Is Anyone Planning To Port Blendder To The Pandora

[Monk]

Sorry if this has been mentioned before, but why do you bother with textures at all?

Repeat after me, again, "these are not textures for use". Repeat after me "these are only templates". If you don't know what a template is, or why one might use a template, just feel free to ask and I'll try to go through it, whether I've explained it in this forum or not.

But try and wrap your head around the idea that, as I've repeatedly said, those bitmaps are not for use as textures in any game, but are only to illustrate the mapping. Capiche? You keep criticising me for doing something I haven't done and wouldn't dream of doing, and which I keep saying I haven't done and wouldn't do, and it's starting to make me question your level of attention, at least.

Using wire frame rendering and per-vertex colors is both many magnitudes faster and produces much cleaner and fully anti-aliased lines (because you must admit that the texturing above - even if it is as good as it can get using pixmap textures - looks kinda weird because of the texture bilinear aliasing)

Repeat after me - those are not how we're going to be displaying models in-game, those are not textures to render by. I don't have to admit that they are as good as you can get using pixmaps because they are, deliberately, almost the very worst you can get. Please - get this. They are not textures. They are merely simple templates. Please understand this.

I would ask you, please, to design some models then instead of bitching that you can do so much better so much easier in a text editor, if I weren't convinced from your comments above that you'd completely missed the point of the way these models were UV-Mapped.

I'm not fond of explaining the same thing repeatedly, but even less fond of doing so to the same person. I hope people listen and understand. So get this, please, this time - the templates applied to show that each facet is individually texture mapped so that we COULD, if we so choose, create nice photo-quality textures to apply to the models is not a photo-realistic texture, it is not meant for final rendering under any circumstances whatsoever. It is merely a template to show that the models are mapped as opposed to simple, unmapped, less capable models such as one might expect a less technically minded 3D newbie to knock out without realising the potential benefits of mapping. The templates' SOLE purpose was to show that the models are UV Mapped. They do that, but no more. That is all they were intended to do.

I feel like "Holly" the computer explaining to Lister that everyone else on-board is dead, in TV series "Red Dwarf":

Lister: Where is everyone Hol?
Holly: They're dead Dave.
Lister: Who is?
Holly: Everybody Dave.
Lister: What Captain Holister?
Holly: Everybody's dead Dave.
Lister: What Todd Hunter?
Holly: Everybody's dead Dave.
Lister: What Selby?
Holly: They're all dead, everybody's dead Dave.
Lister: Peterson isn't, is he?
Holly: Everybody is dead Dave.
Lister: Not Chen?
Holly: Gorden Bennet, yes Chen, everybody, everybody's dead Dave.
Lister: Rimmer?
Holly: He's dead Dave, everybody's dead, everybody is dead Dave.
Lister: Wait, are you trying to tell me everybody's dead?
Holly: *slightly under breath* Shouldn't have let him out in the first place!

They're templates dflemstr.
All of them dflemstr.
They're all templates dflemstr.
They are all templates dflemstr.
They're all templates, they are all templates dflemstr.
Gordon Bennet, yes even the "Spawn Fort", all of them, they are all templates dflemstr.
They aren't textures dflemstr. They are templates dflemstr. They are all templates dflemstr.

I DO so hope this question won't come up again, and that you won't suddenly revert to thinking of those simple bitmaps TEMPLATES as textures. Again. And that the textures are below par and should be redone. They are below par in much the same way that an unmodified bicyle is below par for flying across the Atlantic. Certainly a bicyle is not fit for flying across the Atlantic, but if you're even considering it in that context then I feel you do not sufficiently understand bicyles. Or, possibly, oceans either, and certainly not cyclists.

Also, I as a technically-minded guy would prefer creating these meshes using a text editor. Much easier to get the trigonometry right then (without the right tools it's hard to eyeball stuff like that), and if you use integer coordinates everywhere, you can make mesh files smaller and use extremely simple transformation matrices that might improve performance, even.

Except that a less technically-minded guy probably wouldn't UV map them, not having a clue why UV mapping is useful. Apparently.

Red Dwarf quotes are great. For this instance I like:

Cat: Why don't we just break out the lasers?
Kryten: An excellent plan, sir, with only two minor drawbacks. One, we don't have a power source for the lasers; and two, we don't have any lasers.

Or to put it another way:

dflemstr: Why don't you just admit that the textures are rubbish?
Monk: An excellent plan, sir, with only two minor drawbacks. One, we haven't used textures; and two, we don't have any textures to use, having not painted any.

Do not get that there are just more models to be made than simple, geometric ones?

Apparently not

It feels sometimes like light-bulbs are dying across the globe when certain topics come up, like Alderaan being destroyed by the first Death Star

 

[JayFoxRox]

Monk: Isn't it enough to type in these manually in your text editor?

No. Heathen.

[Gambas following, code tags fucked up, sorry, thats also the reason for the dirty /* hack */ which didn't do anything good neither]

Variables:

PUBLIC geometryList AS Integer
PUBLIC geometryTexture AS Integer
PUBLIC geometryBuffer AS Image

Texture generation and states:

geometryBuffer = NEW Image(32, 32, TRUE)
geometryBuffer.Fill(&HFFFFFFFF&)
FOR i = 0 TO geometryBuffer.height - 1
geometryBuffer[0, i] = &H00003FFF& '/*Y axis (Left)*/
geometryBuffer[i, 0] = geometryBuffer[0, i] '/*X axis (Top)*/
geometryBuffer[geometryBuffer.width - 1, i] = geometryBuffer[0, i] '/*(Y axis (Right))*/
geometryBuffer[i, geometryBuffer.height - 1] = geometryBuffer[0, i] '/*(X axis (Bottom))*/
NEXT
geometryTexture = gl.GenTextures(1)[0]
gl.BindTexture(gl.GL_TEXTURE_2D, geometryTexture)
gl.TexParameteri(gl.GL_TEXTURE_2D, gl.GL_TEXTURE_WRAP_S, gl.GL_CLAMP)
gl.TexParameteri(gl.GL_TEXTURE_2D, gl.GL_TEXTURE_WRAP_T, gl.GL_CLAMP)
gl.TexParameteri(gl.GL_TEXTURE_2D, gl.GL_TEXTURE_MAG_FILTER, gl.GL_NEAREST) '/*LINEAR) Untested*/
gl.TexParameteri(gl.GL_TEXTURE_2D, gl.GL_TEXTURE_MIN_FILTER, gl.GL_NEAREST) '/*LINEAR) Untested*/
gl.TexImage2D(geometryBuffer)

Rendering:

gl.Enable(gl.GL_TEXTURE_2D)
gl.CallList(mapList)
gl.PushMatrix()
gl.Disable(gl.GL_LIGHTING)
gl.Enable(gl.GL_ALPHA_TEST)
gl.AlphaFunc(gl.GL_GREATER, 0.5)
gl.Translatef(128, 128, 2) '/*For testing, center of my debugging map, put whatever you like here*/
'/*Ground plane*/
gl.Begin(gl.GL_TRIANGLE_STRIP)
gl.TexCoord2f(0, 0)
gl.Vertex3f(-1, -1, -1)
gl.TexCoord2f(1, 0)
gl.Vertex3f(+1, -1, -1)
gl.TexCoord2f(1, 1)
gl.Vertex3f(-1, +1, -1)
gl.TexCoord2f(0, 1)
gl.Vertex3f(+1, +1, -1)
gl.End()
FOR i = 0 TO 3
gl.Rotatef(90, 0, 0, 1)
gl.Begin(gl.GL_QUADS)
'/*Front (Y+)*/
gl.TexCoord2f(1, 0)
gl.Vertex3f(0, +1, -1)
gl.TexCoord2f(0, 0)
gl.Vertex3f(0, 0, 1)
gl.TexCoord2f(0, 1)
gl.Vertex3f(0, 0, 1)
gl.TexCoord2f(1, 1)
gl.Vertex3f(1, +1, -1)
'/*Right (X+)*/
gl.TexCoord2f(1, 0)
gl.Vertex3f(+1, 0, -1)
gl.TexCoord2f(0, 0)
gl.Vertex3f(0, 0, 1)
gl.TexCoord2f(0, 1)
gl.Vertex3f(0, 0, 1)
gl.TexCoord2f(1, 1)
gl.Vertex3f(+1, 1, -1)
gl.End()
NEXT

It would probably produce a much better vertexbuffer and you wouldn't run into problems with textures like you do right now (pixelated edges).

Mo it wouldn't, and I don't run into any such problems with textures. Oh - you mean those templates made to be a few pixels tall and wide? those would always look pixelated at that resolution.

4x4 Pixel texture:

32x32 Pixel texture:

Myth busted as shaders could optimize it even further
The stretching on the bottom can easily by eliminated, either by using a second texture or picking texcoords more wisely - or by just generating a better texture which is not a quad.

 

[CasperN]

Keep to the topic for fuck sake...
use the pm for or something for the other stuff.
Edit: or use monks tread: http://www.gp32x.de...me/page__st__45

 

[torpor]

Meh. Lets just port Blender.

 

[Monk]

Myth busted as shaders could optimize it even further

Myth not busted as the "texture" is obviously not being displayed correctly. Why can't you use a 1x1 "texture" to achieve the effect you are getting?

The stretching on the bottom can easily by eliminated, either by using a second texture or picking texcoords more wisely - or by just generating a better texture which is not a quad.

That is correct.

In this particular instance (with the pyramid) if all one was trying to do was achieve a glowy wireframe pyramid, one could simply not fill in the bottom of the pyramid at all, that would improve the look and the poly count. But that isn't the goal here (where I wanted to show stuff simple enough that it could be designed on a Pandora) or in the thread it was first shown in (where I wanted to show that the items were UV Mapped without going to the effort of creating photo-realistic textures).

Monk: Isn't it enough to type in these manually in your text editor?

No.

...

It would probably produce a much better vertexbuffer and you wouldn't run into problems with textures like you do right now (pixelated edges).

No it wouldn't, and I don't run into any such problems with textures.

But just to reiterate - they're all dead Dave. Dave? They're all dead. All of them. They are dead.

I have not designed any textures to use on those models ergo I have not encountered the problems with textures that you are trying to correct me on.

They are all tempaltes Dave, they're all tempaltes, all of them are tempaltes. Read some of the above.

I submit that one could graphically design the models I've illustrated (AND more complex) on a screen as small as the Pandora. I submit that some peopel prefer to use GUIs than to type in text. I conclude that some kind of 3D modelling application would have its benefits. That's about the sum of what I've wanted to say in this thread, and I've seen nothing to convince me otherwise, just some peoiple inventing nits to pick at in order to be argumentative AFAICS.

 

[torpor]

I concur. We *shall have Blender* on the Pandora, I am quite sure. And, for those who use it, it shall be useful. What a concept!!!

 

[Monk]

Meh. Lets just port Blender.

I must admit, more and more frequently I regret even putting my hand up to offer to help out until someone better did so. We seem to have no shortage of people who think they can do better, and yet do not actually offer to do so.

As dflemstr and especially JayFoxRox are so much more knowledgeable in this area than I am, I'm going to do as I said I'd do in the first instance and back out of that project leaving them to produce the models for the GridWars-inspired game, knowing that they have the skill and the time to expend on the project where I have less skill and was less confident of the project than to expend the time it would need to do a better job.

Thanks for stepping up guys, even if it did take you a while

 

[torpor]

I'm going to do as I said I'd do in the first instance and back out of that project leaving them to produce the models for the GridWars-inspired game, knowing that they have the skill and the time to expend on the project where I have less skill and was less confident of the project than to expend the time it would need to do a better job.

No need, I've already integrated your models into my code, and I will release something .. soon .. I promise. Just need a few more spare hours to clean things up and find some nice TEXTURES, lol ..

 

[Monk]

I'm going to do as I said I'd do in the first instance and back out of that project leaving them to produce the models for the GridWars-inspired game, knowing that they have the skill and the time to expend on the project where I have less skill and was less confident of the project than to expend the time it would need to do a better job.

No need, I've already integrated your models into my code, and I will release something .. soon .. I promise. Just need a few more spare hours to clean things up and find some nice TEXTURES, lol ..

LOL, sorry that I haven't had the confidence/faith/skill combination to knock up even a few textures. I had some ideas, but I really wanted to see the code side progress a bit, so I could see an actual game coming into shape. I also didn't want to do lots of painting work when the model import capability was still so much work for you, as I had aspirations towards more complex models that would take more time and effort, and better UV mapping. So these really were as simple and, hmmm, preliminary as I'd warned you before creating the thread, IIRC.

I suspect that, as confrontational as JayFoxRox seems, he/she/it probably can do a MUCH better job of producing models and textures in short time than I can. I'd seriosuly consider making better models as the game comes together and allow for custom model sets - I certainly want a Millenium Falcom model for the player craft if only to pacify my son

That being said, lest my post sound churlish, you do OF COURSE already have my model designs to use however you see fit, and more in future if JayFoxRox lets me down are not a problem beyond time and space

 

[Tempel]

What the fuck guys? I used to respect most of the people in this thread. Why must you all be so upset with each other?

A lot of this insanity seems to be coming from the common problem of "Can X be ported to Pandora?" threads: the responses all become "why would you want to?", when the real question should be "is it possible?" No one ever dwells on that second question.

In summary and conclusion on the actual topic: Blender can probably be ported in some pared-down form eventually, given that it's been ported to the N900. In doing so, it will have some shortcomings, and some benefits. If anyone decides to take on this project, it will have at least a few appreciative users, though not everybody will find it useful.

Can we please be friends again now?

 

[CasperN]

Friends???? MONK is using Cinema 4D!!! :angry: :angry: :angry: we can't be friends with him! but love and forgivness to all the other in the tread

Now port that fuckin program!, I want it to the release date!

 

[Monk]

A lot of this insanity seems to be coming from the common problem of "Can X be ported to Pandora?" threads: the responses all become "why would you want to?", when the real question should be "is it possible?" No one ever dwells on that second question.

Agreed. Reluctantly I think some of the blame lies, incorrectly as it may, with the idiots who try to answer that "why would you want to?" question (myself included). We'd probably do a lot better if we could just let those responses go. Fortunately, this thread HAS been addressing the "is it possible" to a certain degree, and has also brought to light at least one possible alternative that's (IIRC) being actively worked on.

Can we please be friends again now?

Some of this (though perhaps not as much as would be best) is happening via PM rather than polluting this thread any further. Most of it seems to stem, AFAICS, from a mixture of misunderstandings and short tempers.

Friends???? MONK is using Cinema 4D!!! :angry: :angry: :angry: we can't be friends with him! but love and forgivness to all the other in the tread

Bugger - I've been caught out. Can I pretend it's an old Amiga version running under emulation? Or how about it being an ALMOST free copy (a paid "upgrade" to a magazine CD release"? Would either of those get me back in peoples good books?

Personally, I wish we could see more of the Pandora marketplace. I've only recently bought an XBOX 360 (because the XBOX is really on its last legs, it seems, and we have a pile of old games for it) and have seen their version of the AppStore - and I have to say that I understand how MS can make loads and loads of money from that. The XBOX 360 is just the way to deliver the MS Appstore seamlessly - give the Pandora the same kind of connectivity and a nice easy-to-use way of previewing applications - including several 3D modellers (why not?) and... phew, well. We'd be able to see what applications look like on a Pandora now, via a few of the preview screenshots - no installation (heck, no Pandora if it's done via a web site) required!

 

[Chip]

A lot of this insanity seems to be coming from the common problem of "Can X be ported to Pandora?" threads: the responses all become "why would you want to?", when the real question should be "is it possible?" No one ever dwells on that second question.

Because if you can't convince a dev of a valid answer to the first question, the second one is moot. Just because someone somewhere can think up a use for a piece of software, doesn't make it worth someone else's time to port it. You need to come up with some compelling reasons why someone should devote tens (possibly hundreds) of hours to taking a complex piece of software like blender, converting it to OpenGL ES 2.0, stripping out all the unnecessary or too-demanding features, re-arranging the UI for a small screen, and whatever else is necessary to make it work.

You wouldn't ask a stranger to spend an afternoon helping you move, but you have no problem asking a stranger to spend days or weeks writing code. The least you could do is justify your request. Since Fox is already aware of the effort/benefit ratio and decided against it, you can either try to convince someone else or port it yourself.

 

[Monk]

That raises an intersting (to me) question - do devs typically port programs that they themselves have no use for/interest in on the proposed target platform?

Although, in my head at least, Tempel was not so much asking JayFoxRox (or anyone specific, in fact) to port blender as he was asking that we just play nice and chat amicably rather than have these strange spats all the time. Which doesn't seem such a bad thing to want, IMHO

 

[Tempel]

You wouldn't ask a stranger to spend an afternoon helping you move, but you have no problem asking a stranger to spend days or weeks writing code. The least you could do is justify your request. Since Fox is already aware of the effort/benefit ratio and decided against it, you can either try to convince someone else or port it yourself.

On the contrary, I would never ask anyone to do any development; open source work can only be done of the developer's own desire. Of course, that desire is often influenced by usefulness to others, but it's not the sole concern. These threads seem to quickly establish some usefulness, then establish some reasons why it's not so useful (both important things to consider), then devolve into "yes it is!"/"no it isn't!" spats. I hope that, in the future, we can skip step 3 and simply summarize parts 1 and 2 with, to quote myself:

It will have some shortcomings, and some benefits. If anyone decides to take on this project, it will have at least a few appreciative users, though not everybody will find it useful.

At that point, it's up to a dev to find it and decide if it's worth his or her time.

So yeah, Monk got what I was going for. Thanks.

 

[wiimario]

Friends???? MONK is using Cinema 4D!!! :angry: :angry: :angry: we can't be friends with him! but love and forgivness to all the other in the tread

Now port that fuckin program!, I want it to the release date!

well, I'm using maya, and so?

 

[kin]

sorry, blender is too much on a panda in gui terms..

Let alone working with blender gui on a desktop client was hard enough.
Of all the 3d modeling programs, blender feels unnatural and can be annoying in the way it handles selections.

 

[Monk]

sorry, blender is too much on a panda in gui terms..

Let alone working with blender gui on a desktop client was hard enough.
Of all the 3d modeling programs, blender feels unnatural and can be annoying in the way it handles selections.

Without having tried Blender myself, I'll certainly conceed the possibility that you're absolutely correct about ALL of that. It's certainly managed to avoid ever getting on my "must try" list for desktops IIRC, which is a reasonable feat all by itself.

To continue - what would be YOUR best recommendation for something along the lines fo a good GUI mesh editor? I know that personally one of my "Can it run Windows" hopes would always have been Cinema4D 6, simply because that feels fairly lightweight for what it does - very responsive usually - and I know how to use it enough to push a few vertices around from time to time, when I need to. I've never tried Wings but would that or something else be your suggested better alternative?

 

[zapman]

Meh. Lets just port Blender.

Yes! Torpor, you seem to be having the skill and the right attitude to port Blender to the Pandora, could you do that?

Actually it would make most sense and would also probably be feasible to just port the BlenderPlayer to the Pandora: it would enable the Pandora to run games that were made with for blender game engine. THAT WOULD BE AWESOME!!!

I did some game with the blender gameengine (link), so I can say developing blender game engine games on the Pandora would be hard as the screen is just too small but PLAYING BGE games on the Pandora makes perfect sense!!!

--> Game development in Blender is easy and powerful thanks to python. The game engine speed is not the fastest but that's a price lots of game developers are happy to pay as long as they can develop a full game with just one program (Blender ). Also support for fancy OpenGL rendering wouldn't be that imporatant, nice games can also be done with simple textured faces. If we get .blend files to be playable on Pandora we would get tons of homebrew games: develop on desktop machine, play on Pandra (BlenderPlayer).

Torpor, could you look into porting the blender player to Pandra, please? I got a thread up at blenderartists a few months ago, Ideasman42 (who is one of the most skilled Blender devs) said porting Blender isn't really a problem and compiles just like that (link)!

 

[torpor]

It'd be really nice if the Pandora had some sort of modeller that could be used to create artwork for an OpenGL ES compatible engine ..

EDIT: I'm gonna look at this at some point, but I've gotta finish my music-PND first and there's tons of work to do on that ..