[Hool]

DeeperThought, on 08 May 2011 - 09:55 AM, said:

The version available at Fissile Materials is the most recent public release, and I don't think I have changed anything lighting related since then. I wonder what lighting effects you are referring to.

Polymer with DukePlus

• Shadows are casted against surfaces
  
• Flares can cast light
  
• Added lighting from existing Polymer Lighting

Polymer without DukePlus

• Shadows are not casted
  
• No Flares
  
• Standard Polymer lighting

I have also tried changing the commands:

r_pr_shadowcount 2
r_pr_maxlightpasses 2
r_shadescale 0.99
r_texcompr 1

To

r_pr_shadowcount 1
r_pr_maxlightpasses 1
r_shadescale 0
r_texcompr 0

Still crashes. So its DukePlus that is causing the problem. So can you help me out here DeeperThought? Can I lower the lighting but still keep DukePlus (Reloading, intelligent AI Ect.)?

Here is a new recent crashlog:
http://pastebin.com/3MkT2MK9

[Danukem]

WheatleyNukem, on 08 May 2011 - 10:53 AM, said:

Still crashes. So its DukePlus that is causing the problem. So can you help me out here DeeperThought? Can I lower the lighting but still keep DukePlus (Reloading, intelligent AI Ect.)?

I still need to get the latest HRP and test it. I'll try to get back to you soon. Let's try to keep these sorts of discussions in the bug reports forum in the future:

http://forums.duke4....elp-me-threads/

EDIT: I just played a few maps using the new HRP and the latest DP with the most recent build of EDuke32. Everything worked fine and there were no crashes.

[Helixhorned]

@WheatleyNukem

1. Are you absolutely sure you're using a non-ancient version? I introduced

"print 'cached tile' messages only from 5 ms up so as not to flood the OSD/log" in r1784, yet your log shows stuff like

"Load tile 2293: p0-m4-e0 "highres/sprites/decals/2293.png"... 1 ms"


2. Is an eduke32.crashlog produced? That should be much more helpful, but it's likely that it doesn't show up consistently with each different crash.

[MusicallyInspired]

WheatleyNukem, on 08 May 2011 - 10:53 AM, said:

Polymer without DukePlus

• Shadows are not casted
  
• No Flares
  
• Standard Polymer lighting

Standard Polymer lighting includes dynamic shadow-casting lights.

[Micky C]

Yeah, whether lights cast shadows or not really depends on the level you're playing as some maps don't have any shadow casting lights in them.

But I agree with DT that this one crash has been taking up a fair amount of the eduke32 and polymer thread. I've been wondering if this is a good idea or not but I went ahead and did it anyway, I made a thread that could potentially be used to discuss everyone's crashes, rather than have a new thread for each new crash, kind of like how the mapster32 problems and bugs thread works. I hope it can come as some use, here it is: http://forums.duke4....e-crash-thread/

[Jblade]

James, on 18 April 2011 - 04:24 AM, said:

The latest few snapshots give me compiling errors, it doesn't seem to want to accept rotatesprite commands in states anymore:
http://pastebin.com/DJsCednp

and here's my log from the last snapshot that works, dated 24th of March 2011:
http://pastebin.com/zejvV3y5

I'm still getting this problem even with the latest snapshot.

[Mblackwell]

I'd be curious to see AMCWEPS.CON since I'm definitely not getting any errors.

[Jblade]

 Mblackwell, on 13 May 2011 - 04:29 PM, said:

I'd be curious to see AMCWEPS.CON since I'm definitely not getting any errors.

Here's an example of one of the states Eduke has a problem with:

state displaytemp_jplasma

  setvar x 245
  subvarvar x WEAPSWAYX
  addvarvar x weapon_xoffset
  subvarvar x looking_angSR1
      
  setvar y 240
  subvarvar y WEAPSWAYY
  addvarvar y looking_arc
  subvarvar y gun_pos

switch weaponcount
case 1
addvar y 5
addvar x 5
case 2
subvar x 5
subvar y 5
case 3
addvar y 5
addvar x 5
case 4
case 0
  getplayer[THISACTOR].weapon_pos DISPLAY_TEMP2
  mulvar DISPLAY_TEMP2 -20
  rotatesprite x y NORMALSIZE DISPLAY_TEMP2 12104 shade pal 1024 0 0 xdim ydim
  getplayer[THISACTOR].random_club_frame DISPLAY_TEMP
   sin DISPLAY_TEMP DISPLAY_TEMP
   divvar DISPLAY_TEMP 1024 // 2^10
   addvar DISPLAY_TEMP 16 
  rotatesprite x y NORMALSIZE DISPLAY_TEMP2 12105 DISPLAY_TEMP 0 1024 0 0 xdim ydim
  break
endswitch

ends

This code works in the earlier snapshot I've used, it works for a long time. It seems the compiler's having a problem because it's in a state or something.

[Mblackwell]

The state shouldn't matter unless the state gets called in a weird place. I have rotatesprite in states everywhere. It's probably to do with your switch statement.

[Danukem]

 Mblackwell, on 14 May 2011 - 08:54 AM, said:

It's probably to do with your switch statement.

Yeah I would try changing that up in various ways and see what happens. Try adding a default case. If that doesn't work, add breaks after each case (and then make the necessary changes so it is functionally equivalent).

[Hool]

Reinstalled Eduke32, DukePlus and HRP. Still crashing, here is the new log:

http://pastebin.com/2zHECa1c

I don't know what is wrong.

[Danukem]

 WheatleyNukem, on 14 May 2011 - 09:45 AM, said:

Reinstalled Eduke32, DukePlus and HRP. Still crashing, here is the new log:

http://pastebin.com/2zHECa1c

I don't know what is wrong.

I don't know what to make of that, maybe one of the eduke32 developers could help interpret it.

It says that it crashed while the current actor was FIRE, but that's one of the actors whose code is completely unchanged from the original game. Also, I see the "clearmapstate" command being referenced, but that command is not used anywhere in the CONs. So I'm puzzled.

What was happening when the game crashed?

[Jblade]

Adding breaks and default cases didn't work - my states were working fine for a very long time and now the game doesn't want to take the rotatesprites being in a switch statement.

[Danukem]

James, on 14 May 2011 - 01:51 PM, said:

Adding breaks and default cases didn't work - my states were working fine for a very long time and now the game doesn't want to take the rotatesprites being in a switch statement.

Well that sucks, and it should be fixed. But in the mean time, if that's the only problem then you can easily move the rotatesprite command so that it occurs after the switch statement.

[Micky C]

WheatleyNukem, on 14 May 2011 - 09:45 AM, said:

Reinstalled Eduke32, DukePlus and HRP. Still crashing, here is the new log:

http://pastebin.com/2zHECa1c

I don't know what is wrong.

Dude.

Micky C, on 09 May 2011 - 04:41 AM, said:

But I agree with DT that this one crash has been taking up a fair amount of the eduke32 and polymer thread. I made a thread that could potentially be used to discuss everyone's crashes, rather than have a new thread for each new crash, kind of like how the mapster32 problems and bugs thread works. I hope it can come as some use, here it is: http://forums.duke4....e-crash-thread/

Bottom line, use this thread for your crash: http://forums.duke4....e-crash-thread/ The eduke32 thread, besides going off-topic every now and then, is really only used to discuss eduke32 features, or important suggestions or bugs such as what James and Deeperthought are talking about. It's not intended to be host to a post-a-thon of one person's crash. I empathize with your situation, it must be frustrating having the game crash like that, but please do follow the correct channels.

[Jblade]

I was just looking through the source code when I saw this:

#define ROTATESPRITE_MAX 2048

Does this mean there's a limit of 2048 rotatesprite commands allowed? or something else?

EDIT: ok I've just checked and I've already got over 2060 rotatesprite commands so it can't be that.

[Hendricks266]

It might be simultaneous rotatesprites.

[Danukem]

It's probably the maximum value that the coordinate parameters can have. About 2 years ago, I managed to crash the game with excessive numbers (from garbage data leftover in the vars I was using) so TX put in a check for max values. If you go over that it gives a warning, uses acceptable parameters, and does not crash.

[Jblade]

Both of those things would make sense, it's not as bad as I thought it meant in anycase!

Anyways I'm still getting the CON compiling error with the latest snapshot:
http://pastebin.com/Au7p43Bw
The multiple case warnings are wrong since I haven't done that - also I moved one of the rotatesprites out of the switch but it's still appearing in the list there. It's NOT in a state either, it's directly in the event command.

[The Commander]

Where are the three Stooges for this bug James is getting, this is holding up progress of the AMC TC.
I want to play it already.

[TerminX]

Send me the CONs.

[Master Fibbles]

This may come off as a dumb question but...

Is there a definitive polymer light tutorial around somewhere? I have done a search here and on the wiki for both se50 and spotlights but I have come up with nothing that is helpful. Either I'm searching for the wrong thing or something but I can't find out what I'm doing wrong.

[MusicallyInspired]

There's instructions on the first post in this thread.

[Master Fibbles]

So this:

Quote

PLACING POLYMER LIGHTS IN MAPSTER

SE 49 = point light
A SECTOREFFECTOR with lotag 49 is a point light.
HITAG = light distance
The HITAG of the SE is the maximum distance, in build units, that the light will travel. Note that it is impossible to enter a number higher than 65535.
XVEL/YVEL/ZVEL = RGB values
To edit these, press F8 on the SE in 2D mode, then select them and type in numbers. Numbers should be in the range 0-255
TRANSPARENCY = priority
a solid sprite is a high priority light, press T once (33% transparency) for lower priority, again (66%) for lowest priority

SE 50 = spot light
light distance, color and priority are as above
ANGLE = LIGHT ANGLE (duh)
SHADE = width
shade values are between -127 and 127, each gives a different width....experiment
EXTRA = horiz (up/down angle)
set EXTRA by pressing 'M on the sprite. 100 is straight ahead, higher positive values angle up, and values lower than 100 (including negative values) angle the light down.

[Micky C]

Yeah that's about as comprehensive as it gets. Play around with them for a bit and one day you might be as good at placing lights as me

[Master Fibbles]

I should vote your arrogance down.

I wasn't able to find this on the wiki which is rather odd. It should be in a place where anyone can find it. I just copied and pasted it into its own txt file and placed it in the folder I have for map FAQs.

Unrelated, but I searched for SE 50 and it errored because the search terms were less than 3 characters and a search for SE50 comes up with nothing. This actually makes searching this forum for this information almost impossible. Which, of course, brings me back to my first point: this should be in the wiki. Maybe in the mapping section that is really the Map FAQ from the early days.

[Danukem]

Mr.Flibble, on 19 May 2011 - 10:01 PM, said:

Which, of course, brings me back to my first point: this should be in the wiki. Maybe in the mapping section that is really the Map FAQ from the early days.

I guess you would put it here: http://wiki.eduke32....Reference_Guide

And put a more condensed version of the details in the explanation field.

[Micky C]

Done: http://wiki.eduke32....r_Lotag_Listing
That wiki code was a nightmare to figure out

[Jblade]

The latest snapshot fixes my problem with the CON code, big thanks to TerminX for helping me with that - I appreciate it

[Master Fibbles]

DeeperThought, on 19 May 2011 - 11:18 PM, said:

I guess you would put it here: http://wiki.eduke32....Reference_Guide

And put a more condensed version of the details in the explanation field.

Yeah. That was the first place I looked too and it wasn't there. I expected it to be there. Thanks to whomever added it.