Is it dead enough now?

General discussion relating to the Transfusion project.

Moderator: General Discussion Moderators

User avatar
predator
Lead Mapper
Posts: 1100
Joined: Sat Aug 14, 2004 12:51 pm
Location: Warsaw, Poland
Contact:

Post by predator »

Models are fine in either format.
The format change would mean quite a lot of work for us mappers, but for other reasons.
We would also have some problems to solve, like flickering lights for example - they are not present in the standard Q3 bsp format.
User avatar
I Live...AGAIN
Cabal member
Posts: 810
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Post by I Live...AGAIN »

Mathieu,

I think my point didn't come across very well, I AM hinting that an engine change might be necessary because we cant seem to code what we want in the engine we are using, again, as always, we are TOTALLY dependant on coders that are familiar with whatever engine. There are obviously more Q3 coders out there than DP or Q1 coders.

Also on a side note, I still have major concerns about netcode in DP/Quake....
User avatar
I Live...AGAIN
Cabal member
Posts: 810
Joined: Fri Aug 13, 2004 05:33 am
Location: Detroit
Contact:

Post by I Live...AGAIN »

Oh, and on another note I just thought of that is also "killing" this project :D


We are spending entirely wayyy too much time with stuff that dosen't need to be addressed with this project. I thought this project was to re-create Blood as close to the original as possible. We continue to be sidetracked by new goodies like bumpmapping, real-time lighting, re-texturing....etc

All that stuff is nice but it does nothing but make more and more work making this project an even more insurrmountable task. I have said all along we need to keep things nice and simple, get physics right, get the environments correct, weapons as close to original as possible, get a decent looking set of models that are as original looking as a 3d from 2d model can be and lets go....... all the pretty stuff can be done later on or by someone else. Everyone loved Blood for what it was, this is turning out to be something besides a re-creation of Blood

2 things IMHO that will make this project successful:

1) A "Dead Nuts" re-creation of the first episode(Shareware) of Blood. Dead nuts meaning "as close to original as humanly possible" Like I say above,,, keep it simple, the rest can follow...

2) A working internet friendly version of multiplayer. (everyone who played Blood loved multiplayer but on the internet it sucked, I'm sorry to say that TF on the internet sucks as well. If we can fix that it will help draw more interest.)
User avatar
scar3crow
kvlt-anaatmanah
Posts: 386
Joined: Mon Aug 08, 2005 06:22 pm
Contact:

Post by scar3crow »

From what Ive observed, DP seems to be a fine engine, the only problems being the hull concern with Q1BSP format. Which apparently this could be solved with either MD5 support (which isnt added yet) or using Q3BSPs, which are implemented. My problem with Q3BSP is, and I may be wrong on this, or it may be the Quake3 engine and NOT the bsp format that is doing it, but the lack of light changes. Quake3 uses very very static lighting. And by very very static, rockets dont illuminate, they only project overbright textures (firing a rocket down a dark hallway or into shadow in Q3A, unlike in Q/Q2, it doesnt light up the area).

Like I said, I dont know how much of this is defeated by DPs loading of Q3BSP. I dont know a lot about formats. But DP is very different from stock Quake and Quake3. Also I do know that Q2BSP has rather dynamic lightmaps (I can think of many ocassions of such in Q2) and also makes use of crouching as well.

Coding wise, I see more work to be done on the QC aspect. I mean sure, MD5s arent implemented in DP. And you cant duck. Im more concerned about weapons having the right damage values, cultists that actually roam and throw dynamite or shoot the appropriate shotgun or tommy gun, and gargoyles that swoop. I would much rather have a menagerie of functional flesh and blood seeking my destruction then merely having the ability to duck, and still being behind on those much more vital things - that which comprises the gameplay of the game that this project was intended to emulate in the first place.

Then again, I like DarkPlaces and I like the progress I have seen done on Transfusion. But its not just engine coding it needs. I still stand by my belief from back in 2001 that most of Blood's gameplay could be emulated through straight up non-engine enhanced QuakeC. It would look ugly as balls, but a lot of it would be playable and it would bring more of Blood to modern machines then just crouching or using more recent sounding tech could ever do.

You can crouch all you want, I would much rather press barrel to flesh amongst hordes of the undead - even if Im forced to dodge a napalm round by strafing instead of ducking.
User avatar
scar3crow
kvlt-anaatmanah
Posts: 386
Joined: Mon Aug 08, 2005 06:22 pm
Contact:

Post by scar3crow »

Didnt see Joes amendment there.

If you allow gibbing, Blood shall run.

Im all about not re-texturing in some aspects. Though Kazashi and Cruaich have proven themselves at visuals (excuse me while I go drool over the shotgun some more), retexturing is a massive enterprise that at best partially compromises the total asthetic given by the game.

What helps make Transfusion look like Blood right now, is the fact that I am seeing those good familiar textures all over again. They match each other and it works.

What Transfusion needs is QC that makes the gameplay. I need to play Blood, everything else is secondary.
User avatar
Necrosis
Bubble Wrap Man
*not Canadian*
Posts: 1242
Joined: Thu Jul 21, 2005 01:38 am
Location: In The Flesh
Contact:

Post by Necrosis »

I Live...AGAIN wrote:1) A "Dead Nuts" re-creation of the first episode(Shareware) of Blood. Dead nuts meaning "as close to original as humanly possible" Like I say above,,, keep it simple, the rest can follow...

2) A working internet friendly version of multiplayer. (everyone who played Blood loved multiplayer but on the internet it sucked, I'm sorry to say that TF on the internet sucks as well. If we can fix that it will help draw more interest.)
Yeah... The way I see it (in one of my rare optimistic bursts), if episode 1 is finished, people might start giving us more attention, and that will mean a larger community, and with the larger number of people knowing Transfusion, we'll have more coders, mappers, etc. joining us to ressurect Blood. As for the netcode sucking, totally agree.
Guest

Post by Guest »

My views about the main pros and cons of switching to Q3 (versus sticking to DP):

+ stable and complete engine. No feature to wait: it's there and we can use it, or it's not and we do without it. (yes I think it's a pro. We've lost too much time waiting for features IMHO)

+ better (or at least more proven) netcode

- the whole game code to rewrite, from a base code with a nice bot AI, but no single player. That's a huge task and probably the biggest problem given the lack of coders

+ Q3 VMs are more powerful than QuakeC progs. No doubt we could do cooler things without having to find hacks. Also, it's probably easier to find Q3 coders than Q1 coders.

- from a graphical standpoint, DP is superior. No doubt we will miss that

+ Q3 shaders may be a cool thing to have. But I'm not an artist...

- switching to another engine is a risky game. In January 2003, we tried to switch to Q2. As you can see, it was a complete success... :?


If I was in a position to really help the project as a coder (which means if my work let me more free time), I would push much more towards a switch because I think in the end it'll be a win. But I'm not, and so I can't reasonably give this advice while the project has so few coders.
Guest

Post by Guest »

Guest again?! :shock: My God, time to take some sleep I guess... Sorry again.
User avatar
Elric
Cabal member
Posts: 166
Joined: Fri Aug 13, 2004 07:35 am
Location: Paris & Limoges, France

Post by Elric »

Guess what? It was me. :roll:
User avatar
Necrosis
Bubble Wrap Man
*not Canadian*
Posts: 1242
Joined: Thu Jul 21, 2005 01:38 am
Location: In The Flesh
Contact:

Post by Necrosis »

Elric wrote:My views about the main pros and cons of switching to Q3 (versus sticking to DP):

+ stable and complete engine. No feature to wait: it's there and we can use it, or it's not and we do without it. (yes I think it's a pro. We've lost too much time waiting for features IMHO)

+ better (or at least more proven) netcode

- the whole game code to rewrite, from a base code with a nice bot AI, but no single player. That's a huge task and probably the biggest problem given the lack of coders

+ Q3 VMs are more powerful than QuakeC progs. No doubt we could do cooler things without having to find hacks. Also, it's probably easier to find Q3 coders than Q1 coders.

- from a graphical standpoint, DP is superior. No doubt we will miss that

+ Q3 shaders may be a cool thing to have. But I'm not an artist...

- switching to another engine is a risky game. In January 2003, we tried to switch to Q2. As you can see, it was a complete success... :?


If I was in a position to really help the project as a coder (which means if my work let me more free time), I would push much more towards a switch because I think in the end it'll be a win. But I'm not, and so I can't reasonably give this advice while the project has so few coders.
1. Indeed. Doesn't Q3 already have crouching? And I'm not aware of any other big things that needed implementing, but I'm sure it would be easier with Q3. *doesn't own Q3, unfortunately*

2. Not much to say here, but you're right.

4. I have no information on this except that it makes sense and that it is very likely easier to find Q3 coders. You rarely find people making UT99 mods when there's a far superior (in terms of graphics and compatibility and such) UT2k4 to make stuff with.

5. Well, it's been previously stated that we are not going for a graphical overhaul of Blood to put it on par with Doom 3, we are going for a recreation of Blood so that newer PCs can run it. I think that was Kazashi, but I can't be sure.

And that's about all I have to say on the subject. Also, I think it would be a bad idea to lock this because it lowers morale or something, we're getting some productive stuff out of it.
User avatar
scar3crow
kvlt-anaatmanah
Posts: 386
Joined: Mon Aug 08, 2005 06:22 pm
Contact:

Post by scar3crow »

Elric wrote:
+ stable and complete engine. No feature to wait: it's there and we can use it, or it's not and we do without it. (yes I think it's a pro. We've lost too much time waiting for features IMHO)
I still say, code the game in QuakeC and then move on from there, 95% of what matters can be achieved solely through QC, and it already has roots in that. QVM is a nice overhead, but it doesnt get as deeply into the game logic as one would need, and thus you still need to get into the nitty gritty of the DLLs. Also, well, most Q3A modders are more of... modern day game modders. Or id modders. They are working on UT2k4/HL2 and to some extent, Doom3. Q3A is already old news for them.

If you want to find QuakeC coders... attract them. They are out there. I spend each day talking to about 30 of them, but an instable project is the least attractive thing to them, and this projects engine choice fluctuation isnt exactly attractive.

I grant you one on the netcode.

Elric wrote:
the whole game code to rewrite, from a base code with a nice bot AI, but no single player. That's a huge task and probably the biggest problem given the lack of coders
That is bigger than a huge task... American McGee's Alice reported that the biggest task they had to overcome in its development was not the art, gameplay, level design, sounds, scripts, writing, or voice acting - but making Quake3 accept the single player structure and focus in that environment. This is from a professional game dev studio with massive funding, from people who have worked all over the industry. I think this fact grossly outweights all of Q3As pros when it comes to development for anything with a single player initiative.

Elric wrote:
Q3 shaders may be a cool thing to have. But I'm not an artist...
They are cool, but how often would they be of use in Blood ? The only thing that is coming to mind is the electricity jiggling in the Tesla Cannon v_wep. Maybe the pulsating light in the napalm cannon. That is it.

I think switching to a less powerful engine that is more rigid in its development, without a clear supporter of to at least trouble shoot (sure LordHavoc is busy, cause he supports many projects, along with developing a new engine of his own and a whole new game, but Ive never had him not answer a single question of mine thoroughly, carefully and honestly, and have seen him take the time to help anyone when it comes to his engine). The Q3A engine is a fine engine, but it lacks some fundamental elements such as a single player structure, dynamic lightmaps, and dlights on entities in any true form.

The bots are... okay. They roam a map real nice. But a monster doesnt do that. The Q3A bots look nice at a glance and are more close to func_trains that shoot you at a closer look. If you want to see AI in Q3A, check out the Brainworks bots.

As far as bots for Transfusion goes, FrikaC, just last night, started work on a new bot called FBDP. It is DarkPlaces specific, he is coding it for Nexuiz but he assured me that if Transfusion still uses DP it will work just as well in it. It makes complete use of DPs features (of which it has many designed to make it bot friendly).

But bot AI is also unique from monster ai. This project is about making Blood, and Blood had some snappy ai for its time, but they didnt run the map picking up powerups. They lied in wait, and if they couldnt find the player... they picked angles and walked in them. If they didnt make progress in a certain direction, they would alter the angle some and try again. Which reminds me quite a bit of the built-in c function in QuakeC, walkmove and movetogoal.

I think getting real gamecode done is the most important thing, regardless of engine used. And I honestly believe, as I always have even when I first suggested qBlood, that the majority of Blood can be created and enjoyed, through QuakeC.

So in short, switching to Quake3 over DP would create a deficit in productivity now, due to the conversion and recreation process, without any clear advantage in the long run beyond that of conjecture, which rests mainly upon hope that beyond this fence lies a more rich community.
User avatar
da_weezle
Acolyte
Posts: 498
Joined: Fri Aug 27, 2004 11:36 pm
Location: Iceland

Post by da_weezle »

I second that.

Seems very harsh indeed to just switch to Q3 for crouching (and few other new features). Monster AI (as they act in Blood) is not in the Q3 source, so we'll have to fix that ourselves anyways.

I think we're pretty close to creating a proper version TF with the first episode of singleplayer done.

Sure the netcode needs some work, but with the Q3 source coming out, we might be able to speed up the netcode in DP a bit (using the Q3 source as an inspiration is what I'm saying). Another thing is that the netcode (unlike crouching) is more likely to be fixed anyways in the future, because there's also the Nexuiz project using DP.
User avatar
da_weezle
Acolyte
Posts: 498
Joined: Fri Aug 27, 2004 11:36 pm
Location: Iceland

Post by da_weezle »

Here's my updated list:

1. Create a hierarchy for project management
2. Create a clear summary of the project's status and goals
3. Create a clear summary of the current active developers and their activities
4. Create a summary of vacancies
5. Create a guide for (new) developers
6. Find out who's working on what
7. Keep The Cabal private for developers
8. Create one website containing 1, 2, 3, 4, 5 and this forum on www.transfusion-game.com
9. Stick to QuakeC and DarkPlaces (instead of Q3)
10. Use Q3 BSPs (instead of Q1 BSPs) for maps

1.
Suggestions:
Luke (Kazashi): Project Coordinator / Release Manager
Chris (Cruaich): Lead Artist
Joe (I Live...AGAIN): Lead Mapper

9.
+ all current code is in QuakeC and DF, transforming to Q3 would take a grand effort
+ most things can be done by QuakeC and we have access to the DP-source
- Q3 comes with crouching and proper net-code

10.
Seems to be necessary for the crouching issue and the mappers don't really seem to have too much problems with converting the current maps (correct?)
Last edited by da_weezle on Mon Aug 15, 2005 09:43 pm, edited 1 time in total.
Kazashi
Shadow Warrior
Posts: 412
Joined: Tue Aug 10, 2004 09:40 am
Location: Blue Mountains, Australia
Contact:

Post by Kazashi »

Ok, I think I'll jump in with some technical details concerning various formats and engines.

The crouching problem is caused by the q1bsp structure, which as scar3crow mentioned contains but a few different bounding box sizes. A new model format won't help that, what is needed is a change to the bsp itself. LordHavoc proposed using the .map files (which are used to compile the bsp) to provide better collision information so that crouching would be feasible, however we are *still* waiting on that support to be finished, and it could very well take many months, despite being a good percentage towards completion.

The other option here is to use q3bsp, which DP supports fine. IIRC there were a couple of issues, primarily concerning lightstyles, however there is a solution to that. My primary concern is whether the mappers will want to recompile the maps to q3bsp, with a good chance that they will stay in that format. Now, why do I want crouching so much? Not because I want to duck projectiles, but because Blood was designed with a lot of areas and passages where you have to crouch. It's a part of the gameplay which I do want to see fixed.

The issue of game coding is one that often comes up with much debate. We only have one QC coder at the moment. There is no guarantee that switching engines will yield better results in trawling for coders i.e. there could very well be a shortage of Q3 coders, meaning that aspect of the project wouldn't be completed in any case.

Netcode is somewhat beyond me, a lot of this due to me personally being more interested in playing Single Player than BloodBath.

Contrary to Joe's belief, I don't think we really spend much time at all on fancy things like texture recreation and realtime lighting. Texture recreation isn't necessary for the project, and is more of a side thing which people can dabble in if they feel so inclined. But it is certainly not central to Transfusion, and we're not actively seeking people to do that. Similar story with bumpmapping stuff: primarily this has been a side venture (mostly undertaken by myself) simply to see how it would turn out. Bumpmapping on models is a different story as it can increase the amount of detail present. Rtlights I believe is different; while not absolutely necessary, in a way it helps clean up one side of the lighting system. Most people won't be playing in rtlights mode, but if they do then without rtlights they're likely to run into a blinding and unplayable mess. Yes, it's not critical to the game, but it's not a trivial throwaway. There is a second reason why it's not a trivial throwaway - they can be used in normal lighting mode as well. That issue with q3bsp and lightstyles? It can be solved with an rtlights file being used in normal lighting mode.

Shaders? Not really needed. There are already options in place for animating things like the tesla cannon and napalm launcher skins, they just need some code.

Model formats.... I sorely wish that DP had a decent skeletal format which I could use, it would be much better for the models as opposed to vertex formats like MD3 (primarily for character animation). But this is one of those things which could take a long time as the engine coder doesn't consider it a priority.

Ok, that's my speel for now. If I am going to be project co-ordinator I'll be needing to speak with all the team members, which I'll get to hopefully soon.
User avatar
Fiend:Kamikaze
A LUMBERJACK!
Posts: 114
Joined: Thu Aug 12, 2004 03:29 pm
Location: United States

Post by Fiend:Kamikaze »

Oh for f***'s sakes, the suspense is killing me. It's like watching a romantic comedy.

"Will they?"
"Won't they?"

I'm fairly sure that by the time this is finished, if it is, that I won't be able to play it because of my either hectic schedule or arthritis. I've pretty much come to terms with the fact that while there is information being released about the project, it's about as dead as DNF. (meaning not officially, but close enough)

Call the bloody coroner.

The main thing is this:

Image
Post Reply