Reviving Transfusion...somehow..

General discussion relating to the Transfusion project.

Moderators: Slink, General Discussion Moderators

User avatar
I Live...AGAIN
Cabal member
Posts: 762
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Postby I Live...AGAIN » Tue Apr 13, 2010 07:34 pm

Back from my meeting:

OK, I am ignoring Kurt's above post because it should be deleted soon. Also because he is a mental midget....

Now regarding an engine switch:

While I was never a huge fan of Darkplaces in the past, it was mostly because I love multiplayer and the code we currently use in Transfusion is buggy and laggy. Now I have played Nexuiz and the MP there is great. What is the difference between that code and what is in our beta? I dont know because I am not a coder.

The team has already done alot with Darkplaces to mold it into closer resembling Blood's physics, crouching was another huge step. It is truly a beautiful engine and can handle everything we need to implement so why change it at this point?

Unless there is a "team" of coders comfortable with a different engine and are prepared to do all the work necessary to make a switch and not give up in the middle I say we need to stick with what we have.

We just talked about shortening our goals and making them attainable, switching engines is ALOT of work, exactly the opposite of what we are trying to accomplish.

I'm not trying to say be cheap or lazy, but why make things harder than they need to be?

Also, try playing Quake, then play Blood. The 2 games are not similar at all. The physics are totally different, the environment is much more interactive in blood, blood plays much quicker. Blood requires crouching, it's a necessary part of the gameplay.

Try playing Quake and Blood side-by-side, you will see what i mean.

Here is a good place to get starter info for Darkplaces also:
http://dpwiki.slipgateconstruct.com/index.php?title=Main_Page
User avatar
Fiend:Kamikaze
A LUMBERJACK!
Posts: 114
Joined: Thu Aug 12, 2004 03:29 pm
Location: United States
Contact:

Postby Fiend:Kamikaze » Tue Apr 13, 2010 09:34 pm

In terms of gameplay, speaking of Quake vs. Blood, I feel the same issues are raised when it comes to Doom vs. Blood, as id's games all have similar gameplay. id tech games are clunkier than Build was. Build's gameplay was very, very smooth, especially in Blood. The way the player moved is something very difficult to pin down.

Jumping was exaggerated in all of the build games, and in Blood, unlike Quake or Doom, gravity was less harsh. In Doom ports and in Quake, jumping may have raised you about 32 pixels. In Blood, Duke, and SW, it is much closer to 64.

Another thing we recently addressed in HT was view height. In Doom, standard was 32, and I believe Quake also has something similar. Knocking the view height up to 48 was more Blood-like. I remember reading a thread (possibly on here, possibly on PostMortem), where someone was trying to convert the Duke3d Shining TC over to Blood. When doing this, the level came out the same, but Caleb is noticeably taller than duke (or the view height is) by at least 16 pixels.

Blood's use of view bobbing/view swaying also had a great effect on how it was perceived. Think about when Caleb lands after a jump: The view smoothly transitions down to signal the impact on the ground, then up again, down and up, eventually evening out.

All of this contributed greatly to Blood's feeling but it seems like so much of it was a product of the Build engine. The Build engine seemed to use resources in a vastly different way than the way FPS games were (Doom/Quake) and how FPS games have evolved. Makes me wonder if DNF would have resembled a modern build engine game and if it would have used it's technology (supposedly it had a new exclusive engine) in the same way.

Build was a fundamentally 2D engine. It may have been 2.5d and faked 3d effects, but it was solidly grounded in two dimensions and the use of gravity and physics were not part of a 3-dimensional system, but rather were like hardcoded workarounds for a third dimension, so they just don't work the same way.
User avatar
zZaRDoZz
Acolyte
Posts: 416
Joined: Tue Apr 29, 2008 03:13 pm
Location: USA
Contact:

Postby zZaRDoZz » Wed Apr 14, 2010 12:15 am

ILA:
The team has already done alot with Darkplaces to mold it into closer resembling Blood's physics, crouching was another huge step. It is truly a beautiful engine and can handle everything we need to implement so why change it at this point?


That's an excellent point. DP is so developed that it might actually be the most efficient port out there. The DP specific coding that's been done isn't something that can be simply thrown away either.

ILA
Unless there is a "team" of coders comfortable with a different engine and are prepared to do all the work necessary to make a switch and not give up in the middle I say we need to stick with what we have.


Lets not consider the abandonment of DP as the port used even a "proposal" right now. Remember the only valid reason for leaving DP would be to force smoother gameplay. There is absolutely no evidence as I write this that that would indeed be the case. A request that other ports be explored is only a request for information, for hard numbers. No one has these right now.


ILA:
We just talked about shortening our goals and making them attainable, switching engines is ALOT of work, exactly the opposite of what we are trying to accomplish.
I'm not trying to say be cheap or lazy, but why make things harder than they need to be?


Making things harder than they need to be is the exact opposite of what a mini project would be for. Anything that adds a tremendous amount of difficulty or reworking should be abandoned by default. Remember there's a large learning curve that our potential volunteers have yet to overcome. In addition to that, they have their own projects for now.

The only reason to 'slim down' as you might say is to get frame rates at fast levels and get transfusion back on it's feet. It might not have anything to do with the port being used. Tfn maps have some huge vertice counts and large open spaces. Another port may offer nothing to solve such issues. It may be a matter of reducing map complexity.
The most imortant thing to remember is we can't treat this as "the next step" in completing transfusion. Transfusion's reason for being is ultimately a blood reproduction. I don't think using standard blood maps is such a good idea due to those high vertice numbers. Playability has to be central to the goal for a side project. Finding new people is going to be all the more difficult if they're left with the impression that maps have to be so accurate that hardly anyone will be able to play them at more than 18 frames a second.
User avatar
zZaRDoZz
Acolyte
Posts: 416
Joined: Tue Apr 29, 2008 03:13 pm
Location: USA
Contact:

Postby zZaRDoZz » Thu Apr 15, 2010 12:11 am

Me:
The most imortant thing to remember is we can't treat this as "the next step" in completing transfusion. Transfusion's reason for being is ultimately a blood reproduction. I don't think using standard blood maps is such a good idea due to those high vertice numbers.


A correction, A mini project would be the next step. It would have AI as it's focus, seeing as coders would be in abundance if they join. Simple sp maps with a quake like level design would be preferred.
User avatar
I Live...AGAIN
Cabal member
Posts: 762
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Postby I Live...AGAIN » Thu Apr 15, 2010 12:59 am

I'm not sure I understand why you would want quake-like levels? Transfusion is Blood, it has nothing to do with Quake.

I'm afraid I dont know much about AI. What are our options there? What can we do and what cant we? Does Darkplaces have any AI improvements over vanilla Quake? If I remember right the multiplayer bot AI could be tweaked and routed with some work but I have no idea about monster AI.

I'm guessing one if the biggest challenges would be to get a zombie to come out of the ground when it spawns.
User avatar
DustyStyx
2D Artist
Posts: 2069
Joined: Wed Aug 11, 2004 02:12 am
Location: Salt Lake
Contact:

Postby DustyStyx » Thu Apr 15, 2010 01:00 am

Why "Quake like"? That is a broad brush. The focus should be Blood after all.
User avatar
DustyStyx
2D Artist
Posts: 2069
Joined: Wed Aug 11, 2004 02:12 am
Location: Salt Lake
Contact:

Postby DustyStyx » Thu Apr 15, 2010 01:03 am

The reason frikbot got verry dumb with our q3 maps was because they were designed for Q1. Again, Nexuiz has already over come these issues.
User avatar
zZaRDoZz
Acolyte
Posts: 416
Joined: Tue Apr 29, 2008 03:13 pm
Location: USA
Contact:

Postby zZaRDoZz » Thu Apr 15, 2010 01:16 am

Yes, it is a bit vague. What I really mean is low vertice count. Blood replication is the ultimate goal. Still I'm guessing that a emphasis on AI in the short term with simple, smooth running maps would have a quick turn around time. Having maps that run on nearly all systems would mean lots of downloads, lots of testing, and far more interest being generated.

When ai/weapons coding is into the final stages of beta our alleged coders might have far more experience. They could move on to networking and other issues. By that time a more advanced DP or other port might have made an impression, giving the Tfn team more options to proceed and an engine with greater capacity for blood's intricate maps.
User avatar
Liberation
Axe Zombie
Posts: 21
Joined: Fri Apr 09, 2010 06:25 pm
Location: United Kingdom
Contact:

Postby Liberation » Thu Apr 15, 2010 07:17 am

From what ive seen so far alot of the monster code is already in transfusion, Ive even seen some code for the zombies coming out the ground etc. As for AI that is something else to look at but we could maybe get away with using q1 type AI for the moment.

Ive got the cultist model loaded (well i just added the .psa to the .PAK) since the code from the SVN looks to be in the middle of a rewrite when it comes to .psa / .md3s. I've updating the Butcher code to take the .psa atm and looking into animations.

Also i've serched high and low for the HUD code with no succsess and im wondering if its in the DP engine itself.

There a few other things but ive got to get to work so i'll get on later .
User avatar
I Live...AGAIN
Cabal member
Posts: 762
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Postby I Live...AGAIN » Thu Apr 15, 2010 10:45 pm

Did a quick search and found this post from Willis:

Willis wrote:The code is completely engine side at this point, and is not in CVS. Its not in CVS for two reasons. 1) It's still incomplete, and 2), its not properly merged (and currently somewhat broken). Also, there is debate about scrapping it out for a csqc version once csqc is ready. It is fully activated by the -game transfusion switch, but you muse use a the released transfusion executables for now since it isn't in DP CVS. The code will certainly be released in the event of a TFn release, but until then, I need to merge and fix it before I can really post it.



Hope that answers some questions....
Here is the link to that thread:
http://forums.transfusion-game.com/viewtopic.php?t=1349&highlight=hud
User avatar
Liberation
Axe Zombie
Posts: 21
Joined: Fri Apr 09, 2010 06:25 pm
Location: United Kingdom
Contact:

Postby Liberation » Fri Apr 16, 2010 05:45 pm

That would explain why i cant find it :D

Also does anyone have a up 2 date .fgd file for worldcraft? or does the the 1 in the sdk have enemy enterys.
User avatar
DustyStyx
2D Artist
Posts: 2069
Joined: Wed Aug 11, 2004 02:12 am
Location: Salt Lake
Contact:

Postby DustyStyx » Sat Apr 17, 2010 03:22 pm

Just as a recap from some of the above...

Item : Possible Solution
    Multiplayer : People actually need to do this to test and be sure it's right

    Game physics : video comparison
    * Jump hight
    * Crouch level (ability to squeeze into very small places)
    * View bobbing
    * Movement timing (faster strafe forward speeds, how fast to spin around)

    View hight (dead center in Quake, higher in Blood)

    Values extracted from Blood : Some can be taken from Blood's rff files (Build moding communities may know other tricks.)
    * Weapon damages
    * Monster characteristics


I'm sure there are more things we need to focus on, but Friend's post reminded me a lot of the complaints we've had about Transfusion game play in the past.

Imediate Concerns: If we are to actually get some headway and quality feedback for Liberation and eventually Willis to work with, we will need some dedicated testers willing to collaborate and communicate with each other. In the past people have fractioned off, doing their own thing, and team cohesion has suffered to the point where it has been for the last few years.

Something quick and dirty that might get Transfusion up to speed quicker would be to just start with the Nexuiz code and import the appropriate Transfusion values. We would still have to test the end result to get it as close to Blood game play as possible, but it would bring Multiplayer up to speed in short order.
User avatar
I Live...AGAIN
Cabal member
Posts: 762
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Postby I Live...AGAIN » Sat Apr 17, 2010 06:23 pm

May I suggest a bi-weekly(or monthly) ICQ session like we used to do? Lets try to get everyone pulling in the same direction.
User avatar
Liberation
Axe Zombie
Posts: 21
Joined: Fri Apr 09, 2010 06:25 pm
Location: United Kingdom
Contact:

Postby Liberation » Sat Apr 17, 2010 06:47 pm

Ok im getting the gist of the model code but the problem is that the models are incomplete, some parts are .psk format but then other parts are md3's, also skins are missing.


BTW, the Zombie was already fully coded with ai as well, hopefully i be able to use this to understand the new style animations, also from default he comes out of the ground!!
User avatar
Slink

Not to be a dick, but...

Posts: 1912
Joined: Mon Aug 16, 2004 04:42 am
Location: Niagara County, NY

Postby Slink » Sat Apr 17, 2010 09:26 pm

IRC sounds good.

Return to “Transfusion Discussion”

Who is online

Users browsing this forum: No registered users and 3 guests