Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

DarkBASIC Professional Discussion / - DBPro Coding Challenges -

Author
Message
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 30th Sep 2005 00:40
Ah - good to see a little bit of life on this one again!

@Segan - looking good. I like the ease with which you can change the variables. Only problem I had was that when I pasted your example sets of variables, the terrain lost it's infiniteness - ie. I got to the edge. Did I do something wrong?

@Rii - nice textures. The delay in changing zones is the main problem I think, as it's about 3 seconds on my pc (although that means it will probably be less on most other peoples pc's). Is it possible to do the calculations for each new zone in small stages as you approach the boundary, rather than in one big chunk when you get there?

Also, just wondering - and it's a question for everyone really - if you went say for 10 minutes in one direction, then came back to the same place, would the terrain be the same? And if so, does it mean that the further you travel, the more information the program has to store about the entire terrain? To me it seems like a necessary limitation on using an algorithm to create an infinite terrain - the further you travel, the bigger the calculations needed by the algorithm. Reminds me of the problems I had when writing my Mandlebrot set fractal code - I could only zoom in about 12 times before the numbers calcualated by the algorithm became so huge that DBP couldn't handle them.

spooky
22
Years of Service
User Offline
Joined: 30th Aug 2002
Location: United Kingdom
Posted: 30th Sep 2005 00:54
For some reason RiiDii's code just silenty crashes on line 166. I believe it's the point command. For some reason doing a point or dot outside bitmap range crashes some pcs but works on others. Code is quite difficult to debug but I'll have a go later.

Re your terrain question. I know exactly what you mean by zooming into a fractal many times. Even using double floats, DBPro just is not accurate for very small numbers. My terrain proggy has to calculate all the heights into a huge array so you can indeed travel in one direction and either wrap around back to start, or turn around and go back, and terrain will be same as you left it. Seems a bit silly to start off say on top of a mountain, go forward for a few seconds, turn round, go back to where you started and find mountain gone!

Nearly all the tutorials on terrain creation mean you have to precalc all the heights for entire landscape. I have racked my brains trying to work out a method and formula for creating the heights on the fly.

Been too busy this week to finish my code. Will do it over weekend.

Boo!
Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 30th Sep 2005 01:59
@Ric Lol. I found the error (another typo one). When you copy/pasted the new values in, you should see the number of globals decrease by two. I will fix the code right after posting this message.

Quote: "
Also, just wondering - and it's a question for everyone really - if you went say for 10 minutes in one direction, then came back to the same place, would the terrain be the same? And if so, does it mean that the further you travel, the more information the program has to store about the entire terrain? To me it seems like a necessary limitation on using an algorithm to create an infinite terrain - the further you travel, the bigger the calculations needed by the algorithm. Reminds me of the problems I had when writing my Mandlebrot set fractal code - I could only zoom in about 12 times before the numbers calcualated by the algorithm became so huge that DBP couldn't handle them.
"


My code generates the same stuff every time, regardless whether you go 10 minutes in one direction and then come back(mandatory for me, because my matrixes are constantly swapping values, and thus have to always use the same formula.) For the numbers, I havn't tested, but there is no doubt with my current code that the numbers will eventually get too big to handle... However, i think that that will take a REALLY long time (since the values are only incrimented/decrimented by 1 for each matrix used.) This is something I can look into though for my next version...
Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 30th Sep 2005 02:06
Sorry for double post...

I am having problems with the forums right now (may be because I'm still being moderated...) When I try to edit my above post, it gives me an error message.

So, you all will just have to bear with me and make sure that these two variables:



Are somewhere at the top of the program before you run for it to work properly. Is that too much work for you?
RiiDii
19
Years of Service
User Offline
Joined: 20th Jan 2005
Location: Inatincan
Posted: 30th Sep 2005 03:04 Edited at: 30th Sep 2005 03:06
My terrain in theory should return the same map when you return to the same location. "Random numbers" are actually not so random as they are really a sequence of numbers that only appears to be random. Basing the first number in the series on the Timer() makes the series a little more unpredictable, but still a series based on the Timer() value.

In my code, I create a seed based on the unique x/z segment of the section of the terrain (not db units, but terrain segments). All numbers generated in any sequence following that seed will be identical to any other time that same sequence is generated. Since the sequence is predictable and recreateble, there is no need to "store" vast amounts of data. My code does have an issue going into negative segments, but that can fixed (debugged).

Here's a snippet to demonstrate how the random sequences work.
The seed value generates the sequence (use the updown arrow keys to change the seed). Select a seed value and write down the seed and the sequence of numbers. Then increase the seed to any value, and then return to the seed the value that you wrote down. You will see that the 3 random numbers are exactly the same.




As far as the bug in changing terrain (I can't test it now, but I'll give it a go later); I suspect that the Advanced Terrain commands used rotate the image right 90 degrees when creating the terrain, so that the Y-axis of the image becomes the x-axis of the terrain, instead of the Z-axis. Now I just need to figure out what that means needs to be done in order to fix that.

And the 3 second delay between "zones" is about what I get too. I reduced this by splitting up creating the new heightmap and creating the terrain, which reduced the delay a little bit. The real solution would be to try and manage something like 16 heightmaps and 16 terrains (in a 4x4 array). The terrains might be small enough that the delay would be very minimal. If that still causes a delay, the next step would be to manipulate the terrain's "limbs" (or a memblock) using the new vertex commands, basically morphing the terrain as the player went along.


Open MMORPG: It's your game!
Philip
21
Years of Service
User Offline
Joined: 15th Jun 2003
Location: United Kingdom
Posted: 30th Sep 2005 15:16 Edited at: 30th Sep 2005 15:21
This is a really interesting challenge. Please can I interject here?

Having looked at the v. interesting website that Tinkergirl published a http to, and having looked at some of the really clever ideas above (well done guys!) I'm wondering if it would be relatively straightforward to try and create a pretend planetary landscape using some of the methods being discussed.

What I'm thinking about is, say, taking one of the JPG maps of planet Earth that are published on NASA's website, then parsing the jpg file, then applying the diamond square algorithm to create a mock "Earth" landscape.

Thoughts?

I guess it would have to be seriously fractall'd up to create a landscape that actually resembles planet earth ... !

And I'm not asking because of my long running game Entropy and because I'd love to be able to incorporate functionality for the player to land on planets. No, this post is entirely a coincidence. Yes, a coincidence. Thats my story and I'm sticking to it.

Incidentally, is it hot in here or is it just me?

Philip

Cheer if you like bears! Cheer if you like jam sandwiches!
P3.2ghz / 1 gig / GeForce FX 5900 128meg / WinXP home
Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 30th Sep 2005 15:29
Of course you can join! I'd love to have enough time to do these.. I REALLY miss spare-time-programming :-(

Is it classed as an infinite terrain if it loops? On earth you'll never reach the end of the terrain due to the nature of the world not being flat..

My Website:
spooky
22
Years of Service
User Offline
Joined: 30th Aug 2002
Location: United Kingdom
Posted: 30th Sep 2005 15:59
My terrain will wrap around after a while but if you are moving around at different angles, like in a flight sim, I don't think anyone will notice. For example if my matrix is 100*100 for arguments sake but array holding terrain info is 1024*1024, that is over 100 times the area of the matrix. That's way big enough to call an infinite terrain (in my books anyway!). Beacause I have used that diamond square method across whole terrain, even the very edges wrap around nicely.

Boo!
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 30th Sep 2005 19:24
Quote: "Is it classed as an infinite terrain if it loops?"


I guess it's open to interpretation - personally I'd say an infinite terrain would ideally never repeat itself. But that's not to say that an infinte series can never repeat itself, because it can (1.21212121r for example). My argument, though, would be that a circle, whilst having no beginning or end, still has a finite length (ie. its perimeter), and therefore can not be considered infinite. I think.

@Rii - why stick to just a 4x4 grid? I'm thinking that a 128x128 grid, for example, should be as easy to manage, as they would be smaller. And the advantage being that you only need to create an extra row of 1x128 terrains each time you move forward a zone, a total area much smaller than one giant terrain ...?

@Segan - sorry - should have spotted that bug myself!

I'll hopefully be starting my own attempt at this challenge this weekend. Too fun to miss out on, even though I know I can't win!

Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 30th Sep 2005 19:34 Edited at: 30th Sep 2005 19:36
I might try to do this tonight.. I hope my new siggy works here...

EDIT: WOOT IT WORKS!!! Everytime you see it, it SHOULD have a different quote... All I need now are quotes!
FYI: That was the fruits of my afternoon at "work"

My Website:
RiiDii
19
Years of Service
User Offline
Joined: 20th Jan 2005
Location: Inatincan
Posted: 30th Sep 2005 21:27
Quote: "why stick to just a 4x4 grid?"

Once it's small enough that the delay is at it's minimum, it actually becomes a waste of resources/time to make it smaller. In other words; Advanced Terrain requires a given amount of resources and time just being called. At a 2x2 pixel heightmap terrain (I think that would be the smallest possible), it would probably take the same amount of time to create as a 64x64 pixel heightmap terrain. These are just estimates (and may vary from pc to pc), but taking these values as examples, creating 32 2x2 terrains and piecing them to together would take much longer than creating a single 64x64 terrain.

Note: Some initial tests done by Aura for the AGE terrain editor showed that a 64x64 pixel heightmap terrain was actually a few milliseconds faster to create than a 32x32 or 16x16 pixel heightmap terrains. 128x128 was a lot slower though.


Open MMORPG: It's your game!
Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 1st Oct 2005 00:00
@Philip
Quote: "
What I'm thinking about is, say, taking one of the JPG maps of planet Earth that are published on NASA's website, then parsing the jpg file, then applying the diamond square algorithm to create a mock "Earth" landscape.
"


I think the biggest problem with this would be the fact that the colour of the earth does not necessarily represent it's height. Thus while you could probably build a terrain from the map, with recognizable colours and water, the terrain might not be accurate (anything goes with imaginary planets though!)
Darkbasic MADPSP
19
Years of Service
User Offline
Joined: 15th Jun 2005
Location: Uk
Posted: 1st Oct 2005 13:35
when does this finish i've got my entrie ready (completly new)

Gtas back and bigger and better than before
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 1st Oct 2005 15:09
Thursday 6th is the deadline - but you don't have to wait until the deadline before posting your entry!

Darkbasic MADPSP
19
Years of Service
User Offline
Joined: 15th Jun 2005
Location: Uk
Posted: 1st Oct 2005 15:41 Edited at: 1st Oct 2005 15:42

here you go and it's under 20 lines!!!

Gtas back and bigger and better than before
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 1st Oct 2005 16:32 Edited at: 1st Oct 2005 16:34
Getting better DBMad - but don't stop there! There's more to do if you want to get your code working better.

1. To set the sync rate, you need to have first called 'sync on', and you must then use the 'sync' command every loop. If you don't do that, the sync rate is controlled automatically. eg:



2. Only put commands in your do-loop that need to be called every loop. For example, putting 'position camera 50, 40, 70' in your do-loop is going to keep telling it to do the same thing over and over - which is a waste of processing speed. Call this command just once - before your do-loop.

3. You have two do-loops. Without any way for the program to break out of the first do loop, it will never get to second one - so the code



is completely pointless as the program never gets there.

4. Instead of shifting the matrix by a set amount each loop, why not give control to the user - let them use the arrow keys to control how the matrix gets shifted. eg, put



within your do-loop.

5. Try texturing your matrix so it looks more like a landscape than some messed up fishing net! You'll need to create an image on the screen - perhaps using the box command, then grab the image using 'get image'. Then texture your matrix with the image using 'prepare matrix texture'.



Darkbasic MADPSP
19
Years of Service
User Offline
Joined: 15th Jun 2005
Location: Uk
Posted: 1st Oct 2005 21:00
thanks ric I'm glad you like my first proper attempt

Gtas back and bigger and better than before
Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 2nd Oct 2005 21:15
Ok - I tried to make some time to do this over the weekend as I really like terrains and all that goes with it... however, I have not made enough time. I looked into something that could create a Perlin Noise image... This is what I came up with but its not working right. I also have tweaked it so its not what the origional based psuedo code Google found for me was....


Run it and use the arrows to scroll the image.. I think I can see repeating though which is not what I wanted. Its kinda like Magic Eye!!

My Website:
IBOL
Retired Moderator
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location: @IBOL17
Posted: 2nd Oct 2005 23:07 Edited at: 2nd Oct 2005 23:11
i have been thinking a lot about how to do this competition,
but all my work with matrixes is flawed somehow...
specifically, i create 9 matrices as 3x3, but when i try to smooth
the edges from (1,1) to (0,1) it doesn't work.

the reason i haven't used the shift matrix commands is that
you can only move them up,down,left,right...there's no 360-degree
movement.

BUT: if you could only move 4 directions, but the terrain was infinte
and non-repeating, would it be considered??

thanks,
bob

Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 3rd Oct 2005 02:04
I think thats the idea... an infinite terrain that you can "scroll" about.

I've been thinking a lot about my code above and lots of realisations came into my head about improving it, such as..
1) Instead of mapping each pixel to 1 PerlinPixel, scale the image... each pixel could be 1/10th of a PerlinPixel. There is a function that should smooth it all out nicely. There was better code - in my example it uses Linear but there was Cosine smoothing (better but slower) and Bicubic smoothing (slowest but best).
2) Instead of recalculating the entire grid when shifting, instead copy all the pixels up 1 and only recalculate the row/column that is "new"

Hopefully I'll be able to do this tomorrow night...

My Website:
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 3rd Oct 2005 02:52
@Ibol: If only shifts up down left and right, but can't you do a combination of each to give the effect of shifting diagonally? For example, y-rotate the camera 45 degrees to the right, shift 1 down and 1 left, and it would seem like you are moving diagonally. Y-rotate the camera 60 degrees to the right, shift down 1 and left 2, and that would give you another direction - etc. Haven't tried it myself, but it's working fine in my head!

@Nick: glad you could find some time to have a go at this one. I like the perlin noise idea. Looking forward to seeing it converted into a terrain.

Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 3rd Oct 2005 23:43
In THEORY this should be an infinite terrain (allbeit in 2D)




I learned a LOT from this - mainly that Memblocks are FRICKIN quick!

I doubt I'll have anymore time to do this as I'm off on holiday on friday and need to start preparing.

The above code works. It generates a nice looking greyscale image that I imagine might produce nice terrain. On my PC I can scroll in 2 directions simultaneously at about 180fps and 1 direction at 380fps. Each scroll represents moving into a new "square".

Bugs.. Oh where to begin!
1) For some reason, when scrolling diagnoally it distorts.. I think this is a problem relating to point 2
2) The horizontal scrolling isn't quite right... If you "wiggle it" on the spot horizontally, you see lines where it hasn't quite refreshed the edge right
3) Down and Left (both increments) work perfectly (on their own ) but if you go up or right (decerement) as soo as you go into the negative range, it repeats in blocks like it is having negative number issues... This is probably easy to fix - but I'm out of time.

I would be really interested to see what this does in 3D.. So someone please use it!! Even it it is only in a matrix.

My Website:
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 4th Oct 2005 03:00
I discovered a weird optical illusion with your code Nick - if you stare into the image for a while at the same time as scrolling it, it appears to go a light bluish colour. Cool effect! Now don't tell me it's just me - is it?

I noticed that each small block has what looks like a graded fill. It would be even more effective, I'm guessing, if you could get the small blocks to match at the edges with their neighbours - it would give a smoother, more natural landscape I would have thought.

Good work, though.

Nicholas Thompson
20
Years of Service
User Offline
Joined: 6th Sep 2004
Location: Bognor Regis, UK
Posted: 4th Oct 2005 10:52
Did you play with the remarked out Interpolation sections? The options are:
1) None, you end up with a gridded effect
2) Linear, each "block" if gradiented to try to match
3) Cosine, uses Cos to try to make the gradient smoother

There was Cubic - but that required more work and slightly different arguements (needed 5 rather than 3).

I didn't notice the bluish effect - but I use a TFT so it might not show it due to minor ghosting (its a good TFT (Iiyama E431), but modern ones are MUCH better).

Do you have any idea's about the problems I mentioned where is loses the ability to tile neatly?

My Website:
TDK
Retired Moderator
22
Years of Service
User Offline
Joined: 19th Nov 2002
Location: UK
Posted: 6th Oct 2005 01:26
Here's my attempt at an entry to the challenge:

I'm not all that familiar with DBPro and I found quite a few differences with what I'm used to in DBC. As such, it has a few rough edges which I won't have time to fix before tomorrow.

No media required to run it...



TDK_Man

Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 6th Oct 2005 04:06 Edited at: 8th Oct 2005 03:02
Well, here is my newest example:

Same terrain, just using a couple of trig functions to figure out the hill height.



Unfortunately, I just ran out of time to finish it completely (I won't be able to do any work tomorow on it.) What I didn't complete was:

1: Combination of both programs into one, with variables determining which algorythm to use.

2: A way to make the humps in the terrain a little less predictable.

3: Water

4: Some "Low poly trees" as an example of putting objects into the program.

5: An attempt at a skysphere

6: Better textures

So as you can see, I had pretty lofty goals for this project, but it kind of failed at the end. I think the secret will be to build up a library of functions, so that getting a good grass image is as simple as copy and paste!


EDIT: Whopps! I forgot to post the controls! (They are pretty easy though: arrow keys to direct camera, and control and shift for forwards backwords). The camera system seems a little buggy, but as I said I don\'t have enough time to fix it!
IBOL
Retired Moderator
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location: @IBOL17
Posted: 6th Oct 2005 08:18 Edited at: 6th Oct 2005 08:18
well, it looks like i'm not going to make it...got distracted...
RE:Bluish Effect.
nicholas, you said you were using a TFT. i think you might need a THC or an LSD to get that effect.
bob

spooky
22
Years of Service
User Offline
Joined: 30th Aug 2002
Location: United Kingdom
Posted: 6th Oct 2005 11:59
Oooh, running out of time rapidly. I presume we have until midinight tonight (Thursday)

@TDK - I love the way you have textured the matrix - I would never have thought of that idea. I was wondering how to get a smooth texture transition between height levels without seeing sudden change in colours.

Oh well, back to work now

Boo!
CPU
21
Years of Service
User Offline
Joined: 4th Jul 2003
Location: Carlsbad, CA
Posted: 6th Oct 2005 19:22 Edited at: 6th Oct 2005 19:24
Hey all,
good entries around the board, I thought I might just slip this in at the end, more just for people interested in making something like an infinte canyon, don' have enough time to make it fully "Infinte" but its a good start for those that are interested, I guess you could say its not an "offical" entry, but its for those that are interested, (for those that wan't to know, I pulled it out of my archives from about 2 years ago...)

[edit]
Actualy, It's more of a tutorial, now that I remember more about it, I'm fairly sure I classified it as "intermediate",

Well here it is for those interested...



CPU

[center]K-OS Battlefields
IS
///---///---///---UNDER CONSTRUCTION---\\\---\\\---\\\
[center]
spooky
22
Years of Service
User Offline
Joined: 30th Aug 2002
Location: United Kingdom
Posted: 7th Oct 2005 00:00
I've only just finished working for the day, and I've got one hell of a headache, so I'm going to bed. Still got too much to do on terrain proggy so gracefully bow out. I'll finish it over the weekend and post it then. Good luck everyone!

Boo!
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 7th Oct 2005 01:24
Eek - is tonight deadline night? DIdn't get home from work till 10pm tonight, so I won't be able to judge until tomorrow (Friday) night. Any last entries - you've got until then!

Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 7th Oct 2005 01:48
Well, here is my newest example:

Same terrain, just using a couple of trig functions to figure out the hill height.



Unfortunately, I just ran out of time to finish it completely (I won\'t be able to do any work tomorow on it.) What I didn\'t complete was:

1: Combination of both programs into one, with variables determining which algorythm to use.

2: A way to make the humps in the terrain a little less predictable.

3: Water

4: Some \"Low poly trees\" as an example of putting objects into the program.

5: An attempt at a skysphere

6: Better textures

So as you can see, I had pretty lofty goals for this project, but it kind of failed at the end. I think the secret will be to build up a library of functions, so that getting a good grass image is as simple as copy and paste!

EDIT: Whopps! I forgot to post the controls! (They are pretty easy though: arrow keys to direct camera, and control and shift for forwards backwords). The camera system seems a little buggy, but as I said I don\'t have enough time to fix it!
Darkbasic MADPSP
19
Years of Service
User Offline
Joined: 15th Jun 2005
Location: Uk
Posted: 7th Oct 2005 22:38
who has won?

Gtas back and bigger and better than before
spooky
22
Years of Service
User Offline
Joined: 30th Aug 2002
Location: United Kingdom
Posted: 8th Oct 2005 00:25
Well my vote goes with TDK as it looks great but obviously I would have won if I had finished my entry

Boo!
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 8th Oct 2005 02:44 Edited at: 8th Oct 2005 02:45
Ok - judging time - here goes!

DBMad: good to see your two entries. Still at a basic level, but a big improvement between first and second entries which shows you must have learned something from the exercise.

Segan: Very useable code in your first couple of attempts. Your last two posts looked like they had potential, although I couldn't quite work out what I was looking at somtimes, as the camera angle doesn't seem right somehow. Nice work - nevertheless

TDK: Good work - infinite terrain based on the matrix shifting technique, and terrain produced in a similar way to my blue island terrain - using random peaks then averaging the parts between to smooth the terrain. Multitexturing based on terrain height, too.

Rii: Looks good - like the advanced terrain approach, although it means it suffers from zone-lag, which would be a problem in something like a flight-sim. The matrix shifting technique seems to be the way to go, but I think we've only learned this because of your entry. That's a backhanded compliment, by the way!

Nic: Unfinished obviously, but like the perlin noise heightmap approach alot. With the bugs ironed out, it has great potential. Would like to see how it converts to 3d.

CPU: Works well for what it is, and I guess it is an infinite terrain, just not in all directions. Could be very useful though for lots of different games - and top marks for the remarking, too!

Spooky: Looking forward to seeing it when it's ready!

Did I miss anyone?

And the winner is:



Runner up goes to


Well done to all, and good to see some new entrants.

It'd be nice to see the winners code used within a future challenge - like a flight-sim physics challenge, or a lunar lander game, for example, where infinite terrain would be handy.

Roll on the next challenge......

Segan
19
Years of Service
User Offline
Joined: 28th Aug 2005
Location: Canada
Posted: 8th Oct 2005 03:02 Edited at: 8th Oct 2005 03:03
Yay! Results!

Well good job all, and congrats TDK. Bring on the next challenge!

(oh and by the way, both of my last two posts are the same... It was just me fighting with the "edit" procedures again...)

As for camera angle, I have found that it works a little better if you only press one arrow key at a time. The camera feature was something I added at the end because my old camera didn't work (since the heights had got a lot more varied). I didn't get a chance to iron out the bugs in that one...

EDIT: Whoohoo! I no longer have to get my posts accepted, and can use edits! YAY!
TDK
Retired Moderator
22
Years of Service
User Offline
Joined: 19th Nov 2002
Location: UK
Posted: 8th Oct 2005 06:56
Oh bugger! I didn't realise that the winner had to choose the next challenge!

I'd have messed it up a bit more if I'd known. That'll teach me to not read the rules first.

Anyway, choosing the challenge is harder than writing the code...

Give me a bit to think about this and I'll be back with something.

TDK_Man

RiiDii
19
Years of Service
User Offline
Joined: 20th Jan 2005
Location: Inatincan
Posted: 8th Oct 2005 15:35 Edited at: 8th Oct 2005 15:36
Congratulations TDK!


Open MMORPG: It's your game!
TDK
Retired Moderator
22
Years of Service
User Offline
Joined: 19th Nov 2002
Location: UK
Posted: 8th Oct 2005 17:58 Edited at: 8th Oct 2005 18:05
Thanks Riidii!

OK, I've had an idea - just hope it appeals to a few of you. (It's something based on what I did a very long time ago, so it's not an original idea of mine).

The challenge is to produce a version of the classic game Frogger!

If you've never heard of Frogger before (unlikely), then there's an on-line version to try here:

http://www.80smusiclyrics.com/games.shtml

But, there's just one small rule you need to know about this challenge...

The ONLY commands you can use in your program to place anything on the sceen are PRINT and TEXT.

That's right... no 3D, no images, no sprites, no media. Just 100% ASCII!

You are allowed to use CLS and RGB, but anything else that appears on screen has to be printed.

I would think a week is sufficient for this challenge, but if you think this is a bit stingy, please let me know.

Otherwise, the closing date is set at Midnight Saturday 15th October 2005 (GMT).

Good luck!

TDK_Man

Tinkergirl
21
Years of Service
User Offline
Joined: 1st Jul 2003
Location: United Kingdom
Posted: 8th Oct 2005 18:27
Cool and novel! *laughs*

Well done, TDK
Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 8th Oct 2005 23:42
Great challenge TDK - I love it! One week is enough imo - and no excuses, people - everyone has time to do this one, surely!!

Darkbasic MADPSP
19
Years of Service
User Offline
Joined: 15th Jun 2005
Location: Uk
Posted: 9th Oct 2005 00:02
I'll try

(If i can get way from my platformer)

Platformer soon to be wip you'll like it
Peter H
20
Years of Service
User Offline
Joined: 20th Feb 2004
Location: Witness Protection Program
Posted: 9th Oct 2005 00:31
i'll try...if i can get away from dungeon 2.0, PakPerson 3D, ASCII Pong (C++) and the jigsaw puzzle program competition (C++ also)





"We make the worst games in the universe."

Zotoaster
19
Years of Service
User Offline
Joined: 20th Dec 2004
Location: Scotland
Posted: 9th Oct 2005 02:09
hey TDK, somehow I have a feeling youre gonna win again if im not mistaken you made that text rally thing right?


TDK
Retired Moderator
22
Years of Service
User Offline
Joined: 19th Nov 2002
Location: UK
Posted: 9th Oct 2005 03:43
Zotoaster

Apparently, the person who chooses the challenge has to judge the winner - and can't vote for him/herself.

That's why I chose the challenge - because I had done this sort of thing before.

Besides, winning isn't everything - it's the taking part.

TDK_Man

Ric
20
Years of Service
User Offline
Joined: 11th Jul 2004
Location: object position x
Posted: 9th Oct 2005 12:03
Where can I find that text - rally thing? Actually, come to think of it, I wrote one of those on my Acorn Electron about a million years ago. Don't think I've still got the tape, though - never mind the Electron itself!

Zotoaster
19
Years of Service
User Offline
Joined: 20th Dec 2004
Location: Scotland
Posted: 9th Oct 2005 12:29
check out TDK's site


sizer
20
Years of Service
User Offline
Joined: 11th Mar 2004
Location: texas
Posted: 9th Oct 2005 14:06 Edited at: 10th Oct 2005 07:14
would love to do this one, frogger is my fave atari game.
but dont have all that much experience with manipulating text.
also, took me 15hours to read this whole thread.
time for sleep. maybe will hop in the next compo if feasible for me,
school starts on sunday [today] when i wake up.
definitely looking forward to trying some of these.
----
[edit]
decided to enter in this one, but having dome problems with setting user to proper spot, any clue? and i can't get to my classrooms anyway.


-------------------------------------------------------------
AMD Sempron 2800+ : 768MB RAM DDR400 : ATI 9800PRO 128MB AIW
Whistlin Trev
19
Years of Service
User Offline
Joined: 11th Feb 2005
Location: Battle of Nevermore
Posted: 9th Oct 2005 15:41
this is a great challenge which have done a very basic test on and now working on making this better. here is the original test with no collision and it is very basic with cars only and you are the sheep.

sizer
20
Years of Service
User Offline
Joined: 11th Mar 2004
Location: texas
Posted: 10th Oct 2005 13:42
first off, apologies if this show up as a double post.
heres what i got sofar for this challenge.
only 219 lines of code now. still have to add collision.


-------------------------------------------------------------
AMD Sempron 2800+ : 768MB RAM DDR400 : ATI 9800PRO 128MB AIW

Login to post a reply

Server time is: 2024-11-23 15:51:02
Your offset time is: 2024-11-23 15:51:02