Doom3world

What exactly is the issue with the shaders? Wanting GLSL over ARB asm or not wanting to use id's shaders? Because a shader is all math and math, I far as I know, can't be copyrighted so you're probably be safe in using them.

Wanting a confirmed GPL system. Whether it is new ARB shaders, or GLSL. Using id's is a grey area, ie no one from a knowledgable legal standing has said yes you can use them and no one likes being in legal grey areas.

If it was OK, then I'd say TTimo would not have initially opened the ticket over at github(and then closed it when there was lots of initial talk of GLSL although none of it was ever completed). Current, existing ARB shaders from your projects and others, unfortunately, have many compatibility issues with many graphics cards. Particularly of the ATI variety.

so at the moment there is no real, stable, compatible solution to a GPL engine based game, regardless if it's ARB or GLSL. Either would work, but none exist.

Don't mean to sidetrack people, but a game can be done in 6 month. If things are moving poorly with iodoom3 and this project, wouldn't it be easier to add what's missing to Darkplaces engine and have whole a lot of devs happy?

Not a cult, but simply a powerful lightweight free engine, using which indies can make games that sell. It does better job than Doom 3 with rt lights and has shadowmapping, it has global illumination (realtime and through lightmaps), etc. etc.

The reason I popped the question about viability of this project is that I can see indies making good games using stock Doom 3 with minor tweaks, which could be implemented already. Instead, people will be waiting for years for this project or iodoom3 (or add whatever they need if they have skilled coders on the team).

So why not to make milestones and release basic usable Doom 3, then just update it with new features? That's how UDK does it. Why start with complex features and have people waiting on usable builds?

The Dark Mod's "post process shader" has similar driver problems to those seen with Sikkmod or other Doom 3 mods but those have been overcome in the upcoming release simply by renaming the executable to "TheDarkMod.exe" so that ATI and Nvidia's drivers don't try to "optimize" the ARB code. That said, it would be great if the drivers could optimize the code for their respective hardware vendor and GLSL would make that happen by default in most cases.

I think the main drive for GLSL is so that newer techniques and branching can be added at a later time by amateur programmers on an experimental basis in a similar fashion to all the tinkering done with the ARB shaders for Doom 3. Without the complete set, someone tinkering with a small change can't "release" their work and expect people to try it out.

@Sikkpin: Incidentally, I would love to see what you could do with "just" ARB shaders and access to the renderer (remove the depth buffer hacks). I'm sure it would be magnificent in spite of the GL 2.0 limitations. Plus it would be a cool exercise to see what you could achieve visually with SM 2.0 hardware... (I'm sure that's a broken record request by now though. )

Given the alternatives, there's no reason to use idTech4 until it gets better tools and a modern renderer. Which is what's going on here. Given the current rate of development and stated goals, I don't see this taking years, especially if people start contributing in a meaningful way.

this is a thread about advancing the doom 3 engine i really don't even see what the relevance of darkplaces is to this discussion or why it keeps being brought up here

http://www.phoronix.com/scan.php?page=n ... px=MTA5MTI

As much as I dislike Phoronix it appears someone is working on the GLSL backend.

There have been several backends in progress or in testing. That is where everyone stops. No one ever creates the shaders.

Anyone with a fraction of experience in designing a game from scratch will take a graphically mediocre engine with streamlined tools and good documentation over a bleeding edge one thats a clusterfuck of jumbled together parts. Thats why UE3 has seven billion games and cryengine has 3.

I 100% encourage people to push the engine hard in different directions but for the rest of us until these mature there needs to be a base that "just works" and a clean editor and good documentation. I think this is what iodoom3 should focus on, right now the default editor is a steaming pile of shit that no sane artist would touch with a 10 foot pole. Also better model import standards, and blender interoperability, to ensure an open from start to finish content flow.

There really is no good way of exporting from blender as of now.

Edit: while we're on the topic, darkplaces is a great achievement but it remains a bit too messy and untested to build a serious game off. I've tried messing around, and concluded that doom 3 is easier to work with because it has way more documentation and seems to be more holistically designed and robust. Darkplaces will get there, but right now its an oddball frankenstein of code from 1992 and modern shaders and quake 3 map formats and rt lights that are a bit of a joke.

While I'm unaware of the current state of the D3-IdTech4 toolset. If it's anywhere near as bad as our Source(CSS,HL2,L4D,etc) tools, I can understand.

I completely agree with you on everything Whitewolf; a nice, fast, simple engine, that just works. Would be perfect. After that, add another branch that you can add experimental features to. Which you can merge with the primary branch once the feature has been deemed mostly bug free and useful.

Maybe you'll think me a bit pedantic, but a lot of posts in this thread belong as separate topics inside the Doom 3 Source Code forum. It's beginning to bug me that this thread is soooo long, and is wandering far off topic from "Doom 3 modifications".

Rather than derailing jmarshall's thread about his work (which again probably doesn't belong here since it isn't a D3 mod) talking about iod3, GLSL backends and DarkPlaces, perhaps you'd like to start a separate thread over there talking about where you want the engine to go, or what changes you propose to make.

It'd be good if that forum saw some activity, because those who don't read this whole board but are only interested in D3 engine coding are seeing this place as pretty dead (personally, until I found this thread I typically stick only to the "coding" forums)

Could staff split the topic?

A lot of it isn't off topic though. The heading of this thread is renderer/post process/tools... All the talk on ARB/GLSL falls directly in that scope and if Jmarshal is one of the few active people coding on the GPL engine then I find it highly important to bring this issue directly to that person. This specific topic was discussed in the GPL specific forum and everyone pointed fingers to everyone else.

Always use the 'View Active Topics' link..

search.php?search_id=active_topics

i strongly agree

personally i am tired of hearing about darkplaces in here at all

darkplaces is an entirely different engine that hasn't got a damned thing to do with anything that is happening in this thread other then "geewiz i wish someone would work on darkplaces and add all these slick features there instead" which to me is a pointless and borderline rude request

i honestly can think of a single legitimate reason that anyone should be even bringing the word darkplaces up in this thread

the only thing it is accomplishing is to further abstract the thread and make it harder to find any useful information about the topic that this thread is here for which is jmarshals modified doom 3 engine

at least iodoom stuff is somewhat more excusable as it is a parallel branch of the same engine and cross development could possibly prove beneficial (note the word possibly and maybe not probably) darkplaces is an entirely different codebase with its own forum for its own users to plce their own feature request

likewise this isnt the place to compare the default doom 3 toolset to x other system (udk etc) that is also entirely irrelevant as this branch of the engine is working on its own toolset ("doom 3 has really annoying in editor viewport controls i hope this will do a better job" is a good example of relevant on topic subject matter "doom 3's editor is shit and so is valves hammer udk is da bomb" is NOT relevant subject matter)

and yeah its going to take a long time before this engine is ready for prime time

so if you need to have an engine that is ready to go today then please by all means go ahead and choose any one of the many out there

but again there is no reason to come here and tell everyone to give up interest in this engine because it isnt ready today

Darkradiant isn't really that different from the UDK editor in core functionality once you get your models into it. Granted there's no material editor, terrain editor, or any of those extra tools, but most of that stuff is better accomplished with 3rd party stuff. What I really want in the tools is a simple 1 button import for models which generates the the appropriate mtr files (which I think Jmarshall is accomplishing with his FBX import). The current DeadlyLaw editor is already a big improvement over the stock D3 editor already so I think it's getting there.

But anyway my point was to make a big +1 to this post on the Blender integration. I have licenses for Mudbox and 3ds Max (as a "student" licensee) but that's $3700 worth of stuff 2 years from now. Zbrush is $700. Blender does most of the stuff that all three of these programs can do in a game dev context for exactly $0. Yeah, the interface is weird if you're coming from Max and it's got its own workflow but the difference isn't worth $3700 for a hobbyist. I think Blender integration is the one of the paths of least resistance for the uptake of a modified D3 engine. I'm not sure supporting a Collada/Blender importer alongside the FBX importer would be terribly difficult, and the inclusion of a more open format (and a plain text format) could bring more people in. D3 derivatives are (potentially) competitive because of open access.

why keep bashing so much doom3 engine as it is? geez i've been having fun with it since the alpha (if you think doom3 engine sucks you had to see the alpha engine! )
yes jmarshall23 is adding all those features to the engine, but this doesn't invalidate doom3 engine as it is at all! there is no need to constantly bash doom3 engine as it is!

who is bashing the doom engine?? I never bashed the doom engine

my point was simply that if you have something to say about the (vanilla) doom engine or udk or darkplaces or any engine that IS NOT this engine that is being discussed in the thread then there is no reason to bring it up at all

wether you are bashing or defending your engine of choice there is a place for that and it isnt here

my only problem here is that this is a 20 page thread which would seem to indicate that the project is getting a heavy following but then you come in and have a look and what you actually find is that every 6 pages someone mentions darkplaces and then theres a 4 page pissing match to follow and it isnt relevant to anything

to be fair i am just as guilty as anybody of engaging earlier in the thread but even i cant take anymore sidetracking

i have really high hopes for this project and it seems like the more we bicker back and forth about irrelevant crap the less actual progress gets posted (from my observation only)