Screen goes black in Polymer renderer under consistent circumstances, R7793 and R7800
#1 Posted 16 July 2019 - 03:23 AM
Specs of system:
- Intel Core i7 4770K
- Z87 Chipset
- NVIDIA GeForce GTX 970 (drivers have been updated as part of testing)
- Windows 10 1809.
- EDuke32 64-bit
EDIT: R7743 is the first release on Synthesis this happens to me. Any older release has no problem.
This post has been edited by Mitch Richters: 16 July 2019 - 04:14 AM
#2 Posted 16 July 2019 - 08:32 AM
The Battlelord, on 29 June 2019 - 06:07 PM, said:
it happens when there are mirrors in the area.The game keep running, indeed the player can move, fire, jump etc... but is pratically impossible to play.If you switch from Polymost to Polymer in front of a mirror, boom, black screen, then, I guess that it could turn back to normal if you switch to Polymost again, but i was unable to do it as everything is black, you should learn the path via keyboard and try, but i don't think that is worth the time .
This post has been edited by The Battlelord: 16 July 2019 - 08:42 AM
#3 Posted 16 July 2019 - 09:44 AM
EDIT: 7657 was good. the next public revision 7743 was bad. So the issue is sandwiched in all those unreleased revisions. Unfortunately its a whole lot of them. 7733 and 7736 in the changelog mentions mirrors but for software and polymost. But I suppose the bug could come from many other places.
This post has been edited by Mark: 16 July 2019 - 09:58 AM
#5 Posted 16 July 2019 - 02:16 PM
#6 Posted 16 July 2019 - 02:37 PM
The Battlelord, on 16 July 2019 - 08:32 AM, said:
Ah ok, no worries. Apologies, I didn't know it was reported but glad I'm not the only one
Mark, on 16 July 2019 - 09:44 AM, said:
My findings also.
LeoD, on 16 July 2019 - 01:54 PM, said:
Awesome, thanks for doing that leg work! I don't compile EDuke32 so I rely on builds in Synthesis.
#7 Posted 16 July 2019 - 02:40 PM
#ifdef USE_OPENGL if (videoGetRenderMode() != REND_CLASSIC) polymost_completeMirror(); #endif
#8 Posted 18 July 2019 - 02:14 AM
Are these forums the way to submit bugs? Is there a formal bug tracker like Bugzilla or Flyspray etc?
#9 Posted 18 July 2019 - 02:27 AM
Polymer at the moment is lowest of the low priorities regarding dev due to all hands being on deck with dealing any remaining things for fury.
#10 Posted 18 July 2019 - 02:31 AM
oasiz, on 18 July 2019 - 02:27 AM, said:
Polymer at the moment is lowest of the low priorities regarding dev due to all hands being on deck with dealing any remaining things for fury.
Wouldn't Iron Fury be using Polymer code though, or does it use Polymost?
#11 Posted 18 July 2019 - 02:56 AM
Mitch Richters, on 18 July 2019 - 02:31 AM, said:
Ion Fury uses Classic and Polymost only.
#12 Posted 18 July 2019 - 03:06 AM
If it didn't offer a notable visual advantage to polymost, it would have been axed a long time ago.
#13 Posted 18 July 2019 - 03:11 AM
Lunick, on 18 July 2019 - 02:56 AM, said:
Thanks for clarifying
oasiz, on 18 July 2019 - 03:06 AM, said:
Thanks for the insight to this. It's amazing how out of date the Wiki is. It makes it sound like it's the future lol. I'll give up on it and stick to Software or Polymost .
#14 Posted 18 July 2019 - 03:50 AM
#15 Posted 18 July 2019 - 03:57 AM
Mark, on 18 July 2019 - 03:50 AM, said:
I do like it better because its FOV feels more natural and less stretched in the corners (Polymost feels 2.75D, if that makes sense? Seems stretched further at the corners). Just feels very hit and miss though. Bug that's existed for months getting no love, lead developers working on Ion Fury have abandoned it, etc.
Don't worry though, as per the attached I have many scripts ready to chop and change between mods and renderers as necessary
#16 Posted 18 July 2019 - 10:48 PM
Mitch Richters, on 16 July 2019 - 02:40 PM, said:
#ifdef USE_OPENGL if (videoGetRenderMode() != REND_CLASSIC) polymost_completeMirror(); #endif
These if statements were definitely the cause. Confirmed fixed in R7809 and tested with R7811.