Castle Paradox

Mr B · Joined: 20 Mar 2003 Posts: 382

I don't know if this has been tried before (in fact, I have not even tried it myself, so this is all theoretical), but I came up with an idea for saving location information on maps in a single global variable instead of two -- which could be used to 'remember' where a vehicle was left, for example.

A map cannot have more that 32,767 tiles on it, right? Well, a global variable cannot hold a value larger than 32,767. Because there cannot be more maptiles than the maximum number that a variable can hold, it should be possible to hold the X/Y coordinates of any one place on the map in a single global variable.

Using these pseudocode algorithms:

Saving: var=(Y value of location)*(width of map)+(X value of location)

Loading:
Y=var/(width of map)
X=var,mod,(width of map)

This depends, of course, on the division operation performing an integer truncation instead of rounding. Does OHR division truncate or round?

If it rounds, then have the Loading algorithm go like this:
X=var,mod,(width of map) #same as before...
Y=(var-X)/(width of map) #needs X value

Unfortunately, there is not plotscripting function to find the maximum X widths of maps, so I'd have to use constants.

Criminy, did that make any sense? Rolling Eyes

What do you all think? Would this work?

Flamer · Posted: Sat Sep 27, 2003 8:35 am Post subject:

...what's the point in using one global for a two global variable??

confused!?
it means, good thinking on the idea, but where on earth does this lead to, coz it'll be much easier to use two global variables instead of the one...
unless of course you are using a great number of globals for other things.
_________________
If we were a pack of dogs, IM would be a grand Hound, CN would be a very ficious little pitball, and Giz...well, it doesn't matter breed he is, he'd still be a bitch Raspberry!

(no offense to anyone that was mentioned)

Me · Posted: Sat Sep 27, 2003 9:00 am Post subject:

While your method does make sense, keep in mind that the script compiler can handle up to something like 32767 global variables (or was it 16000 global and 16000 local . . . I forget), so unless you're making an absolutely huge game, you won't bee needing to conserve variables. But good idea, very good idea . . .
_________________
UP DOWN UP DOWN LEFT LEFT RIGHT RIGHT A B START

Cube · **Dimensional Traveller** Joined: 02 Feb 2003 Posts: 294

Local variables don't have any limit (Although having a LOOOOT might cause some problems while the script is running, like a buffer overflow... Maybe), and you can only have 1000 global variables. Still, 1000 numbers is a lot and not many people will fill that up. Using 100-200 is easy to do, but if you're efficient you should never have any problems.

Mr B · Joined: 20 Mar 2003 Posts: 382

Oh, I don't particularly care whether it makes sense or not, just as long as it can be done. Wink

Actually, I don't know if anyone would ever use it, really. Unless if there was a game that had a custom battle engine, but even then... Hmm. Maybe someday I should try to make something that uses up all those variables.

Nah.

But I'm sure that there are some uses for it; they just haven't been found. Or needed, at least. I just wanted to know if it could feasibly be done.

P.S., and I think that the OHR supports 1000 global vars. Yeah. Something like that.

Flamer · Posted: Mon Sep 29, 2003 10:16 am Post subject:

1000 globals is plenty to use for a two or so games...
even games with custom battle systems wouldn't need 100 globals, unless it has all these variables like FF8's zell's limit break(ok not a good example, i'm blank on any battle systems that would go over 100)

besides locals can be "recycled" when one is no longer needed, it's value can be reset (like default 0)
_________________
If we were a pack of dogs, IM would be a grand Hound, CN would be a very ficious little pitball, and Giz...well, it doesn't matter breed he is, he'd still be a bitch Raspberry!

(no offense to anyone that was mentioned)

GMan1988 · Joined: 01 Nov 2003 Posts: 16 Location: Nowhere

I think that if you're going with plotscripting, you can do just about anything. Don't worry, if you really want to make a rogue-style game, you'll think of something eventually. Just don't ask me for help, okay? I'm pretty new with this stuff. Neutral

_________________
.hack//SIGN is freakin' awesome.

Flamer · Posted: Tue Nov 25, 2003 6:31 pm Post subject:

ok... i know this sounds stupid, but i've got two things to say...

one: This is a dead thread, DO NOT REVIVE DEAD THREADS!
two: yes, it is possible to do many things with plotscripting... though not everything...
_________________
If we were a pack of dogs, IM would be a grand Hound, CN would be a very ficious little pitball, and Giz...well, it doesn't matter breed he is, he'd still be a bitch Raspberry!

(no offense to anyone that was mentioned)

Iblis · Posted: Tue Nov 25, 2003 6:37 pm Post subject:

Also, if you absolutely must revive a dead thread, at least make sure you have something worthwhile to say. Saying "I'm new with plotscripting" is not a worthwhile thing to say.
_________________
Locked
OHR Piano

TMC · Posted: Tue Nov 25, 2003 11:26 pm Post subject:

Oh. I never realised this was a dead thread when I was reading it. But I'll sum up on the topic:

1) Of course that would work B, and the OHR doesn't round.

2) There is a maximum number of local varaibles, and when you exceed it, you get a 'script heap overflow' error. That's why there's a (pink) bar in PS debugging mode for locals.

3) You can eat up huge amount of globals when using fake arrays (which means custom battle system also typically use alot) . For instance, for the script on Moggle1s PS library page that tweaks palette by percentage, I believe it uses 768 globals.
_________________
"It is so great it is insanely great."