[Micky C]

I'm one of the testers, that's my screenshot two posts up, but it's hard to see what's going on there so I made a video. The curious thing is that the glitches only happen when you're in the top TROR layer, not the bottom layer.

Here's my log file: http://pastebin.com/raw.php?i=E5pfcS5B

[Mark]

I have no glitches when playing the map.

[Gambini]

care to post your log? it may be useful to gather information of the people who see glitches vs the people that see it all fine.

[Mark]

here's the log

Attached File(s)

•  eduke32.zip (1.23K)
  Number of downloads: 630

[Player Lin]

Mark., on 16 February 2013 - 08:54 PM, said:

here's the log

Mark. eduke32 log said:

EDuke32 2.0.0devel r3423
Compiled Jan 21 2013 13:21:07
Using D:/EDUKE3423/ for game data
Windows Vista Service Pack 2 (build 6.0.6002)
Initialized nedmalloc
Initializing DirectDraw...
Searching for game data...
Using "DUKE3D.GRP" as main game data file.
Using group file "autoload/polymer_hrp.zip".
Compiling: EDUKE.CON (3116 bytes)
Including: GAME.CON (151281 bytes)
Including: DEFS.CON (35992 bytes)
Including: USER.CON (42939 bytes)
...

I guess that autoload one is not needed, isn't it? When we're testing polymost...

==============

I got some visual glitches like attachments, also my log and cfg file.

(I'm using my custom CONs so the player stat and some game string is differ from default but I guess it's not matter with this test...right? )

Attached thumbnail(s)

• 
• 

Attached File(s)

•  eduke32_r3480_log&cfgs.zip (4.99K)
  Number of downloads: 520

[Gambini]

Thanks! You took shots from above the island. I think it is expectable to have visual glitches from that point. From the other spots, where most people showed glitches did it look alright?

[Player Lin]

Gambini, on 17 February 2013 - 10:16 PM, said:

Thanks! You took shots from above the island. I think it is expectable to have visual glitches from that point. From the other spots, where most people showed glitches did it look alright?

I can only get that glitch happens on that position, other points are just looking fine as well. I may try again later...

EDIT: OK, I found two other one, both are still above the TROR sector so maybe still not useful too? The -4 one don't needed jetpack but jump on the shelf.

Attached thumbnail(s)

• 
• 

[Gambini]

[Plagman]

So everyone gets the same glitches then? All the screenshots are from different viewpoints so I don't see any conclusive evidence. The one shot where the glitches looked like they happened where Gambini was looking from was taken with a very old version of EDuke32.

[Gambini]

So far I, Mark, Player Lin and from what I´ve spoke to, Mike Norvak can see TROR constructions with the same quality than when using Software. People that either posted here or talked to me that see it all glitchy are Quakis, Micky C and Mikko Sandt. From those three Micky owns an Nvidia card. Many Eduke32 versions are used, true, but you can´t expect people updating their copy every two hours. I and Mikko Sandt have tried both with several snapshots and there´s no difference at all.

And I don´t see the need of posting screenshots of the exact same spot. This is not a trial. If they say seeing it all glitchy/glitchs free we can trust them.

EDIT: I can keep performing as many tests as you want to get to the bottom of this issue. Just tell me what to do and I´ll do it.

[Plagman]

You're right, Micky's video is very weird. I'm not seeing the glitches personally. Micky, if you create a brand new Duke3D directory, only drop the latest snapshot and Gambini's trortest.map in there, make sure to remove any duke3d.cfg, etc, do you get them as well? I see some autoloaded stuff in your log and I want to narrow it down as much as possible in order to find what's causing this different behavior.

EDIT: That's assuming The Big Cheese is Micky.

[Micky C]

The problem is still there with only duke3d.grp, Gambini's test map, and the latest eduke32.exe:

http://pastebin.com/raw.php?i=GsTdAMte

[Plagman]

And they happen both in the game and editor, exactly the same? And both classic and Polymer are fine?

[Micky C]

As far as I can tell, it's the same glitches in both mapster and eduke with polymost. Classic and polymer look perfect though.

[Gambini]

So. Yet we have to find an ATI user that sees it glitch free. That would be the exception that confirms the rule .

Anybody reading this, that uses those cheaper video cards, is more than encouraged to try the test map. And if it looks all smooth, in Polymost, come here to tell us and we may gift you a Nvidia, Plagman´s courtesy.

[Plagman]

This is a real mystery to me. Micky, which NVIDIA driver version are you using? I strongly doubt this is related, but who knows.

[Micky C]

Version 306.97 apparently, at least that's what it says when I press the home button in nvidia control panel.

I'll try updating to the latest driver and see if that changes anything.
Unless I should wait for something first?

[Mark]

I'm running a 2+ year old driver 258.96 At the time I bought this computer, used, I did some browsing on various tech sites and this version was listed as the best at the time. I never updated since. Most of the fine print that accompanies those updates didn't seem like they had anything to offer my now outdated card.

[Player Lin]

I was planned to update the driver for my 8600GT when I noticed I still using 285.58 but I just don't want setting my override refresh rate again...

The latest one for my system is 314.07 WHQL, will try it and see any differences or not...

EDIT: No, it still totally the same at all.

[Fox]

Is there any planning into including translucency (0-255) for normal sprites and rotatesprite? Since Polymost and Polymer doesn't use lookup palettes, I don't understand why would it be limited to 33% or 66% pre-defined flags.

[Gambini]

You can embed an alpha channel to your texture.

[Fox]

Which is not a good solution. And I want specifically to display a text with STARTALPHANUM which gradually fades away. That makes up 93 sprites × all palettes × 255. There aren't enough tiles avaiable in Eduke32 for that. Yek.

[Mblackwell]

If you want a gradual fade you need ... (let me think) probably 3-4 versions of every tile. Fade until right before 66% opacity, and then flip orientation bit 1 and start back at the first tile, then flip on orientation bit 32 after you hit close to 33% opacity and start over yet again.

Not the most convenient, but it would work.

Or you could make the text always transparent and stack multiple layers (and play with orientation to boot).

Again, not convenient but possible.

[Plagman]

Isn't that what the spriteext alpha member is? Is the issue here that it's not hooked up to rotatesprite, Fox?

[Helixhorned]

Yup, it's accessible from CON using {get,set}actor[].alpha or the sprite[].alpha inline array syntax even if it's not a member of either the internal actor[] or sprite[] arrays, but of the spriteext[] array. The values are on a 0-255 scale. Full alpha control for rotatesprite etc. is a thing for the future.

[Fox]

It's Polymer only. =S

[blizzart]

I just want to ask what´s the max numbers of pals I can define in a .def file.
For Brute Force I want to use different pals for model skins, so we can vary skin colors and clothes of enemies and NPCs.

Also I want to define new fogpals, too.

[Mark]

I took a quick look at the HRP defs and saw a few tiles that used 16 pals. No idea about the max allowed.

[Mblackwell]

Wasn't it 0-255?

[TerminX]

Yes, but a few of the higher values are reserved.