[Mister Sinister]

Dear Eduke devs, can you pinpoint me the moment when shrink ray behavior was altered? Was it altered even?

[The Battlelord]

Mister Sinister, on 06 April 2021 - 03:11 PM, said:

Dear Eduke devs, can you pinpoint me the moment when shrink ray behavior was altered? Was it altered even?

I think that it is related to the cliping behaviour that has been changed in the r7432 and many later revisions, where you can litterally push monsters and freely kick'em.
You can try r7395 if you are looking for a "normal" behaviour, well, if the shriker was ever considered to have a normal behaviour.
(Please, someone correct me if am wrong)

[spacekebab]

Getting a CTD with at least the last few versions when I shoot the hungjury actor. Same issue vanilla or modded - not sure if limited to linux

CONLOGVAR: L=45586 picnum  (Global) =4943
./eduke32[0x567d7c]
/usr/lib64/haswell/libc.so.6(+0x43290)[0x7fb1d5c5d290]
./eduke32[0x4622d4]
./eduke32[0x46f99d]
./eduke32[0x45c317]
./eduke32[0x40c2e7]
./eduke32[0x45c3e4]
./eduke32[0x42a804]
./eduke32[0x41fe99]
./eduke32[0x5b9c5e]
./eduke32[0x44e9e5]
./eduke32[0x4076be]
/usr/lib64/haswell/libc.so.6(__libc_start_main+0xd5)[0x7fb1d5c44ba5]
./eduke32[0x40a58e]

    0: -16: 0
    1: -15: 0
    2: -14: 0
    3: -13: 103
    4: -12: 0
    5: -11: 4
    6: -10: 1
    7:  -9: 1
    8:  -8: 8
    9:  -7: 0
   10:  -6: 436
   11:  -5: 103
   12:  -4: 0
   13:  -3: 4
   14:  -2: 1
   15:  -1: 1
   16:   0: 8
   17:   1: 0
   18:   2: 436
   19:   3: 103
   20:   4: 0
   21:   5: 2
   22:   6: 1
   23:   7: 1
   24:   8: 16
   25:   9: 0
   26:  10: 436
   27:  11: 103
   28:  12: 0
   29:  13: 8
   30:  14: 1
   31:  15: 1
   32:  16: 4
   33:  17: 0
   34:  18: 436
   35:  19: 103
   36:  20: 0
   37:  21: 10
   38:  22: 1
   39:  23: 1
   40:  24: 7
   41:  25: 0
   42:  26: 436
   43:  27: 103
   44:  28: 0
   45:  29: 1
   46:  30: 5
   47:  31: 0

current actor: 3738 (0)
g_errorLineNum: 0, g_tw: 1

either way - weird issue

[Danukem]

Might be linux only. I loaded up E4L5 and clipped into the hung jury room, jibbed all the JURYGUY sprites with no issues.

[FRVIND]

Great update from Plagman and Pierre-Loup Griffais. Now at least Polymer has gained 10 fps and and sorry if this is little!! .
Heartfelt thanks

[Kristian Joensen]

Wait a minute... Plagman started working on Polymer again? Wonderful news!

[NightFright]

And I thought the renderer had been abandoned. Well, it still has potential.

[LeoD]

Na na na na na na na na na na na na na na na na... PLAGMAN!

Spoiler

[MusicallyInspired]

Plagman is the (Plag)man.

[Mark]

FRVIND, on 17 August 2021 - 02:08 AM, said:

Great update from Plagman and Pierre-Loup Griffais. Now at least Polymer has gained 10 fps and and sorry if this is little!! .
Heartfelt thanks

Is there a link to the source of this announcement?

[FRVIND]

https://voidpoint.io...ge_requests/154

Regarding the fps I have tried in E1L1 in two very precise points and with precise angles, then also in some of my personal maps. There is no doubt especially if there are skies on the map. However, the fps problem was already partially overcome with the new RTX cards

[Mark]

So far I checked in only 1 of my Polymer projects which was using an eduke revision from about a month ago and updated to the latest. No difference in FPS in a few spots in a map. This is on ancient hardware. A 12 year old comp and 4 year old video card. I'll test in other projects of mine that are using older versions of eduke32.

[Mark]

You mentioned a 10 fps increase. What was your framerate so we can tell if it was a large pecentage increase or not.

[FRVIND]

Ok, ok in E1L1 when you get off the roof in the open space I did, at the worst angle and in the absence of monsters in action (already killed), about 38-40 frames. Now after the update 48-50

[FRVIND]

Integrated Asus Rog Strix Scar II Intel laptop plus RTX 2060.
But these advantages are not always visible, it depends on the type of map. I think open spaces are typical ...

[Mark]

Any little bit helps. I still have a backlog of projects on my drive waiting to get completed. All using Polymer renderer. Been using it since my first mod in 2009.

[FRVIND]

My belief is that for the pc things get heavy when there are a certain number of lights (SE49 or SE50) colliding with each other. But more when the lights intercept the light sources written in the maphacks ... but the latter are fundamental for an incomparable brightness of the scene

[MusicallyInspired]

Man the Polymer HRP pack looks so bad without texture filtering.

[FRVIND]

MusicallyInspired, on 19 August 2021 - 06:41 AM, said:

Man the Polymer HRP pack looks so bad without texture filtering.

Texture filtering? My r_texfilter is set to 2, my r_pr_shadowfiltering is set to 1. As default and this is good.

[MusicallyInspired]

Was it reenabled then? I thought they removed it from the the engine for Ion Fury?

[FRVIND]

MusicallyInspired, on 19 August 2021 - 07:32 AM, said:

Was it reenabled then? I thought they removed it from the the engine for Ion Fury?

I know it's not selectable now from video settings, but from console commands...

[MusicallyInspired]

Well then.

[Mark]

I was happy to see it put back as an option. I have projects that I deemed unplayable because the swirling moire patterns and texture edges were so bad without filtering. I posted about it back in January and Hendricks kindly pushed a change to allow filtering again. Just not by default. I had to add a few lines to my autoexec.cfg file.

[FRVIND]

https://voidpoint.io...fd762aa72a6cbb4

and...

https://voidpoint.io...e590b55cc3ddcfa

[FRVIND]

FRVIND, on 19 August 2021 - 12:52 AM, said:

Ok, ok in E1L1 when you get off the roof in the open space I did, at the worst angle and in the absence of monsters in action (already killed), about 38-40 frames. Now after the update 48-50

I forgot to indicate one important thing:
-surface lights 8
-shadows per surface 6.
If I lower these values ​​respectively to 4/3 then there is a boost in performance. I attached a my complete map to evaluate its potential (I reach 18-22 fps in the worst case). Is a map created specifically to study the collision of lights but also multiple TROR layers (by the way if someone thinks of playing the map with polymost it's a disaster, polymost could never achieve the same quality and functionality)
[the map has an unimportant bug: it may happen that sometimes a monster manages to escape from its prison, but the confusion is so great ... ]

Attached File(s)

•  Ciccior.zip (45.35K)
  Number of downloads: 406

[Mark]

On a 12 year old Dell computer and an old nvidia 750ti video card. Using the HRP, Polymer, your settings and at 1920 x 1080 I had 14fps at the start and a max of 28fps when I faced an outside wall. Averaged about 18fps while playing. If I turned down my settings to their usual 4 and 2 I gained 2 fps.

Were you using the HRP and models also?

[FRVIND]

... a fantastic result. I use also models . So does it mean that the duke engine has remained 10 years ago and benefit very little from the new video cards?

[MusicallyInspired]

My framerate averages between 180-220. It dipped under 100 when I blew up the building in E1L2. Sometimes in quiet areas it gets up to 400. This is with the Polymer HRP. But I have a 2060. (1080p)

[FRVIND]

I also have an RTX 2060 and I didn't realize how You can reach these results. Did You refers using classic or polymost? And without HRP models, isn't it?

[MusicallyInspired]

I was using Polymer because of the new update Plagman made.
In Polymost without any HRP or models I get 2400 fps.
In Software I get 300-500 fps.