its a possibility, i was avoiding suggesting it merely because i haven't got around to looking at pandas source and have no idea if it will be understandable to me.
OK, I am going to try to compile panda, I've been putting it off for too long.
this will answer lots of questions I have about panda 1.6.0 (the as yet unreleased version)
EDIT: just got panda 1.5.4 compiled (im working up to 1.6.0) and took a look into the function, but the couple of changes i made had no affect (in fact, its slower, presumably because im building it in a slightly less performancy way).
anyway, its 4:30 and i have lectures tommorow, so sleep.
Started Coding
-
finnbryant
- Posts: 360
- Joined: Sat Feb 17, 2007 11:18 am
- Location: England
Last edited by finnbryant on Mon Feb 23, 2009 4:29 am, edited 1 time in total.
-
fuzzylunkin1
-
finnbryant
- Posts: 360
- Joined: Sat Feb 17, 2007 11:18 am
- Location: England
heh yeah, Ive been doing assignments pretty much constantly since my last update....
and then i bought a wiimote and some electronicy stuff and am trying to turn my screen into a psuedo-touchscreen
ive been pretty disctracted
and ive still got 2 assignments to go for next thursday *sigh* (and then 2/maybe 3 more for after easter >_<)
that all said, im not going to turn this thread into my personal emo-blog, ill just try to get through it all and get on with this as soon as i can (hopefully ill get some done over easter)
Ive been suffering from a quantus-related problem too, which is why i havnt been able to do anything, even in the little tiney bits of time i get every now and again (its one of those problems that requires an extended period of thought to solve), its basically, after deciding where to place a path, how do you then flatten the terrain? blurring could do it, but i think there was a reason that couldnt be used (thats why it needs 1 long think, i forget everything otherwise
). whay would we need this? because it is the only way that people will be able to get around in mountinous regions (except flight of course), plus i think it would be awesome to have a map with a spirilling path to the top of a mountain, where the buildible area will be
Anything where you have to rely on natural advatages makes for awesome stratagies, and even more awesome gameplay changes when someone works out how to sidestep a particular "advantage" on a particular map, I want this game to be riddled with game changers like galvs were in the alleg beta (in the dev diaries, basically they kept upping galvs power, but no-one used them cos they seemed weak, then one day someone got a team together and convinced them to all galv at once (quite a feat i expect!) and suddely everyone was amazed by how powerful this disregarded tech was, of course it was severly nerfed soon after, but it still must have been awesome to be that comm) of course I dont want anything that unbalanced in the game, but a few minor ones (maybe map specific, like discovering that there was a second way through a particular mountainous area which usually avoids the enemy probes, but only infantry could get through) - this is entirely possible (and even probable a few times at least) thanks to the fact that most of any map will be computer genererated and even the map creator may not know all of its secrets .
there, ive given people reason to keep watching this project and reignited my own interest in one less-than-swift paragraph! it may have lacked some punctuation, but i cant be arsed to go through it again...
and then i bought a wiimote and some electronicy stuff and am trying to turn my screen into a psuedo-touchscreen
ive been pretty disctracted
and ive still got 2 assignments to go for next thursday *sigh* (and then 2/maybe 3 more for after easter >_<)
that all said, im not going to turn this thread into my personal emo-blog, ill just try to get through it all and get on with this as soon as i can (hopefully ill get some done over easter)
Ive been suffering from a quantus-related problem too, which is why i havnt been able to do anything, even in the little tiney bits of time i get every now and again (its one of those problems that requires an extended period of thought to solve), its basically, after deciding where to place a path, how do you then flatten the terrain? blurring could do it, but i think there was a reason that couldnt be used (thats why it needs 1 long think, i forget everything otherwise
). whay would we need this? because it is the only way that people will be able to get around in mountinous regions (except flight of course), plus i think it would be awesome to have a map with a spirilling path to the top of a mountain, where the buildible area will be Anything where you have to rely on natural advatages makes for awesome stratagies, and even more awesome gameplay changes when someone works out how to sidestep a particular "advantage" on a particular map, I want this game to be riddled with game changers like galvs were in the alleg beta (in the dev diaries, basically they kept upping galvs power, but no-one used them cos they seemed weak, then one day someone got a team together and convinced them to all galv at once (quite a feat i expect!) and suddely everyone was amazed by how powerful this disregarded tech was, of course it was severly nerfed soon after, but it still must have been awesome to be that comm) of course I dont want anything that unbalanced in the game, but a few minor ones (maybe map specific, like discovering that there was a second way through a particular mountainous area which usually avoids the enemy probes, but only infantry could get through) - this is entirely possible (and even probable a few times at least) thanks to the fact that most of any map will be computer genererated and even the map creator may not know all of its secrets
.there, ive given people reason to keep watching this project and reignited my own interest in one less-than-swift paragraph! it may have lacked some punctuation, but i cant be arsed to go through it again...
Last edited by finnbryant on Sun Mar 29, 2009 4:56 am, edited 1 time in total.
-
fuzzylunkin1
We could define a slope, path width, depth, ect and then statistically randomly generate points (lets say 10 for simplicity) away from the line of regression going from point A to point B. We then repeat the process until it gets smaller and smaller and eventually to such a small scale it will no longer matter.
Divide and conquer.
This process would be used at key points (every specified distance, or maybe corners/turns), and we would smooth out the terrain between each point with a similar process
Think about using the pen tool in Photoshop or GIMP and making curved paths. Something along those lines. We would be able to change *mean and standard deviation/variation* to experiment with different terrain types (rocky, grassy).
Sounds like that might be pretty inefficient, but it's an idea.
EDIT:
* Maybe just standard deviation/variance, depending on how we approach it, I suppose.
Divide and conquer.
This process would be used at key points (every specified distance, or maybe corners/turns), and we would smooth out the terrain between each point with a similar process
Think about using the pen tool in Photoshop or GIMP and making curved paths. Something along those lines. We would be able to change *mean and standard deviation/variation* to experiment with different terrain types (rocky, grassy).
Sounds like that might be pretty inefficient, but it's an idea.
EDIT:
* Maybe just standard deviation/variance, depending on how we approach it, I suppose.
Last edited by fuzzylunkin1 on Sun Mar 29, 2009 5:50 am, edited 1 time in total.
-
finnbryant
- Posts: 360
- Joined: Sat Feb 17, 2007 11:18 am
- Location: England
love that one! ive been watching it for a while and its the main reason i decided to use perlin noise.
hock, good point on the cap ships, maybe ill have paths and "roads" (if i end up doing fuzzes suggestion ill have them as defaults and allow people to do other options themselves)...
fuzz, i will have to look over your post again another time but it is similar to how i thought i would end up doing it, its quite similar to perlin, though it does sound like it would be very processor hungry...
though it has given me an idea, since the terrain is an image, i could just use the image manipulation functions and *actually* just draw a line
, then i would just have to work out where it goes from and to, which is both easier and more interesting. Problem may be that i dont think the line drawing allows blurry edges (beyond basic anti-aliasing), so i may have to implement a blur on top before applying, but that should be ok.
oh, bit of good news, ive got on good authority that the next version of panda will be put any day now , since it implements threads i may be able to get terrain flattening to work in realtime finally! (it will require that "building" has a lower limit to the time it can take, and we will have to find that limit on a really slow machine so we can be certain we can support everything. plus i dont know how ugly it will get at switchover time, it *should* be undetectable, but the method I'll be using will be messy, unless i bother to hack the panda source and do it *properly*, which could take all of easter, just for that...)
hock, good point on the cap ships, maybe ill have paths and "roads" (if i end up doing fuzzes suggestion ill have them as defaults and allow people to do other options themselves)...
fuzz, i will have to look over your post again another time but it is similar to how i thought i would end up doing it, its quite similar to perlin, though it does sound like it would be very processor hungry...
though it has given me an idea, since the terrain is an image, i could just use the image manipulation functions and *actually* just draw a line
, then i would just have to work out where it goes from and to, which is both easier and more interesting. Problem may be that i dont think the line drawing allows blurry edges (beyond basic anti-aliasing), so i may have to implement a blur on top before applying, but that should be ok.oh, bit of good news, ive got on good authority that the next version of panda will be put any day now
, since it implements threads i may be able to get terrain flattening to work in realtime finally! (it will require that "building" has a lower limit to the time it can take, and we will have to find that limit on a really slow machine so we can be certain we can support everything. plus i dont know how ugly it will get at switchover time, it *should* be undetectable, but the method I'll be using will be messy, unless i bother to hack the panda source and do it *properly*, which could take all of easter, just for that...)
Last edited by finnbryant on Mon Mar 30, 2009 11:36 am, edited 1 time in total.
-
finnbryant
- Posts: 360
- Joined: Sat Feb 17, 2007 11:18 am
- Location: England
WOO!! panda3d 1.6.0 is out!
all my assignment work is in!
im on easter break!
ive got more assignments actually, but they're not in until after easter sometime, so ive still got a good amount of time left over.
so what does 1.6.0 do for us?
It gives us threading (this means that some extremely slow calculations ive been fretting about can be run in the background, this means the game can play as normal, this is only useful for things that are expected to take a long time in game, but you want to game to go on despite that. (in particular - building bases *should* be possible on rough terrain now, since the 5-10 second calculation to flatten the terrain for it can be done in the background whilst the bulid sphere is up, of course if any players have a really slow computer, they may have problems, but it should just be that the base looks really stupid and floating off the ground a bit for a few seconds until they catch up)
The collada importer will allow people to use sketchup to create models for the game.
ill be able to modify the game engine code to fit our game better without the horrible anguish of knowing ill have to do it all again for the next version (not for a while at least).
EDIT: oh, 1 more, appartently i will be able to use 16 bit grayscale heightmaps now, that will mean much more interesting terrain, with more subtle slopes and huge mountains possible at the same time, plus less (hopefully 0) danger of hitting the max value and losing the data
all my assignment work is in!
im on easter break!
ive got more assignments actually, but they're not in until after easter sometime, so ive still got a good amount of time left over.
so what does 1.6.0 do for us?
It gives us threading (this means that some extremely slow calculations ive been fretting about can be run in the background, this means the game can play as normal, this is only useful for things that are expected to take a long time in game, but you want to game to go on despite that. (in particular - building bases *should* be possible on rough terrain now, since the 5-10 second calculation to flatten the terrain for it can be done in the background whilst the bulid sphere is up, of course if any players have a really slow computer, they may have problems, but it should just be that the base looks really stupid and floating off the ground a bit for a few seconds until they catch up)
The collada importer will allow people to use sketchup to create models for the game.
ill be able to modify the game engine code to fit our game better without the horrible anguish of knowing ill have to do it all again for the next version (not for a while at least).
EDIT: oh, 1 more, appartently i will be able to use 16 bit grayscale heightmaps now, that will mean much more interesting terrain, with more subtle slopes and huge mountains possible at the same time, plus less (hopefully 0) danger of hitting the max value and losing the data
Last edited by finnbryant on Mon Apr 06, 2009 3:34 am, edited 1 time in total.
-
fuzzylunkin1
I'm currently busy with school and work. End of my senior year's coming along quickly, and I'm taking four Advanced Placement tests.
Currently working on my first game (2d zombie shooter). It's coming along great and is teaching me a lot more about programming. It's written in C++.
Hopefully that will help me a lot with design factors and understanding of Quantus .
BTW, I know Factoid works in a game [small] development company/[startup?]. Maybe we could get some tips from him.
Currently working on my first game (2d zombie shooter). It's coming along great and is teaching me a lot more about programming. It's written in C++.
Hopefully that will help me a lot with design factors and understanding of Quantus
.BTW, I know Factoid works in a game [small] development company/[startup?]. Maybe we could get some tips from him.