Aum48

Posted By: George

Aum48 - 06/01/05 04:35

Dear friends,

The newest magazine is uploaded. Get it from here:

Aum48

Please post your questions regarding Aum here.
Thank you.
George.
Posted By: Grimber

Re: Aum48 - 06/01/05 05:18

Great! thanks George looking it over now
Posted By: Nems

Re: Aum48 - 06/01/05 06:33

Oh man, I been hanging out for soooo long too.
Thanks George.
Posted By: immersionfx

Re: Aum48 - 06/01/05 07:14

Thank you once again George.
I was eagerly waiting for this!

One question that came to mind when reading Peter Churness's interview and, if not answered in this forum, I think it should be placed in a "Beginner's Corner" in one of your future AUMs:

Have you ever covered in one of the AUMs the idea of stacking objects and pushing/popping them on the right time? I mean how can this programming theory be implemented with C-script?

Thank you for another great AUM issue!
Posted By: FBL

Re: Aum48 - 06/01/05 07:28

Thank you, George. Some really interesting stuff in there.
Posted By: George

Re: Aum48 - 06/01/05 07:57

Thank you for the kind comments.
immersionfx: I thought that you are talking about objects like boxes, barrels and so on for a moment. Stacking entities can be done pretty easy using a starter function that includes a while loop. You can save the position of the entity (maybe the angles, the current frame, and so on) every x frames, and when the "distance" between player's current position and the virtual position of the entity is smaller than some value, you recreate the entity in the level. I might cover that in a future Aum article.
Posted By: immersionfx

Re: Aum48 *DELETED* - 06/01/05 08:07

Well actually I was referring to stacking variables, game states, etc.
Like, say, stack messages each time the player does something which needs clarification and display them on screen with a FIFO-type procedure.
Posted By: George

Re: Aum48 - 06/01/05 08:08

Please read my reply again. I've changed it.
Posted By: nuclear_winter

Re: Aum48 - 06/01/05 09:59

10X man!!!
Posted By: Nems

Re: Aum48 - 06/01/05 10:05

Finally I have a lightning script I can understand and am having fun with it right now with my chopper game, thanks again.
Posted By: Serlink

Re: Aum48 - 06/01/05 10:24

Thanks George I can use a lot of this
Sam
Posted By: fogman

Re: Aum48 - 06/01/05 22:54


Oh I havenīt seen it, new aum!!!

Thanks as always, George!
I really like your magazine, without it I wouldnīt
understand a lot of things

So, itīs not time for sleeping, yet.

greets, fogman
Posted By: Daedelus

Re: Aum48 - 06/02/05 03:03

The morrowing RPG is looking awesome so far!
Im really looking forward to seeing it develop further.

Posted By: Braxton

Re: Aum48 - 06/03/05 03:03

Awsome... I only thought, tho, that the models need to be bigger to fit the scenery. Other than that great!
Posted By: George

Re: Aum48 - 06/03/05 05:31

Thank you. Yes, you are right about the size of the models. I'll try to make the change and see what happens.
Posted By: Green_Grass

Re: Aum48 - 06/04/05 15:34

Thanks George, Morrowing RPG looks really nice!

I also want ask you, did you can update some things in your sword_combat script.
Example i need solve how i can make when player get sword (but without frames), i want place this sword in corect position in player hand, and also when player have this sword i want hurt enemys with this sword.

Thanks!
Posted By: George

Re: Aum48 - 06/04/05 16:43

Hi,

I plan to update the sword combat script sometime in the future; I'd need a warrior model with a separate sword first. An easy way to solve the problem would be to animate the sword and the player together, and then to save them as separate models. That's how the guard model from Morrowing works and most of its associated code should work fine for your project.
Posted By: Locoweed

Re: Aum48 - 06/04/05 19:46

Very nice work George.

The morrowing series is going to be very useful to alot of people. I think someone mentioned that the size of player, etc, are a bit small for the level geometry, other than that, it's coming along great.

Loco
Posted By: George

Re: Aum48 - 06/04/05 21:21

Hi Loco,

Thank you for the kind words. I am aware that the size of the player is too small, but hopefully that will be fixed in the following Morrowing episode.
Posted By: Lennart_hillen

Re: Aum48 - 06/04/05 21:53

Hhm... I'm a little more late then usually, but still, better late then never
Looks cool George As always I must add
Not many things this episode I can use, but it's always fun to read
May you bring out many many more mags

-Lenn
Posted By: Grimber

Re: Aum48 - 06/04/05 21:59

"An easy way to solve the problem would be to animate the sword and the player together, and then to save them as separate models"

I got one, but its Absolutly horrible to look at I don't even know why i keep it, its the first humanoid model i made. It has a sword and shiled animated to the models aniamtions. so you just need a vec_set(x x) vec_set( pan pan) and sync the animation frames

( he drops the sword in death animation sword goes flopping backwards over his shoulder to stick in the ground )

actualy the mesh isn't bad now that i look at it ( not great either), just the skin is terrible, and the animation frames need fixing because i mess up several vertexs ( somehow got ones didn;t intend to grab and they realy get moved bad)

animated for stand, walk, run, death, attack, jump, swim.started on a duck aniamtion but looks like I gave up there

if you want it i could try redoing the skin and fixing the animations ( hehe unless you wanted to do that )



Posted By: George

Re: Aum48 - 06/05/05 19:00

Hi Grimber,

Thank you for the generous offer. I was thinking to get a free quake model that has the needed animations from the net, but if I won't be able to find one I'll definitely ask for your help.
Posted By: Wade_Adams

Re: Aum48 - 06/05/05 21:37

Quote:

I plan to update the sword combat script sometime in the future; I'd need a warrior model with a separate sword first. An easy way to solve the problem would be to animate the sword and the player together, and then to save them as separate models. That's how the guard model from Morrowing works and most of its associated code should work fine for your project.




I believe the old sword code's major problem was that it was dificult to manage exactly how much damage was done and there there was a lot of traces done. Maybe a scan cone or do a trace at two points like venture did. I haven't tested these methods out but those have been my thoughts on the subject.
Posted By: Grimber

Re: Aum48 - 06/06/05 00:21

ive tried a few things a while back on a melee combat. Best i found was just totaly ignore trying to utalize an actual collision system aspect.

Instead just useing a simple vec_dist for a min distance check and if attacking entity is facing in a particular arc angle to the target an attack can be made ( animation )

then just fall back on a mechanical RPG system to calculate a % chance to hit based on character level, stats and modifiers vs target level, stats, armor. doing the check 'to-hit; once per animation cycle then will solve the major problems
Posted By: Wade_Adams

Re: Aum48 - 06/06/05 00:52

Believe it or not I did do something simliar to that for a turn based system I was tweaking around with, but in my case the attacker was always pointed the correct way and I knew ahead of time which enemy I was attacking.
Posted By: Darko

Re: Aum48 - 06/06/05 01:34

Great.I am going to download it now
Posted By: George

Re: Aum48 - 06/06/05 12:10

Another idea would be to check the position of a few key vertices on the sword; if you have set a few key vertices on the enemy model as well, you could even create damage depending on the body part that was injured. A few simple vec_dist instructions would do the job.
Posted By: Nems

Re: Aum48 - 06/21/05 22:47

Hello George,
I have been seeking to implement the mesh deformation from AUM 8 and whilst it works to a certain degree, there are some areas I would like to see clarifed.
For instance;
The deformation as shown in the AUM8 article works only on a terrain size you have demonstrated on.
Any variations in terrain size will return random results (it either will deform or wont, it will deform away from impact point etc.)
I have kept within the scope of the formulae like inputing the correct vertices count per terrain and played with the other number variables but cant get an acurate deformation at point of impact so I keep to the size your test terrain is.
Secondly, it seems that deformation is a visual effect only as I cannot get my player to enter into the cratered areas at all, it justs rides up over a ghost (or the original terrain)surface.
Have tried both c_move and ent_move, no differrence.

I am seeking to build a clone type "Scorched Earth" game and would really like some help on this one.

My question is, can you give another project example of terrain/model deformation article for A6 pleas,e seeking to resolve the above issues?.
Thanks.

*Edit
Discovered that by using 'normal' (vec_to_angle(my.pan, normal);) the ghost prob dissapears but cannot yet implement it correctly. All knids of crazy things happen to the position of the player or it wont initialise etc...
Also, because of event_impact, any object impacting will cause an empty pointer "Empty pointer in create_crate: vec_set(hit_coords,pocket.pos)"
I can get rid of the empty pointer by placing an if statement for the
existence or the projectile but any impact will cause a deformation to occur.
Have tried event_detect but no workable solutions yet.
Posted By: George

Re: Aum48 - 06/22/05 06:46

Hi Nemisis,

The problems that you are experiencing look like a minor bug in your code; the code from Aum8 should work ok for any terrain. Well, I was planning to write an article that deals with terrain deformation in the near future so now I've got another reason to do that.

I'll see you.
George.
Posted By: Grimber

Re: Aum48 - 06/22/05 07:16

nemisis 2 things im not sure if your checking or not but will have impact on mesh deformation.

be sure the terrain has the 'high pressision' checked in med on your terrain ( it should by defualt)

also right after you deform the mesh, you need to run
ent_fixnormals(entity,frame) on the mesh for properly collision after the deformation. it's a slow instruction because ti has to run through every polyface of the mesh and recaclualte the nomrals. so you need to give it a couple frames to make sure it's been completed
Posted By: Nems

Re: Aum48 - 06/22/05 07:33

Thanks George and Grimbar,
Will keep plugging away at it.
Grimbar, got a quick reply on how to implement 'fixnormals'?
Everything else is within parameters, no variations there.
If you guys like and or have the time, I could PM you's a link to the level.
Cheers.
Posted By: Nems

Re: Aum48 - 06/22/05 08:54

OK George, looks you you were right again.
I did a 68x68 terrain and paid attention to filling in the right value for the var and it works perfectly every time now.
Thanks for pointing it out.
Posted By: George

Re: Aum48 - 06/22/05 09:03

Just wanted to point out the solution for those of you that might be interested: "Maybe you have forgotten to change vertex_dist[1090] to a new value?". This is what I've asked Nemisis in an email.
Posted By: Grimber

Re: Aum48 - 06/22/05 09:09

ent_fixnormal is easy to use. just after you deform the mesh you call the function and you just plug in a pointer to that terrain and the frame number ( which for most terrains unles animated is usualy frame1)

looking at georges AUM 8 code I would place it in the deform function
Code:

function deform(vertex_number)
{
vec_for_mesh(temp, terrain1, vertex_number);
vec_scale (temp, 0.7);
vec_to_mesh (temp, terrain1, vertex_number);
ent_fixnormals(my,my.frame);
}



it but should solve most NORMAL problems with the terrain
Posted By: Nems

Re: Aum48 - 06/22/05 09:35

Thanks grimbar, jumpman mentioned something about a hullupdate function so am researching that if I can find it.
Cheers again and will try the fix you gave as I cant intergrate the 'normal'fix I mentioned previously.
Posted By: Grimber

Re: Aum48 - 06/22/05 10:41

seems like c_updatehull an updated version for use with the poly collision system

application of both seems identical. so depending on how your moving your entitys (c_move or ent_move) I'd give both a try.
Posted By: Nems

Re: Aum48 - 06/22/05 21:11

And out of them all, its the one that works for me .
My opponant drops as its land is shot out from underneath...coooool!!!
© 2024 lite-C Forums