D-Day dev ideas and how to make them work. (long read)

Topics relating to D-Day modeling/coding/skinning etc.

Moderator: Rab,d

Post Reply
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

D-Day dev ideas and how to make them work. (long read)

Post by Grim »

Some development ideas rolling around...
There are a lot of little ideas in various areas of the mod that i've seen methods and have had ideas on how to develop them. Some of these ideas have rolled around the forum before and ive developed a few methods to accomplish some of them.
Lets start with some ideas on the basic look of the game and work toward changes that effect the play and feel.

note - these ideas can be taken literally if you want but this is just some random banter to toss around and to enlighten people on a few things and how they work so that others can mabey take some of what is here and further develop it.

Weapon models.
Let's face it. the dday weapon models are the most ass ugly things in the universe.
I made some semi high poly but not really weapon models to replace them, most of which are done and most of which have been seen on this board.
What are the benefits? Well as needs not mentioning it makes the game look better, but more importantly it makes some other nice changes possible.
Standardized hands, gives it a more polished look overall.
Animation frames, I'd like to change the method of animation that the weapons use.
Currently i do not like how the idle animations cause warping of the vertexes, I think that a changing of frame order and number of frames (reduced) would go a long way in improving not only the look and feel but the overall performance as less frames means less memory used by the model and faster loading.
And now a look at a code aspect, creating a whole second set of animations in the model for use in truesight is messy in my opinion, I have wondered about other methods for this, things like making the weapons truesight animated only and hiding the gun when hipped, or another idea i have also considered is the use of the gunoffset and gunangles that are available. These functions could possibly be utilized to reposition the weapon to the hip (or ts) position without requiring extensive ammounts of extra animation.

Of course this leads me to another point. Idle animations. I don't think dday can afford to keep em without switching to a lossless model format.

Player models are currently the most complex and performance intensive models in the mod.
A few things that could be changed with them is the removal or severe reduction of excessive idle frames and point/taunt/wave/salute. Remove these animations and this greatly reduces the amount of frames needed for a player model. With less frames the models would take up less space or alternatively this frees up space for more animations that can be used for different things such as varied runs, reloading etc.

Of course to do this we would need new player models right? high poly? maybe. semi hi poly? SURE!
I have some candidate models for that as well!

Lets expand on this idea of new player models. The ent for the player uses the modelindex value for the player model and the modelindex2 for the gun. This leaves us modelindex3 and 4. These can be utilized in a number of ways. the most practical application of these that i can see and consider easily accomplished is to use them for the players helmet and gear. The gear would consist of the pack and belt items and possibly more. This would allow the different classes to carry different gear and have an actually different look past the use of skins.

Another thing to consider would be keeping the players from axis and allied in the same animation sequence, so that they could share the same weapon models on the player model instead of loading a whole new set of models filling up the model count. Also they could share in the helmet/gear idea.

Benefits? Better look overall of course, but also alot of extra functionality and the ability to make the different players have a unique look without the expense of creating a lot of player models. And with less overall to load the games performance increases.

K i just made supper so ima start writing this again.
...I had something to add to that last paragraph when i went to cook... damn train got lost in the woods... I'm sure it will come to me.
hmmm brb i cant think at the moment.

With new weapon models and new player models we get a better and more standard look, with the frame changes we get a nicer more efficient feel.
Since i just sat down again and I still can't recall where i was on that one paragraph ill just continue on.

With a new look we should probably work on the gameplay and various issues pertaining to that.
Lets start with the player.
Player movement, players run ridiculously fast. How does this affect the overall game?

Let's look at some benefits.
You can get around the level fast. You can do rambo assaults with fast and surprise attacks and you have generally more freedom of movement.

What are some poor side effects of this?
Teamwork is trashed. You are more likely to run off and do the rambo and forget you have a team.
You die more. You run at top speed and get to die even faster. Other people can kill you faster and easier.
Unrealistic, running at 30 mph and diving off a cliff with a bolt rifle~~~SUCH LIES

Possibilities?
Perhaps a better balance of weapon effectiveness versus speed.
give each weapon a more accurate weight value to affect the person using it.
Give weapons more realistic handling.
Add accuracy modifiers such as speed, velocities, position and or even stamina? Perhaps add a modifier that improves your soldiers effectiveness for playing in a more proper manner, such as a rank system. The longer your soldier stays alive and the better you play the better your soldier's performance.

Which leads into the weapons.
Bullet handling for the most part is great. It's simple and effective. Perhaps a few things can be trimmed/added to the function to improve performance and gameplay.
Bullet weapon ideas?
Ballistics, create accuracy functions and range functions to affect how the bullet travels/performs.
CODE STUFF-> Create a primary trace to the weapons effective range and use as the main bullet routine, create a secondary trace if the first trace exceeds the range. Perhaps even use a trace fraction variable per ammunition/weapon type to affect the final direction/accuracy/damage of the impact.
note: create material based impact in a similar manner of the footstep sounds sourcing the material. For performance boost use a limiter to prevent excess temp ents, no one will miss a few invisible bullet impacts.

o snap i just remembered what i wanted to adda few paragraphs up
For players.
The foot steps are currently handled by the client's think in the view stuff, I would like to move this to a server frame function and trigger the ev_footstep on the actual frame on the player model where the foot hits the ground. ps ill get back to this client think stuff in
ALMIGHTY CLIENT THINK!


Ricochets. you know you all want them.
shooting thru thin walls.
could be lame could be awesome. That one can only be known thru a practical application.

Tracers.
The hmgs shoot a lovely tracer, BUT on larger maps or even small maps it looks silly for a bullet to travel so slowly thru the air.
A solution to this is to use a normal bullet trace and give an origin at or near the impact site and place this in the tracers think
CODE STUFF-> Create the ent and give a forward velocity no more than 2000 because more than this causes prediction screw ups thanks to dum albeit efficient netcode. The tracer will spawn with its velocity for the server frame it was made in and will travel as such until it hits the next frame and moves forward to the origin from the trace. The space in between will be interpolated nicely and the effect you will see is the tracer moving as fast as a bullet to its target.
Once it reaches its target it resumes its 2000 velocity and rejoins the universe and will of course hit what ever its at.

This gives the illusion that tracers are as fast as bullets and skips the extra think frames in between meaning improved performance and better ent handling.

I wanted to get to some more weapon stuff but FIRST!

ALMIGHTY CLIENT THINK!
The client think is called for every frame the client uses (cl_maxfps).
Until all clients have frame/net seperation </pimp>LIKE R1Q2! http://r1ch.net</pimp>
The client think is going to kill servers. some servers cap the maxfps which is one measure that i approve of, but let's go to some cleanup and ideas as such
The client think atm handles some things like stamina, view functions, kick, weapons and such.
There are a number of things such as view functions and stamina as well as some jump and stance functions I want to move to a client ent think which would be once per server frame.
What are the benefits?
By moving as many functions as possible out of the client think you can save a lot of performance on the server and save bandwith as well as removing any ill effects/exploits/problems that a high maxfps can cause.
Downside?
people who boast that they can render q2 at 99076^345 fps will bitch that they cant waste the server's bandwith and or be lame mother fuckers.

Im sure ill add more to this later but lets get back to the wonderful weapon issues.

Explosives.
Grenades... sigh
When you get hit by a grenade explosions you will not immediately disintegrate and spray 27 YES 27 gib and blood spray entities.
Why? because its just a grenade not a rocket launcher. And 27 entities per dying victim of huge explosion radius can flood clients and give the WONDERFUL FUNDERFUL JOY OF OVERFLOW.
What to do?
Limit gibbing on nades and instead toss bodies around in an almost ragdoll fashion. looks good tbs.
Give nades an actual decreasing radius of damage instead of duck, duck, goose, GIB.
Also use an infront(targ, nade) to add a blindness. This makes nades useable not only as a weapon to damage but as a means to blind and confuse the enemy.

Dead bodies.
since players wont really gib on nades make dead bodies gib on nades.
as well dead bodies should disappear faster to once again improve performance.

Blood... Currently the bloodspray looks nice... but I think that it should be tweaked and scaled down to remove excessive performance bleeding(pun intended) and changed to albeit less dramatic but a better looking blood spray.

Vehicles.

Yes vehicles can be added and can work in dday. I have some tank models(PIV and Sherman) and the code to use them.
It's just a matter of hammering out how they would be used ingame and how they would handle.

RAWR!

SO MUCH MORE TO GET THRU! CAN WE MAKE IT ANY FURTHER?!? I THINK MY FINGERS ARE BLEEDING!?!
... that wasnt a question it was a statement of perturberance
ps i cant spell

Ideas on how to change the gameplay hmmm

The class function works well and really defines dday.
There are some variations changes possible that may be an idea to move to.

Class costs - make different classes require a certain ammount of kills forcing players to not only work together more as a team but would make players not so ready to sacrifice their snipers to a rush dive attack

weapon purchasing - Buying weapons and equipment using points from killing.... too cs?

Rank/levels - Improving rank and such would give the players access to better equipment as they accomplish more. This rank can be carried over levels

Structures - Buidling structures using a set number of team points so that players can purchase equipment from them?

In my opinion I like the class frag cost idea the best because it gives a certain sway balance to the teams and allows one team to move forward while another team is pushed back. It also attaches a value to your "life" so that you may not be so ready to die tragically and feed the other team.
Combine this with gloom like spawn structures and it may give dday a whole new level of fun.

Structure ideas?
simple, flags and ammo depots. barbed wire. and mines.
Who would build them? the engineer and medic.

Now with that you have an objective not only to defend but to destroy as well you can even add things like deployable mgs and pillboxes and etc

Airstrikes! and artillery.

Being able to call in either randomly would be nice. maybe a flare for airstrikes thrown like a nade, and leave the binocs for arty.
There are a few neat tricks that can be used to create a very nice plane bombing or maybe even a nice dive bombing.

which brings us to map tricks and effects to push the limits!

Map models to die for.
You can create some wicked terrain and use a model editor to modify it to a map model.
Why do this? once you have tweaked the map model which should fit nicely with clip brushes you can extend the map model past the skybox and create faux giant levels and beautiful enviroments.
if terrain is mapped and then placed as a model you can make wonderful terrain maps.
Make sure the actual map brushes are set to no draw so that they are not rendered.
There are a number of applications for this which are amazing, trees, grass, plants buildings etc etc. The simplest is of course terrain mountain/hills/buildings to line the map instead of ending at a sky brush or some other lame thing. One other idea is to have a giant church or large buildings that are too impracical to map out or a forest which would be too much to map out.

One thing to note is that the map models render and perform faster than actual map brushes and are smooth shaded!

Another thing to add is a number of mapping ents such as target_free to remove things, better sound ents, more functionality in other ents such as misc_md2.

blah blah blah still more.124,15jisd15

K k k k k

Here is another area that may be a bit much but HEAR ME OUT PLEAASEE!!@!

Monsters. These are not bots, they are dum single player monsters. they can use the same model as the players and be given simple monster code and placed thru levels.
This would give a wonderful opportunity to make a ddaysp or a coop map where players have to accomplish goals aggainst ai opponents

Bots. Yes the real deal respawning bad things(TM)
with names, weapons, personality etc.

The basis of these would of course be the monsters, but with expanded support and an extended think routine and with a waypoint system. These would be able to react to players and would try to play the map without having to see a player and or all the other things monsters are dependant on.

meowl

And the last thing I will touch on !FOR NOW!

Improved spawning functions for players. Give them and or mappers improved ability to do things like spawn forward or spawn closer to where they are needed.

o wait... lets touch some more

Melee....
I like melee and i would like to improve it both knife and fists starting with using the new hand models.

RAWR

Medics.
Medics should be fast until they pull out or pickup a weapon.
They should heal partially wounded people and be able to heal dead people..
HEAL DEAD PEOPLE WTF?!?!
I mean when you die, if a medic heals your body before the respawn timer is up then your team can recover the point you lost.

Also health boosts. A medic should be able to heal you up past 100 hp. giving them a usefulness past trying to heal those rare injured folk.

O yes... Morphine over doses to counter those lame self healing fgts.

and another small note.

when a player is hit and limping, change the run animation from frames 40-45 to 40-43 to give the actual look of limping, and since the footsteps sounds would be moved to frame dependant the footsteps would actually sound like limping too.

exceterawr

CONTINUATION

Once/if all/mostpopular/voted/mosteffective changes could be put in place then another aspect is hud and scoreboar. And menu gfx.

Create all that. Should be relatively quick and painless.

START-----------------SNIPPET
Standalone stuff
For a standalone would a standard q2 be placed in the dir?
or perhaps make a modified q2 geared to dday, this would allow a lot of functionality to be moved client side which would improve performance.
BUT......
....
...
..
.
It's probably easier to keep compatibility with base q2.
but package the standalone with a modified q2 engine to give an option for certain features such as better menus and hi quality texture support and perhaps some better effects but make it all optional so that the minimum q2 requirements are not exceeded.
The other reason for this is so not to take away from other engines such as q2max. I love you psychospaz...
END-------------------SNIPPET


What, Why? and sometimes how.

What.
A new standalone version of dday that is highly polished, looks new and performs better and may be more fun to play.

Why?
Free.
If all packaged together as a new standalone game with as many traces of q2 as possible removed we can make something that would be fun, small to download and very easy to use. Fast to run.
Free.
Less demanding and faster to get into than other games.
Free.
If a free decent looking WW2 fps thats small to download appeared on planetquake news than im sure I and many other people would jump at downloading it and then jump right ingame.
Free.
A lot of people who dont want the hardcore simulation but want more realism than say WOLFENSTEIN would start filling the community. The community would overflow with new and old players. People wouldnt require q2 to play either.
and
It's Free.


How.
all of the above >:X

and it's free.

-Grim o/~ all hail the seriously pissed of bunny from hell!
Image
User avatar
Luiso
Posts: 3377
Joined: Sun Oct 19, 2003 11:57 am
Location: Madrid, España.

Post by Luiso »

Class costs - make different classes require a certain ammount of kills forcing players to not only work together more as a team but would make players not so ready to sacrifice their snipers to a rush dive attack

weapon purchasing - Buying weapons and equipment using points from killing.... too cs?

Rank/levels - Improving rank and such would give the players access to better equipment as they accomplish more. This rank can be carried over levels

Structures - Buidling structures using a set number of team points so that players can purchase equipment from them?

In my opinion I like the class frag cost idea the best because it gives a certain sway balance to the teams and allows one team to move forward while another team is pushed back. It also attaches a value to your "life" so that you may not be so ready to die tragically and feed the other team.
Combine this with gloom like spawn structures and it may give dday a whole new level of fun.
NO! plz, no. If u want that, play gloom. really; dday is better like it is.

all the rest is ok.

cept 1 more thing: i find annoying that medics can heal themselves as they want. I'd like something like medic gets automatically healed when he heals some1. But not when he wants.
"[...]As I take hits from the bong[...]"

Image
User avatar
[CFE]BUTTHEAD
Posts: 525
Joined: Thu Nov 20, 2003 9:42 pm
Location: São Paulo, SP - Brasil

Post by [CFE]BUTTHEAD »

i think was good to make what we discost in shaef 2.0, and in jap team ,when ofcer cal lairstrike, an plain go to the target, hit it and blow, like kamikazes ;)
Rezo 1/3
Para arrumar 1/2
De te levar a 1/4
User avatar
bahamut0013
Posts: 763
Joined: Sun Jan 18, 2004 8:16 pm
Contact:

Post by bahamut0013 »

I have to mention something about the tracers....

Real tracer rounds go slightly slower than normal rounds, but not at a point that the human eye could recognize (like 10-15 fps, like saying the sleeve on a shirt is a milimeter too short). But the light it gives off goes much slower than the actual round. After about 100 yards, the bullet has moved a few feet ahead. Meanwhile the phosphorus that is glowing is falling further and further behind, and usually it all burns off (for a 5.56 tracer, that is) within 500-700 yards.

So DDay is actually kinda accurate in that manner, but the light is still too slow. about 90% speed of the bullet would work.
WARNING - HOLE IN FLOOR ABOVE DOORWAY BELOW UNDER FLOOR
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

bahamut0013 wrote:I have to mention something about the tracers....

Real tracer rounds go slightly slower than normal rounds, but not at a point that the human eye could recognize (like 10-15 fps, like saying the sleeve on a shirt is a milimeter too short). But the light it gives off goes much slower than the actual round. After about 100 yards, the bullet has moved a few feet ahead. Meanwhile the phosphorus that is glowing is falling further and further behind, and usually it all burns off (for a 5.56 tracer, that is) within 500-700 yards.

So DDay is actually kinda accurate in that manner, but the light is still too slow. about 90% speed of the bullet would work.
And thats exactly how the new effect looks
Image
User avatar
L.T. Mark
Developer
Developer
Posts: 775
Joined: Wed Mar 10, 2004 10:24 pm
Location: Denver,CO
Contact:

Post by L.T. Mark »

This all should be good for dday
Image

Pain is temporary but Immortally is for ever
User avatar
InFerNo
Developer
Developer
Posts: 4376
Joined: Mon Aug 25, 2003 11:07 am
Location: dday2
Contact:

Post by InFerNo »

I think mappers should make every building destroyable :D
and about a new menu, i made a menu once but it only works on clients that support PNG textures

i could try to get RScript to work decently and have the cloud effect like in the menu from RtCW: ET, since i stopped working on it a year ago
but i can pick it up again and try to build a new menu
i'll do a layout in photoshop for your convenience

edit:
Image
this could be done without RScript
If you want anything moving around, then you probably need a client that can see the RScript too
PS: does dday actually have a logo?
I would like to add a logo, if there isn't any i'll just write DDay: Normandy in a font on top
Image
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

good idea, some logo concepts should be thrown around.
I love the input inferno
and the menu is top notch
Image
User avatar
bahamut0013
Posts: 763
Joined: Sun Jan 18, 2004 8:16 pm
Contact:

Post by bahamut0013 »

Logo? The allied and axis dude has always been an unofficial one. The other was the pill from dday1.

But here is my submission. Excuse my sarcasm.
[img=http://img171.exs.cx/img171/7250/dday8cv.th.jpg]
WARNING - HOLE IN FLOOR ABOVE DOORWAY BELOW UNDER FLOOR
Mr. YOur no fun
Developer
Developer
Posts: 5496
Joined: Sun Aug 10, 2003 9:48 pm
Contact:

Post by Mr. YOur no fun »

I figured out how to make the building blow up, but using ploys. Perhaps one could have them blow up with a model inside, and then clip that area off so people couldn't get at it? At least with the polys you could crawl around on the rubble.

How is the Bulge thing going?
User avatar
InFerNo
Developer
Developer
Posts: 4376
Joined: Mon Aug 25, 2003 11:07 am
Location: dday2
Contact:

Post by InFerNo »

i did everything
the menu, with the transparant white selection thing, the background and all
but..
i can't find a way to get the menu to the upper left corner
can anyone look into that in the code?
i'm not good with the code
Image
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

For a standalone the exe can be changed (for things like menu)
just keep it compatible with baseq2
Image
User avatar
InFerNo
Developer
Developer
Posts: 4376
Joined: Mon Aug 25, 2003 11:07 am
Location: dday2
Contact:

Post by InFerNo »

well i did everything except putting the menu in the upper left corner
kthx

i'll upload it soon if i get webspace (which will be my BMC space next week)

edit:
i wanted m_main_back to have raindrops "on" the screen so it would look as if the picture was taken while the water is splashing around
but the problem is either they are transparant and they're basically just some 'things' shaped as a drop which you can look right through
or it would have a blue-ish texture so it's actually water

problem i wanted to do: i wanted the waterdrop to have the magnifiying effect of looking through a drop of water
if anyone has an idea please share it
i'm working on an idea right now and it just might not work :/
but it would be a cool gimmick

edit2:
to clarify, m_main_back is transparant and contains the ID Software logo which is on top of the actual background image for effects

i'll probably have to a pak for non-opengl-clients-which-do-not-support-png-textures too
Image
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

I mean for a standalone we can do anything we want for the menu.
Including moving things around.
Why include a basic q2.exe in a standalone when you can spiff it up!
Image
/
Developer
Developer
Posts: 793
Joined: Fri Jan 16, 2004 3:51 pm

Post by / »

Oh God I pray you every night for a better dday, thanks for making my dreams come true.
And just in case you don't hear me I pray to Satan too :mrgreen:
User avatar
Lewin
Posts: 2693
Joined: Thu Jan 15, 2004 10:47 am
Location: Bialogard , Polska
Contact:

Post by Lewin »

lol :D
User avatar
InFerNo
Developer
Developer
Posts: 4376
Joined: Mon Aug 25, 2003 11:07 am
Location: dday2
Contact:

Post by InFerNo »

anyone have a decent logo concept?
Image
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

k here is a quick animated stg44 teaser model
best with ggl and intensity 1
http://grim.moonside.org/dday/stg44.zip
Image
User avatar
Lewin
Posts: 2693
Joined: Thu Jan 15, 2004 10:47 am
Location: Bialogard , Polska
Contact:

Post by Lewin »

thank you god Image

bad pcx file :? but model looks great
User avatar
L.T. Mark
Developer
Developer
Posts: 775
Joined: Wed Mar 10, 2004 10:24 pm
Location: Denver,CO
Contact:

Post by L.T. Mark »

Nice stg44 god
Image

Pain is temporary but Immortally is for ever
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

not meant for the pcx
its meant to use skin.png
Image
User avatar
Lewin
Posts: 2693
Joined: Thu Jan 15, 2004 10:47 am
Location: Bialogard , Polska
Contact:

Post by Lewin »

so?
Grim
Developer
Developer
Posts: 790
Joined: Fri Sep 19, 2003 2:04 am

Post by Grim »

so what?
Image
User avatar
Lewin
Posts: 2693
Joined: Thu Jan 15, 2004 10:47 am
Location: Bialogard , Polska
Contact:

Post by Lewin »

what i must do if i want see skin? :D
User avatar
SHINY
Posts: 2599
Joined: Thu Jan 22, 2004 5:29 pm
Location: the netherlands
Contact:

Post by SHINY »

good stg but i dont beleef its done right
if u aim the gun is so thin and if u aim and then shoot the gun flys up and if u aim u can look true the gun and then reload is not done i think
Image
Post Reply