Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[SAS~Storebror]

Arctic Summer

[danperin]

In my game simply never becomes night 

+1!

[SAS~Storebror]

A night plus 1? That's two nights then.

[danperin]

A night plus 1? That's two nights then.

Extra dark, mate! 

[SpongeBob]

Okey, Vega is right.

I've got Graphic Extended on a BAT install.

These are two views over the Aleutians at 12:00 (daytime):






These are more or less the same views at 23:00 (it should be night):





As you can see, the plane shows the typical night illumination, but the sky is as blue as during daytime.
I cannot tell if in the two last pictures we see the Moon or the Sun (I bet for the Sun), but it doesn't show the 'solar corona' as in the two first pictures.
Pay attention that the Sun is at the same location in both sets of pictures. I haven't checked yet, but it seems that it doesn't move along the day (I mean 'apparent movement', yes I know that it is the Earth what moves  ).

[PO_MAK_249RIP]

Hi Slibeni have you got anywhere with the 'lines' visible on the terrain yet? Guessing a map tile issue?

[WxTech]

I wonder if the grid-like lines on the terrain/water might be caused by the way map_c is handled? The tga is broken into blocks and put together so as to ignore any non-water blocks. Then the game 'breaks apart' this shore/river-only image into the correct placement, with the formerly left out land-only and water-only blocks filling in automatically.

Furthermore, the screenies so far suggest that the new render maps the terrain onto a spherical planet. Is that so? If so, mapping a purely Cartesian, planar map onto a sphere without correcting for distortion could lead to such artifacts as gaps between sections.

As to the handling of the cirrus cloud layer. In my older game at least, the cirrus is mapped to a shallow dome! Done so because of the artificially near horizon imposed by the limited terrain draw distance. This dome is centered upon the player, highest overhead and gradually sloping downward into the distance in all directions. With this mod the cirrus layer would have to be made flat, or curving appropriately to follow the Earth's curvature (especially if the game world is indeed mapped to a sphere.)

[sowercrowd]

Hello Slibenli,
your modwork is absolutely awesome, however i got a ctd at 100% mission loaded using latest build 108864309.
End Il2ge.log file shows:

Error occurred on Friday, November 2, 2018 at 10:49:09.

Registers:
eax=00000000 ebx=774d17d9 ecx=00000000 edx=00000000 esi=00000000 edi=00000000
eip=77dd1d2d esp=0018f188 ebp=0018f1f8 iopl=0         nv up ei pl zr na po nc
cs=0023  ss=002b  ds=002b  es=002b  fs=0053  gs=002b             efl=00200246

AddrPC   Params
77DD1D2D 0000012C 00000B38 FFFFFFFF  C:\Windows\SysWOW64\ntdll.dll!NtSignalAndWaitForSingleObject
5BA217D6 5BB646C0 0018F27E 00000001  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!abort  [/builds/vrresto/il2ge/common/exception_handler.cpp @ 221]
5BA21B15 5BB65A80 5BB65918 00000166  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!_assert  [/builds/vrresto/il2ge/common/exception_handler.cpp @ 340]
5BA25E5C 0018F2EC 00A1DBE0 6D44038B  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/wgl_wrapper.cpp @ 358]
5BA2C48B 0018F35C 0018F3DC 00000000  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/gl_wrapper_main.cpp @ 765]
5BA3190B 0018F3EC 5BA25784 0018F3E4  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/arbprogram.cpp @ 653]
5BA3224E 00008620 00000012 0018F468  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!wrap_BindProgramARB  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/arbprogram.cpp @ 706]
1009BDD7 00000000 00000000 00000000  E:\IL-2 _4.12.m_VPmedia\il2_corep4.dll!Java_com_maddox_il2_engine_CfgFlagsEngine_Set

[slibenli]

I remember there was some issue with the sun light direction - I'll have to look into that again.
The lines on the map aren't fixed yet.
I'm taking a break currently, so it might be some time until I get to do this.

I would like to use the issue tracker at gitlab (https://gitlab.com/vrresto/il2ge/issues) to keep track of everything.
Maybe somebody could help me out creating tickets for the open issues?
You would have to create an  account at gitlab.com - then I could give you access to the issue tracker.

[slibenli]

Hello Slibenli,
your modwork is absolutely awesome, however i got a ctd at 100% mission loaded using latest build 108864309.
End Il2ge.log file shows:

Error occurred on Friday, November 2, 2018 at 10:49:09.

Registers:
eax=00000000 ebx=774d17d9 ecx=00000000 edx=00000000 esi=00000000 edi=00000000
eip=77dd1d2d esp=0018f188 ebp=0018f1f8 iopl=0         nv up ei pl zr na po nc
cs=0023  ss=002b  ds=002b  es=002b  fs=0053  gs=002b             efl=00200246

AddrPC   Params
77DD1D2D 0000012C 00000B38 FFFFFFFF  C:\Windows\SysWOW64\ntdll.dll!NtSignalAndWaitForSingleObject
5BA217D6 5BB646C0 0018F27E 00000001  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!abort  [/builds/vrresto/il2ge/common/exception_handler.cpp @ 221]
5BA21B15 5BB65A80 5BB65918 00000166  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!_assert  [/builds/vrresto/il2ge/common/exception_handler.cpp @ 340]
5BA25E5C 0018F2EC 00A1DBE0 6D44038B  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/wgl_wrapper.cpp @ 358]
5BA2C48B 0018F35C 0018F3DC 00000000  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/gl_wrapper_main.cpp @ 765]
5BA3190B 0018F3EC 5BA25784 0018F3E4  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!getContext  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/arbprogram.cpp @ 653]
5BA3224E 00008620 00000012 0018F468  E:\IL-2 _4.12.m_VPmedia\il2ge.dll!wrap_BindProgramARB  [/builds/vrresto/il2ge/core_wrapper/gl_wrapper/arbprogram.cpp @ 706]
1009BDD7 00000000 00000000 00000000  E:\IL-2 _4.12.m_VPmedia\il2_corep4.dll!Java_com_maddox_il2_engine_CfgFlagsEngine_Set

I can't tell from the log for sure, since it's not complete, but to you have set water to 0 in conf.ini?

[slibenli]

I wonder if the grid-like lines on the terrain/water might be caused by the way map_c is handled? The tga is broken into blocks and put together so as to ignore any non-water blocks. Then the game 'breaks apart' this shore/river-only image into the correct placement, with the formerly left out land-only and water-only blocks filling in automatically.

Yes, I suspect it has to do with map_c.

Furthermore, the screenies so far suggest that the new render maps the terrain onto a spherical planet. Is that so? If so, mapping a purely Cartesian, planar map onto a sphere without correcting for distortion could lead to such artifacts as gaps between sections.

The terrain is mapped to a sphere. The further away from the camera the stronger the distortion is.
But it's still small enough to not be noticed.
Since every vertex of the terrain is individually mapped to the sphere there should be no gaps.

As to the handling of the cirrus cloud layer. In my older game at least, the cirrus is mapped to a shallow dome! Done so because of the artificially near horizon imposed by the limited terrain draw distance. This dome is centered upon the player, highest overhead and gradually sloping downward into the distance in all directions. With this mod the cirrus layer would have to be made flat, or curving appropriately to follow the Earth's curvature (especially if the game world is indeed mapped to a sphere.)

Yep, is should be a dome that follows the Earth's curvature.
I guess there will be tiling effects because this dome needs to be much bigger.

[WxTech]

It would be worthwhile to look into the way flat sections are fitted to a sphere. If a flat plane is warped to the spherical surface the linear segments between the corner vertices can diverge, producing gaps. If the flat sections are kept flat, so that the global surface is polygonal instead of perfectly spherical, then such gaps cannot occur.

This is somewhat analogous to the demand that the game screen be drawn on the gnomonic projection. Such a projection has straight lines everywhere drawn as straight, thus ensuring no gaps at the joining sides of polygons. No other projection can do this, because anywhere off axis (screen center) straight lines would be curved, with a failure to correctly fit polygonal sides; gaps or overlap would result.

To further impress the characteristics of the gnomonic projection, as a means to divine what might be causing the gaps we see thus far in this graphics extender...

In the real world, the straight line between any two points seen by the observer is a segment of a great circle. That's because the viewpoint and the two endpoints define a plane, and the viewpoint is always the center of the environment. This plane is thus a section of the larger plane that bisects a sphere centered on the observer.

The only projection which retains great circles as straight lines no matter where crossing the FoV is the gnomonic. As an indicator of its real world usefulness, note that this is the projection used for plotting the paths of meteors on the sky (no matter where they appear, the brief flight of a shower meteor is essentially straight, and thus represents a great circle segment.) Any other projection preserves the straightness of a great circle when passing through the center of projection (view center) only; everywhere else a straight line becomes curved.