going from light to dark:

z=255
for y=0 to 255 step 1
    line 0, y, 320, y, z, 1
    z=z-1
next

Note that 'step 1' may be removable (most basic dialects default to 'step 1' for 'for next' loops).

Originally Posted by diablothe2nd

I've noticed you programmers start out with a vague idea and just go with the flow.

Not really, because that is the worst way to write programs; you really have to think it out properly, or you'll end up with a mess. It's similar to the good (bad!) old: 'Type first, think later' way of programming.

[pmc]

Originally Posted by Thorham

Not really, because that is the worst way to write programs; you really have to think it out properly, or you'll end up with a mess. It's similar to the good (bad!) old: 'Type first, think later' way of programming.

Agree.

Think first, code second!

This gets quicker as you get more experienced as you start getting to the stage where you kind of know how to go about coding something as you're coding, or at least a few split seconds before you write the actual code. That's how it feels to me at least.

But if something's larger, more unfamiliar or new the first thing I do is spend time coding things "in my head" first.

[Graham Humphrey]

Originally Posted by diablothe2nd

my first hurdle seems to be colour codes. the circle command syntax has only one spot on the end for a colour code and i can't find the correct syntax for a non grey colour. i've tried using RGB hex values, both with and without a hash (#) but it's not having it.

'Colour' in this case has to be a colour number from the palette you're using. If you're using 256 colours for example you have to pick a number from 0 to 255 and the circle will be drawn using that colour register.

[diablothe2nd]

thanks for clearing that one up graham