Untethered Dark Mod
Posted 10 December 2011 - 02:18 PM
Posted 10 December 2011 - 02:46 PM
Edited by Shadowhide, 10 December 2011 - 02:47 PM.
Posted 10 December 2011 - 02:49 PM
Depending on if we can get community involvement - I could make a nice little upload site to manage this in a 'friendly' manner, but at the moment it's all a bit of a shot in the dark as where to spend time most efficiently.
Posted 10 December 2011 - 02:58 PM
Posted 10 December 2011 - 03:12 PM
It's not just the assets in the main mod itself, we would also have to track down all the assets used in the various maps released thus far (or just let them be broken).
Posted 10 December 2011 - 03:38 PM
Yes if they were going through one marathon anyway, they'd want to add these to the list too.
The alternative would be to have to make a fork release to keep the old FMs on the main trunk, which would be a bad idea.
Posted 10 December 2011 - 04:10 PM
TDM The Best MOD. Period.
Please visit TDM's Moddb site and help promote the mod:
(Yeah, shameless promotion... but traffic is traffic folks...)
Posted 10 December 2011 - 05:18 PM
Right. This duty is suitable for mappers who are familiar with assets I believe.
Posted 10 December 2011 - 05:29 PM
In any case, understanding what exactly is needed is the first step. At the moment the material side is actually quite small, while I have yet to do a quick check, it seems to be under 300 files - of which I think the majority are particles/decals/lights(which also has many duplicates) and a few character textures.
I'm going to try cover the generation of lists for particles and sounds tonight.
Lastly, there's no need for this to be a single step process, if we can get nice lists - just tick things off as we go. That's my plan anyway.
Posted 10 December 2011 - 05:29 PM
If I take a D3 head model and move a few verts around, can we release that?
If I make a new model from scratch that duplicates the position of the verts in the original D3 model, can we release that?
If I take pieces from D3 models (a torso here, an arm there) and reassemble them in a different way, can we release that?
Can we release a normalmap that has been cut up and reassembled?
Can we release a texture that has simply been mirrored or rotated?
If I take the zombie md5mesh file and export it as a .lwo, can we release that?
I genuinely have no idea, but it has a distinct bearing on how many assets we'd have to replace and how difficult it would be to do it.
Posted 10 December 2011 - 05:49 PM
I don't think even the lawyers have figured that out. What you're observing is that on some fundamental level the concept of "illegal data" just doesn't make any sense.
I think most of your examples would qualify as derivative works, but at some point a small-enough item ceases to be protectable (what if you copy a single pixel from a single frame of a Hollywood movie?). The line is exceedingly blurry.
Posted 10 December 2011 - 06:24 PM
If I take (...)
Posted 10 December 2011 - 07:26 PM
What I do with working from samples is just try to use certain elements from them, to take them and make them unique but similar. For example a lot of textures that I make, I work from photos, however the photo source itself is far removed from the end result - usually the diffuse has actually been built from the height map which I build by hand. If I'm just replacing a normalmap, I convert it to a height map, do manual edits to remove any problems (normal->height is messy), add in some flavour and a little bit of overlay samples or whatever. Then throw that back and generate the normalmap, testing it against the old one. Once it's nice I'll take the height map and build the diffuse from it - so that it matches closely. Sometimes this uses a blend of the colour from the old diffuse, but even then I try to water it down and abstract it as much as possible from the final product. It's a fairly relaxing process all in all, and it does have it's own 'this is mine' reward.
With models, would it not be possible to take the existing model, use some tessellation/subdivision (and perhaps bring in details from the normalmap - displacement from a heightmap), and from there work to cut down to simplify the mesh which would in the end be quite different from the initial model - perhaps do the texture at the same time? (I'm pretty crap at skinning non-statics ). I think that would give a credible quality improvement too
Posted 10 December 2011 - 08:52 PM
Yes, we're using at least five D3 heads.
How would anyone tell the difference between the original and a recreation that used the exact same coordinates?
Not sure what you mean by 'not much point'. This has already been done for the beggar character and the noblewoman, for starters.
I wouldn't do that also for other reason: Doom's material is dated. Models are low poly and textures are low res.
You're talking as if this hasn't already been done. Broken record alert, but we already use tons of D3 material. To go standalone, it must be replaced. Knowing to what degree it must be replaced is important information.
For example, the wearbeast uses a head-mesh created by Oddity on top of a body mesh that is a modified imp with a new material. Is it modified enough that it could be released? Or would the fact that it is based on an imp mesh be a concern? How do we know? (In this case it's a moot point, since it uses the imp skeleton and all the animations which would also have to be replaced, but the question is still valid for other cases)
Posted 10 December 2011 - 10:53 PM
From what I've gleamed from google, most of the cases in which similar topics arise come from trademark infringement, where a character likeliness is retained or from cases where people have used unlicensed things in commercial or commercially backed productions. In any case, the content has been part of TDM for a long enough period in which no one has complained.
You'll never find someone that will give you a real stance on this stuff. So long as you're comfy with it — which I damn well would be — then why not; after all, I'm fairly sure TDM has helped sell a fair few copies of d3
Posted 11 December 2011 - 12:19 AM
OTOH if you make a new head and put it side-by-side with the old head and it's visibly trying to look like a different head, then that's probably outside the standard. Speaking realistically, I think it's as much about the impression it gives off about its intention (is it trying to copy & look just the same or trying to look different while performing the same function, which of course you can't copyright) as the actual similarity in some "objective" sense.
Edit: In short, just make it look like it's trying to be a different head (body, post, whatever), like a good faith effort that shows. What you don't want to do is make it look like, when you put the two side-by-side, that you're trying to blatantly copy the look & feel (and that's exactly your intention) and you're changing just a few negligible details in a cynical attempt to "effectively copy without literally copying". It's about the attitude that comes across looking at them, are you in good faith or bad faith. That's how the standard works.
Posted 11 December 2011 - 04:25 AM
Posted 11 December 2011 - 05:19 AM
I take it this applies to re-creating the meshes visually by inspection, rather than creating a new mesh by editing the original D3 one? I assume in the latter case it would be considered a derivative work because of the copyright on the digital data, irrespective of the actual visual appearance.
Posted 11 December 2011 - 08:11 AM
That's exactly what we plan to do, isn't it? We can't replace the D3 assets with things that don't look the same...if we remake the D3 steam particle, it has to be the same size and shape as the original or it's going to look misplaced in all the maps that already use it. If we make a new zombie, it has to have the skeletal joints in exactly the same place, or maps that have things def_attached to zombies (like Glenham Tower) will be broken.
Furthermore, what do you put "side by side" for comparison? Just the final head, as it appears in game? Or every part separately? Including the mesh, the diffusemap, the normalmap, the specular map, and the material file?
How do people imagine this process working? It sounds like a fairly subjective criteria. Do we establish a committee to decide if a replacement has the "same look and feel" as the original for every asset that is replaced?
Well, it remains to be seen if we are or we aren't. But if you're suggesting we replace the (professionally done) assets with even better assets, that reduces even further the number of people who can actually help.
Posted 11 December 2011 - 11:14 AM
It has to fill the same function, but that's not itself copyrightable. That's why I kept talking about intention playing a practical role, and being in good faith. This situation is interesting since it's not the usual bad faith case of wanting to copy, but there's a good faith reason. You actually intend and want the things to look different, and your basic question is the good faith one of how different do we need to be while still filling the function (instead of how close can we get, which is the bad faith angle). We love id for making the engine and releasing the sourcecode, and we want to respect their property. If someone raises any question, you want to be able to say we worked to respect the copyright and make these assets visually different, and that has always been our intention, and it's documented, e.g., in this thread, in every thread the topic comes up.
Practically speaking, you raise great questions, and some of the tougher cases. We should think about some kind of standard & process (before "publishing", we can still collect stuff in the meantime). That's a good idea. I don't want to jump the gun on anything, so we can sit on it and think for a bit.
Offhand, I don't think it's a matter of what some committee says, since nobody else has to care what they think. Ultimately it's what id people & an arbitrator or judge (not that it would ever go to arbitration or court, but their hypothetical perspective informs the standard) would say looking at the two, which amounts to a detached perspective of a reasonable person on the street. When you look at the two, is your gut reaction "Come on, it's the same thing." or "These are different; just look at them."? Also if the "same thing" is some public domain source, like photosource textures, then that's on safer ground.
What we might do is line up some case studies to see how this practically works. Make and collect some in-house replacements, then line them up with the originals and let's have a discussion about it when we have a better idea of everything involved and how it's turning out (e.g., how the replacements work in FMs, and how much latitude we have to make stuff different, etc.). That's what we can do for now, that and just read up on case studies; what has id done before; how does the standard work for games... (Also if there's a movement like over at d3w to make their own replacements, we can watch how things fare for them.)
Posted 11 December 2011 - 01:55 PM
Are you sure about this? That would mean I could take normal-maps straight from D3 and use them un-altered, as long as the diffuse-map and/or mesh is different enough? How can that be, when I'm still distributing an original D3 file? Could I take a whole skeleton and series of animations from Prey and distribute them as long as I rig a unique mesh to them? (I'd love to do that, as they've got an awesome hawk rig) That doesn't sound right.
Posted 11 December 2011 - 02:13 PM
No, copyright does not work like that. While Doom3 is a "work" in its entirety, one could also argue that single files are a "work" (and Doom3 is just a collection of smaller "works"). You can't take a file unmodified and re-distribute it like that, unless there is a licence that covers that.
There are some border-line cases like a 1x1 pixel texture - you cannot copyright that like an author can't copyright the individual letters - be he can (and does) owncopyright on a a whole paragraph or even the entire book. So you can't re-release a paragraph on its own (unless you use the american "fair use" joker card), but you can take the letters and write your own story.
So for us:
For sound assets, rerecording them with your own equipment is the way to go (regardless on wether the end result sounds close to the original). Likewise for any textures (luckily we don't use much and the ones we use can be painted easily in Gimp anew) - paint them from scratch or from photos we own.
As for the meshes, rebuild them from scratch - if you are lucky, they look different *AND* better - if they end up looking alike, well, though look.
The gray area are shaders and scripts and entity defs so on.
Luckily, after greebo rewrote the entire AI in the SDK, I exterminated all AI scripts and I think almost everything else except a few definitions already. And these definitions are likely not copyrightable (just like header files) as they are too simply and too technical (you can't define them in a different way without them breaking). There are already court rulings that establish as much so renaming all D3 event functions to different names is just silly and unnec.
That leaves us with entity defs (we removed almost all of them anyway, and the ones left - see above for definitions), and shaders.
Maybe I forgot something, I don't know off-hand. I think the easiest way to see whats missing is to build a special "darkmod.exe", and install it along with the entire mod in C:\Darkmod and try to run it without D3 even installed. Then we will see what files it thinks it needs.
"Remember: If the game lets you do it, it's not cheating." -- Xarax
Posted 11 December 2011 - 03:07 PM
Posted 11 December 2011 - 09:07 PM
Posted 11 December 2011 - 10:26 PM
(2) They won't sue us because nothing in TDM copies LGS's old property. You can't copyright gameplay mechanics, and none of the assets or AI or setting comes from Thief. (And the standalone part is irrelevant in this situation anyway.)
(3) Besides, they already know about Dark Mod and at least the former community director let us know they're also fans of what we're doing. They've had years now to say something to us and whatever little indications they've made have been positive.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users