[Spiker]

I'd rather we never had a final release

[Danukem]

supergoofy, on Jun 13 2009, 01:19 PM, said:

that's true

but all the official snapshots are beta builds that considered stable, but not final.

Yes, but consider the expectations people have for polymer, after all the attention it has gotten. Also, there is a difference between releasing something with a new feature that is finished-as-far-as-you-know, and releasing something with a feature that you know is not finished, or has known bugs.

Having said that, I have no idea what is holding up the release and I doubt that our speculation would make any difference one way or another.

[gt1750]

supergoofy, on Jun 13 2009, 09:54 PM, said:

I believe that r1429 is stable enough to be candidate for official snapshot

Might be stable enough for you but not for others.

For me, all the recent polymer builds crash randomly (even when standing still with little to no action happening), usually after 10 minutes of gameplay. R1431 is no exception. I've attached the log if anyone is interested. Also it doesn't run very well on my system (Phenom X4@2.5GHz, 4GB RAM, ATi Radeon 4850 512MB, Vista 64) the framerate in E1L1 with polymer HRP and Parkar's light setup drops below 1fps at the very beginning and keeps below 20fps in the cinema cloakroom.

Attached File(s)

•  eduke32.log (13.53K)
  Number of downloads: 1126

[supergoofy]

Probably you have problem due to bad opengl driver. ATI drivers are known to cause such problems. I have a slower/older system (cpu/gpu), but at least I have nvidia gpu. The frames are usually above 30 during gameplay.

[gt1750]

I've used ATi cards since 2003 and I've never had any "ATi-specific" OpenGL issues in a game until now. I'd love to enjoy Polymer in it's full glory but there's no way I'm buying a new nVidia card just for that.

BTW the previous log was created while playing with older (9.4) Catalyst installed. I've just updated the driver to 9.5 but everything remains the same - performance, crashing...

[supergoofy]

Question: Have you tried to run eduke32 with only one cpu core?

In Windows XP:
1. Run eduke32
2. Don't press start yet to load the game
3. Open Task Manager go to Processes and find eduke32.exe
4. Right click on it, choose Set Affinity and then uncheck all cores except Core0
5. Click ok and close Task Manager
6. Go back to eduke32 window, click on start
7. Lets see if it works

For Vista users, probably you should follow similar procedure.

Though I doubt if this has any effect. The problem is with the Ati drivers.

[gt1750]

No. Why would I do that? Isn't EDuke32 already optimized for multi-core systems?

[supergoofy]

Just for testing purposes. You never know with computers. Sometimes the things that seems to be impossible to cause problems, actually are the troublemakers. Just test it you never know.


Personally I played with r1431 E1L1, E3L1 and E3L2 and all were working well. And the frames were above 30.

[atlas]

thank you for the new screenshots, they are lovely : )

i think before any candidate is released to the public there should be a good set of artwork to go along with the release - so we don't have any features that prospective players would miss. what i mean is glow maps fine tuned and adjusted, and maybe (if we can gather a team together) polished normal maps for 8-bit and hrp. this is what i was expecting when i first heard of the polymer engine. i think alot of people had their idea of what they wanted from the new version of eduke, and we should focus on keeping them all happy. the idea is to build the community, not lose people because we didn't listen to everyone's ideas.

if you need beta testers to help make this new engine the best and most polished engine it can be before release. count me in.

[TerminX]

I think the only thing really holding up a release--and even then, it's more just something I want done rather than something that has to be done--is a possible performance improvement from changing how sprites are handled when rendering. I don't know when Plagman will get around to that... if it's going to be later rather than sooner I might release a build before then.

I also still haven't coded any of the CON hooks for lighting, nor have the spotlights on recon cars been redone yet, et cetera. There's also the issue of needing new menus for Polymer related stuff at some point... and the issue of the menu code in Duke being so horribly awful that it basically NEEDS a full rewrite to do anything useful or to be flexible at all.

And then there are the problems in the editor...

And then there are the bugs with maphack based lights...

Needs more polish!

[0815Jack]

gt1750, on Jun 14 2009, 01:02 AM, said:

Might be stable enough for you but not for others.

For me, all the recent polymer builds crash randomly (even when standing still with little to no action happening), usually after 10 minutes of gameplay. R1431 is no exception. I've attached the log if anyone is interested. Also it doesn't run very well on my system (Phenom X4@2.5GHz, 4GB RAM, ATi Radeon 4850 512MB, Vista 64) the framerate in E1L1 with polymer HRP and Parkar's light setup drops below 1fps at the very beginning and keeps below 20fps in the cinema cloakroom.

i have managed to compile the polymer source 1429 by myself and i have played it with the polyper hrp a couple od hours....not a single crash. i didnt found any ati related issues.....

E1L1 is not running with high framerates too .... mostly below 20 fps ..... but this was explained by plagman..

my system: Core 2 Duo 8400, 4GB RAM, ATI Radeon HD4350 512MB, WinXP sp3 x86 / catalyst 9.4#


btw. can someon explain me the light setup

//light sector x y z range r g b radi fade angle horiz mins maxs prio tile

[Chip]

I'm getting really curious of how / if this will run for me since I see people saying it doesn't work for them at all well and that's in between crashes!!! Those people have a better CPU but the same graphics card as me.

Oh dear, I may have waited all this time for nothing
I just hope that my computer is more amazing then the equivelents out there and all this works for me.

[SwissCm]

Chip, on Jun 14 2009, 09:18 PM, said:

I'm getting really curious of how / if this will run for me since I see people saying it doesn't work for them at all well and that's in between crashes!!! Those people have a better CPU but the same graphics card as me.

Oh dear, I may have waited all this time for nothing
I just hope that my computer is more amazing then the equivelents out there and all this works for me.

You'll be OK if you have an nVidia card. ATI cards have some trouble with it due to a buggy OpenGL drivers and god help you if you have an intel chipset.

[NightFright]

I agree that a release should not happen before the devs - in this case, TerminX and Plagman - are 100% satisfied with it. I think most users prefer to get a polished snapshot rather than being beta testers of some version that still has issues of some sort. All this Duke-related stuff is about patience, anyway, and I am willing to wait as long as it takes. Unless it is not... forever.

[supergoofy]

There is also the possibility of releasing a snapshot of the latest code without polymer.

[Danukem]

Setting mdflags to 1 on a sprite does not work in Polymer (the sprite is still rendered as a model).

[Plagman]

DeeperThought, on Jun 14 2009, 06:33 PM, said:

Setting mdflags to 1 on a sprite does not work in Polymer (the sprite is still rendered as a model).

Ah, yeah; I forgot about that. Thanks for reporting it.
In the future, can you file any issues in this tracker: https://sourceforge.net/tracker/?func=brows...amp;atid=706724
I often miss stuff on the forums or just forget about reports like that, so it would help me a great deal with keeping track of outstanding bugs/missing features.

[Sobek]

SwissCm, on Jun 14 2009, 09:59 PM, said:

You'll be OK if you have an nVidia card. ATI cards have some trouble with it due to a buggy OpenGL drivers and god help you if you have an intel chipset.

What's the go with Intel chipsets, if I may ask?

[Jimmy]

TX, on Jun 14 2009, 12:27 AM, said:

I think the only thing really holding up a release--and even then, it's more just something I want done rather than something that has to be done--is a possible performance improvement from changing how sprites are handled when rendering. I don't know when Plagman will get around to that... if it's going to be later rather than sooner I might release a build before then.

I also still haven't coded any of the CON hooks for lighting, nor have the spotlights on recon cars been redone yet, et cetera. There's also the issue of needing new menus for Polymer related stuff at some point... and the issue of the menu code in Duke being so horribly awful that it basically NEEDS a full rewrite to do anything useful or to be flexible at all.

And then there are the problems in the editor...

And then there are the bugs with maphack based lights...

Needs more polish!

TX, if you recode the menu, do you think you could recode so that they can be manipulated through the cons? That would be an awesome feature, and mods like DukePlus could really benefit from it!

[Tea Monster]

Sobek, on Jun 14 2009, 06:17 PM, said:

What's the go with Intel chipsets, if I may ask?

Intel chipsets were made for office PC's Like those little HP desktop models that are in nearly every office in creation. They are made to run Office Apps and Internet Explorer. They are designed to be manufactured cheaply and do typical office tasks. They are NOT designed to play Quake 4 at a decent frame rate as most offices do NOT need to do that!

A lot of OEM's grabbed these motherboards to make cheap home PC's.

Only after this happened, did Intel belatedly try to shoe-horn in some Direct X functionality to their drivers and chip sets. Sadly, thier attitude to OpenGL is "Huh? What's that?"

[Sobek]

Tea Monster, on Jun 15 2009, 01:49 PM, said:

Intel chipsets were made for office PC's Like those little HP desktop models that are in nearly every office in creation. They are made to run Office Apps and Internet Explorer. They are designed to be manufactured cheaply and do typical office tasks. They are NOT designed to play Quake 4 at a decent frame rate as most offices do NOT need to do that!

A lot of OEM's grabbed these motherboards to make cheap home PC's.

Only after this happened, did Intel belatedly try to shoe-horn in some Direct X functionality to their drivers and chip sets. Sadly, thier attitude to OpenGL is "Huh? What's that?"

Oh you mean Intel's integrated video chipsets? Could have just said that... By 'intel chipsets' I automatically assumed you meant the physical chipset like P35 or X38.

Naturally, just about any integrated video solution is a no go, for this game or any other. I'd consider that a given

[SwissCm]

Sobek, on Jun 15 2009, 01:45 PM, said:

Oh you mean Intel's integrated video chipsets? Could have just said that... By 'intel chipsets' I automatically assumed you meant the physical chipset like P35 or X38.

Naturally, just about any integrated video solution is a no go, for this game or any other. I'd consider that a given

There are some higher-end intel video chipsets (GMA X3000 onwards) that may be able to theoretically run Polymer, though the drivers are likely to not be mature enough to really pull it off with any reasonable success.

[TerminX]

SwissCm, on Jun 14 2009, 09:03 PM, said:

There are some higher-end intel video chipsets

SwissCm, on Jun 14 2009, 09:03 PM, said:

higher-end

Quote

intel video chipsets

[SwissCm]

TX, on Jun 15 2009, 02:23 PM, said:

Ha

[Benichon]

Two new screenshots from rev1433:

Episode 1 Level 1:


Episode 1 Level 2:


PS: I've reported two bugs for Polymer (rev1433) in sourceforge.

Edit : it is just an idea of what it could be after the HRP is finalize (normals textures are created by myself not by official workers). I think some official people who helped since many years will do really better normal than me (bad normal and bad defs from me). And like what developpers said, the developpement can take many times before finish. when we are in some place with many light and effect it goes slow and there are still bugs and many things to add so don't think it's going to be finished (i hope i didn't say anything wrong). .

Continue it Plagman and TX ^^.

[HJ-kos]

i'ts all great... but i think ONE thing is must be fixed. I mean HUD weapons. Would there be any way to make reflections\lightnings on weapons? Like for now pistol always have light on top left edge, even if it's dark or light source is on right or below duke or even switch glowing red close to duke...

[Piano Man]

Now that we got the new render being finalised... I think we should now have a no-holds rule on texture limits and polygon counts...

I mean some of the models looked great 2-3 years ago... Now they're getting old and look kind of ancient....

ANd with the new render looking the way it is... I say to anyone creating textures or models - DON'T HOLD BACK... Give us 10,000 polycount if you want... We have computers these days that can render 1,000,000 triangles and points on screen... Why should we hold back with the HRP project. Would go nicely with the new render.

I also proposed an idea before 3DRealms shut down (due to incompetence on their part for not releasing a game in 12 years) that if we were to re-make the Duke Nukem Atomic levels using the new render and add better backgrounds and better level design... NOW BEFORE anyone says it's illegal... we wouldn't be using the original maps or editing them as THAT IS illegal... We would be using them as a basis and creating the new maps from scratch to better incorporate a more realistic design and push the game to it's limits. Would be nice to know a rough release date on the new render as I'm dying to try it out.

[Jimmy]

Lights are too purple, and the maps on that metal texture look fucking atrocious.

[Jaxx]

Is there any release date?

[TerminX]

Congratulations, you're the millionth person to ask the same question! Your prize is a punch in the butthole.