[Omniscaper]

Bobboau, can we bribe you or something to elevate FSO to Halflife 2 or Doom 3 standards? I can hear the REGRESSORS now. I'm willing to make the first donation.

[Flaser]

No need for bribes, the man is a saint already set an aim:
Outdo HL2 and Doom 3.

Status: 17% - Matsystem.
            15% - Multimapping (glow, shine, evn)
             7% - Superdetailing

[Nuke]

i updated my drivers and it kinda sorta works now. still i havent got to messing around with it yet.

[Kosh]

Wait a sec. If this involves stuff needing to be added to the tables, then wait about old campaigns that use their own tables? Is this not backwards compatible? Or did I just not interpret everything correctly........

[Nuke]

i figure old stuff would use the default materials set, else the campaign deigners could make a materials patch to take advantage of the new eye candy abilities of the engine.

[WMCoolmon]

Beyond that it should be possible to just stuff a materials.tbl in the mediaVPs, rather than hack it in. Pretty much everybody should have mv_core, and the only reason the materials.tbl would be overridden is if a campaign were to take advantage of it.

[Raptor]

Still not really following this...

So we could do the following?

1) Assign 'Armour Types' to different ships
2)Have said Armour types be really good against some weapons, but poor against others

That's about all I can follow...

[Bobboau]

that's the armor thing WMC made, the material system is more about makeing things look good.

[Vasudan Admiral]

'nother question:
In terms of variables, we have ":global.time:" atm. Are there/ will there be others we can use to base calculations off?

Ie, velocity in certain directions, fred variables, ship damage, global location, light hitting object (ouch), distance to an object, ship orientation and many others i just can't think of right now.
All of those could quite nicely be used in various circumstances, such as basing light flickkering levels (or frequency of the flickkering) off the ammount of damage the ship has sustained etc.

And while i'm here: triggers. How will they work, and will it be possible to link an animation sequence to them?
Some things i can think of that would be awesome are: weapon links (ie, triggered by firing the associated weapon), object/weapon impacts, fred variables or sexps, docking process, object destruction (ie, only triggers when the object blows up - fancy non-standard explosion effects or special debris etc), warpout sequence and again many others that i can't think of.

Once people (myself included) learn how to use this to it's full potential, i think the results will be incredible.

[Bobboau]

there will be many many many game variables, I already have localy access to an object's position and orientation, as well as the eye/view pos/orient. adding in any of the variable you have mentioned would besimple and will probly be done anyway. the idea is to give access to as many game variables as posable.

the triggered animation system has a totaly diferent way of working. it has basicly a newtonion simulation running and when an event happens it's told to move an object in one direction or another. I realy had thought that by now a lot more would have been done with this, adding animation triggers is (in it's self) a very simple process, the only tricky part is figureing out were to place the few lines of code to add a trigger. as far as the variables here are concerned, it would probly be relitively easy to grant access to some of the variables involved. and a seperate animation system that uses variables to define object orientation is something I've had in mind for quite a while now.

I have implemented looping and conditional passes recently, I also added in an optional additional translation matrix (mostly for scaleing of thrusters but can be used for other things), this offers some very interesting posabilities. with these tools you could (if I hadn't already hard coded it) make weapon models that draw at each of the weapon points through the materials system.

[Vasudan Admiral]

there will be many many many game variables, I already have localy access to an object's position and orientation, as well as the eye/view pos/orient. adding in any of the variable you have mentioned would besimple and will probly be done anyway. the idea is to give access to as many game variables as posable.

That's awesome. Just give out a list of implimented variables when they're done, and i can't wait to use them.

In regards to the triggered animation systems, sorry, i think i mislead you there :\
I was still refering to the materials system rather than the sub-model one - ie, material effects that are only brought about once triggered. Pretty much like the current sub-model animation system triggers, except instead of moving something, it would initiate some sort of material change.
A really basic example would be to have a sexp fire after 20 seconds, which triggers a 'scripted' (mathamatically) material animation, such as gradually lowering the alpha values on difusse map of a ship, starting at the time it was triggered rather than as soon as the mission loads.


And don't worry about the sub-model animation code not being used. I was performing a rather interesting experiment the day it came out, as i'm sure many others were/are.
Within that same day i managed to get this:
http://www.geocities.com/vasudanad/New_OrestesDockingClaw.zip
I didn't release it then and am reluctant to release it now due to the terribly incomplete-ness of the whole thing, but since the docking ship there is going to be delayed in texturing, i figure i may as well post this now.
I was hoping to post a better one with both ships completely HTLed and with no glitches, such as some better docking clamps that were actually timed properly and looked decent, but the effect is still there in that one. So ignore everything but the animation.
(I also have a pair of far far far more extravagant sequences completely planned out, but they're for TI. )

And submodel scaling matricies you say? Ooooooooh......
*scurries off to replan a few things*

[Black Wolf]

Argh. Your geocities account is out of bandwidth by the looks, and I don;t have space to mirror that ani anywhere Which sucks, because it's rather cool.

[Vasudan Admiral]

arrgh, i was afraid of that. :\

*wants Sectorfiles back*

[DaBrain]

"save as" works fine for me.

Looks like the file is still there.


Edit: Wow, that's cool.

[Taristin]

If Raa recieves the file in an email, he may upload it to game-warden.

[Bobboau]

keep in mind that the geometry matricies will only effect what they look like, not what happens with colisions.

and you can make any/all passes conditional, so if there is some mathematical expression that will give a zero for everything that you don't want to draw, that's already implemented.
(you can also make looping passes and go through a serese of elements)

[WMCoolmon]

How are looping/indexed passes done (Just a line or pointer to another post is fine, I just want to see how it's done as it may be useful for things like the escort gauge...I'm not sure when I'll get back to coding work, next few weeks are gonna be busy...but maybe I'll come up with some good way in the next couple of weeks or something.)

[Bobboau]

there is a for loop and rather than useing 'i' it uses a mamber of variable_call_data that I added for that purpose, and I made another expression determine how many times to go through the loop, if that expression is NULL, it does it once, otherwise it does it as many times as that expression says.
although, I supose I could even go further and make an expression for every part of the for loop, the initaliser the comparison and the intrement step, just in case, but what I have now is more than enough to get done with most of the cases that will ever show up and the statement is rather ugly as it is.

Code: [Select]


for(call.iteration = 0; (mat->pass[p][get_environment()][0].loop)?call.iterationpass[p][get_environment()][0].loop->value(&call):call.iteration<1; call.iteration++)

//use quote, HTML parseing screws this up a little

[Bobboau]

ok, tonight's party time funfunfun was rewriteing the stupid assed way I was handleing model loaded textures and adding inline conditionals to the expression system.
it works similar to the c style inline conditional
()?():()
exept sence we already are useing ':' for stuff I had to use '|'
so it's like (:condition:)?(:true_action:)|(:false_action:)

example:
(:object.engine_output:>0)?(:object.engine_output:)|(1)
will return 1 when object.engine_output is <= 0 and it's value otherwise.

so... good for me.
now off to use this to put the warp effect's scaleing behavior into an expression wich is what initaly ent me off on this tabgent.

[wow smileies fuk'n ate this post alive]

[edit]
now that I think about it I wanted '|' for logical OR

the only common charicter left that I can see is '@' :/, I guess I could use '_'. anyone got any ideas here? it can only be one charicter

[Black Wolf]

Perhaps "&"?