Et: Xreal

[BackSnip3]

Okay thanks again, it was that, using your lighting q3map commands gave me the same result
I’ll be patient, I just really hope it will be finished and super stable one day with super graphics effects and my models inside
There’s still a bug that I already told here but probably you didn’t saw it as this topic has so much pages, it’s a bug with the playermodels(probably the gpu acceleration), it stretches models where a vertice’s weight is shared by multiple bones(like elbow,shoulder,foot)

So you think it’s not that big, but when I use a custom model that looks like that in ET:

… it looks like that in ETxreal:

[Tr3B]

It is an arbitrary limitation within the vertex shaders. It wouldn’t be fast to have more than 4 weights per vertex. You have 2 options: fix your models or use r_vboVertexSkinning 0. ET’s skeletal animation implementation is a huge mess and it’s also slow.
I profiled the engine and R_CalcBones() is the 9th worst function in the entire code. ET recalculates all bone positions each time the client game wants to know where a certain bone position is.

Do you know good tools that work with .mdm and .mdx? I’m already thinking about writing a converter to the Doom 3 or Unreal formats or even COLLADA so it would be easier to import those models into Blender or another modelling app for further improvements.

[BackSnip3]

Well I do the exporting with Lightray3D, a non-free ****ty program, but it imports and exports .mdm/.mdx. Lightray website

Those import/scripts are .dll files so I don’t know if you can see how he reads the ET files, I’m no coder at all I’m maybe saying stupidities

I already tried exporting to md5mesh/md5anim, it works well (I tried creating new anims in blender, re-export to .md5, import into lightray and export into .mdx…), though the main problem is that Blender importer for .md5anim is taking ages to import 5000 frames !

Also, .md5anims are constructed to be one file per animation, so they just have to be called by a script, while ET’s .mdx has all anims in one file(body.mdx) and uses .animc files.

I think you would also have to recode this boring animation file system to doom 3 style .def files to make it easier to implement .md5 into first person view and characters.

I’d be glad if .mdm/.mdx **** would be removed and replaced by something working only with .md5mesh stuff, I tried using .mdm/.mdx for first person view animations but it crashes the game, it’s only made for biped skeleton!

I can export the stock ET models and anims and share them into .md5mesh/.md5anims if you want…

[RazorKiller]

Dushan what is that “Newton physic” inside your OpenWolf project? Can you give us some preview and files so we can test that?

And can you please concentrating yourself on fixing bugs first, not to compete with others? then you can add and can be creative how much you want and add features what you want.

[Dragnadh]

Is this still going strong or did it fade/die again?

[KeMoN]

relax man
You can’t expect them to work every day on that project. That’s insane. I haven’t worked on Pointe du Hoc project for the entire last week until today. Sometimes you just aren’t in the mood to work on it.
If we put pressure on them they will soon loose all their mood so let them set up their own schedule. There is really no lack of time for that

[Shownie]

[QUOTE=KeMoN;314246]relax man
You can’t expect them to work every day on that project. That’s insane. I haven’t worked on Pointe du Hoc project for the entire last week until today. Sometimes you just aren’t in the mood to work on it.
If we put pressure on them they will soon loose all their mood so let them set up their own schedule. There is really no lack of time for that :)[/QUOTE]

Agreed! “Quality demands time!”

[Dushan]

Newton??? Newton its opensource physics engine, I added that inside OpenWolf several days ago… Server stuff is OK, game specific physics is missing only from project … Currently am thinking on ditching XreaL engine from project and start using and updating old vanilla renderer engine.

[IndyJones]

what do u mean by ditching?

[Dushan]

abandon XreaL engine

[IndyJones]

that’s what i was afraid of…
so we won’t see et coming out on new engine with a lot of new capabilities?

[Dushan]

don’t have to be … in vanilla ET renderer engine before I had celshading. I didnt say that we cannot update vanilla engine, we can still update it with features from XreaL engine (rotoscope, md5 models, jpg/png images, vbo/glsl … stuff like that) and add some other features what XreaL dont have … like raytracing …
I have think about this for some time, personally am not happy with XreaL performances with ATI/AMD GPU cards. I dont know is that because of AMD OpenGL driver’s or any other reason … Nobody can say that XreaL isnt good and that dont show advantages over any IDTECH3 engine, its huge improvement.
But it would me take too much time to learn XreaL rendering engine. In my XreaL build’s there was/is extremely big problem on Radeon 4850 and Omnibots ingame … Impossible for playing with 6+ bots ingame

[daz2007]

[QUOTE=Dushan;316035]don’t have to be … in vanilla ET renderer engine before I had celshading. I didnt say that we cannot update vanilla engine, we can still update it with features from XreaL engine (rotoscope, md5 models, jpg/png images, vbo/glsl … stuff like that) and add some other features what XreaL dont have … like raytracing …
I have think about this for some time, personally am not happy with XreaL performances with ATI/AMD GPU cards. I dont know is that because of AMD OpenGL driver’s or any other reason … Nobody can say that XreaL isnt good and that dont show advantages over any IDTECH3 engine, its huge improvement.
But it would me take too much time to learn XreaL rendering engine. In my XreaL build’s there was/is extremely big problem on Radeon 4850 and Omnibots ingame … Impossible for playing with 6+ bots ingame[/QUOTE]

The Vanilla ET Renderer Engine is very Bottlenecked as well, I get more FPS is Half Life 2 then in ET, Specially on Radar Map and with Air Strikes, Its Needs Speeding up.

Xreal has alot of Potential, would be a shame to give that up.

[n00btime]

[QUOTE=Dushan;316035]don’t have to be … in vanilla ET renderer engine before I had celshading. I didnt say that we cannot update vanilla engine, we can still update it with features from XreaL engine (rotoscope, md5 models, jpg/png images, vbo/glsl … stuff like that) and add some other features what XreaL dont have … like raytracing …
I have think about this for some time, personally am not happy with XreaL performances with ATI/AMD GPU cards. I dont know is that because of AMD OpenGL driver’s or any other reason … Nobody can say that XreaL isnt good and that dont show advantages over any IDTECH3 engine, its huge improvement.
But it would me take too much time to learn XreaL rendering engine. In my XreaL build’s there was/is extremely big problem on Radeon 4850 and Omnibots ingame … Impossible for playing with 6+ bots ingame[/QUOTE]

I’m pretty sure many graphics people would agree that if you’re having a performance issue with raster graphics, ray-tracing will not help at all. Plus it also means you’d have to change the way the geometry was stored, so you’d probably get rid of the BSP format as well and break compatibility with all ET/Q3 maps which can’t be recompiled. As someone who’s worked with a lot of large software projects, I think it would be very unwise to throw away all of Tr3b’s work, especially since you just said you’re going to reintegrate/reimplement a lot of it.

The code written for the XreaL renderer is card-agnostic, so any performance issues are likely because of drivers (as you said). They have a bad reputation as it is. You could, in theory, make separate execution paths for each card, but it would quickly turn into a horrible mess.

I would also profile the engine before you assign blame to any certain component. Bots should have no hooks into the renderer at all, so if you have trouble with adding too many bots (which should be completely server-side), it probably means that the bots are chewing up your CPU. Since the engine is single threaded, whatever timeslice was spent in the renderer previously is now smaller, so fewer FPS.

I’m not trying to sound rude or condescending. Your project has stirred a lot of interest and it would be a shame to see it go down in the flames of feature creep.

[IndyJones]

there aren’t (m)any good games coming soon, so sad…
i thought at least et will keep providing some decent fun, just like in the old times…

i assume i am not the only one who just don’t like brink or qw (am i allowed to say that loud?)

[aaa4]

yeah, you pretty much stole et:xreal, placed it in et, are now having the attention from the community, revert the engine, make it another thing, selling it as the old, stolen one, and established your old plan of finally setting up something successful. it is not hard to read that from your lips actually. all your mod projects from the past concerning et have been one big failure as far as i see, so now stealing et:xreal to gain attention for your project is the way to go ?, poor form…

also your project is pretty bad from a conceptual point of view. you pretty much merged everything you see. it is like one big playground for you, when i see your changelog, there is actually 90% useless stuff for et in it. you also do not really listen to the community, telling you exactly what you should be told. you are “overfeaturing” your project, noone needs most of the stuff you implement.

also you do not really care for those coming after you. for example the more important thing would be contributing to a general code base which focuses on patching/improving the engine. help establishing such a code base first, and without your overloaded features pls! merging thing after thing chaotically, not caring at all for those after you is not the way to go. only caring for YOURSELF and YOUR project. believe me i have watched you now pretty long, and you actually care only for yourself. contribute to a bigger bug fix project first and help establishing this in a well thought out conceptual basis, and that is surely not the same concept as you have now. do things in a conceptually convenient way pls. thank you !!

[IndyJones]

this might be well-known aaa4 trolling again
however, i do agree at some point. many of us - et players - would like to contribute to the new engine project, but most of us are pretty useless in this case - i’m not a coder, but you dushan are.

this is not a rat race, it’s not about who’s going to release et first and get all the bravo’s…
why not joining diffrent project then? after contributing to one, u could always add your features later…

[Shownie]

I must say, all your flame posts towards dushan will not likley make him change his mind…

Honestyl this is big lack of respect from the “community”. Instead of flaming, why not try start a well discussion… He never said he is going to abandom this project, just move to the vanilla engine.

If I were dushan and heard all these flame posts, I would abanom you all for beeing jeerks… sad, but true.

[IndyJones]

really? would u quit doing something just because an one or two said something negative about the project?
that’s not smart either… there are always black sheeps in the community…

[Shownie]

Something negative? Man they are like releasing hell on this dude, I see no fair point…

Well maybe I wouldn’t abandom just like that, but I would for sure not have the motivation to keep doing something for a community when treated like this o___O