[Helixhorned]

...since it's easier than searching for them in a thousand places. I'll try to take a look at them once in a while (which could be anything from immediately to a few months )

Some issues I've seen around here and at the SVN tracker for which I don't know a (good) solution:

- No fullscreen in linux: see here (it's rather suboptimal)

- No lights ?

[Helixhorned]

I also remember someone posting who couldn't pan walls. I can't reproduce this. Do you (the poster) mean panning with KP5 and KP8/2? If so, this caused by the wiring of the keyboard and can't be fixed (There's a workaround, see mapster32.cfg at the bottom).

[Plagman]

I think TerminX re-added the necessary code for SE lights to display in mapster32 now, so this shouldn't be an issue with recent builds.

[XThX2]

Helixhorned, on Oct 1 2009, 02:00 PM, said:

I also remember someone posting who couldn't pan walls. I can't reproduce this. Do you (the poster) mean panning with KP5 and KP8/2? If so, this caused by the wiring of the keyboard and can't be fixed (There's a workaround, see mapster32.cfg at the bottom).

I have the same issue with KP5. I have to use the annoying slow panning which is time consuming.

[NF64]

I have a minor issue. When you are in mapster and the new feature is (T)est level but it doesnt seem to work? is it just me or is this common...

[Stabs]

is there a max point lights per level because now when i put one down it prevents weapons from creating light when they are fired

[TerminX]

PR_MAXLIGHTS is 256... are you anywhere near that?

[Mark]

I spent about 20 minutes searching previous posts and did not find the definitive answer to the lack of wall panning and what to do to fix it. Pressing right shift and the Keypad numbers only acts as though the shift key is not being pressed. The closest answer I could find was someone saying "look at the end of the mapster config file". I did and don't know what to make of it.

[Stabs]

yeh iam past 256, well, that crept up on me, i think its also because of the fire aswell.

can it be increased or should i start scaling stuff back?

[Gambini]

For first time i can´t complain about anything with Mapster32. I´m getting no crashes and the windowed mode is now very usefull with the increased zoom.


oh! wait a minute! there is ONE thing: I always use ¨map shade preview enabled¨ because it transmits better the final product´s looking when mapping but there is one thing that doesn´t work with it. sprites take the sector´s shade and palette except in cases where the palette is not 0, 1, 2, 4, 7 or 8. That means that if i have a sector´s palette of 17 sprites look different from the ingame result.

[Mark]

I used the wrong terminology in my previous posting. Instead of wall panning not working I meant wall alignment is not working. Holding down the right shift key along with a keypad number no longer works as it should. Mapster is not recognizing the shift key, so walls are panning instead of aligning. I checked just to make sure, and the shift key does work for other things so its not a faulty keyboard. Anyone else having this problem?

EDIT: I guess it is something on my end. I found an older version of eduke on my computer and it is happening there also. I'll have to find out why my keyboard mapping is not 100 percent compatible. This is a new computer with a multimedia keyboard with programmable keys but I haven't run anything yet to remap the keys. Plus the shift key is doing what it should on everything else besides Mapster. Odd.

[TerminX]

DanM, on Oct 4 2009, 06:53 PM, said:

yeh iam past 256, well, that crept up on me, i think its also because of the fire aswell.

can it be increased or should i start scaling stuff back?

It'll be increased significantly soon. We're thinking 1024.

[Stabs]

woooo woop woop woop woop

thanks tx

[Mark]

I solved my keyboard problem. For some reason on this new computer and keyboard I have to turn off the NumLock button in order for things to work right in Mapster. On my old computer it didn't make a difference if the Numlock key was on or off. So if anyone else out here is running a Microsoft Sidewinder X-6 keyboard thats your fix.

[Forge]

Nfelli64, on Oct 4 2009, 05:16 PM, said:

I have a minor issue. When you are in mapster and the new feature is (T)est level but it doesnt seem to work? is it just me or is this common...

It only works for me if mapster is in windowed mode. And after eduke has exited, I still need to hit the escape button to bring mapster back from some white screen.

[Mark]

I'm pretty sure its version 1524 of eduke I'm using. In mapster, custom textures that are not part of the HRP are showing up in a very low resolution but look as they should in the game. In most cases they are so blocky I can't even align walls and floors properly.

EDIT: I have had so many crashing and texturing problems for about the last 2 weeks in both Mapster and Eduke I'm about to give up for now. I bought a new computer at this same time so its hard to figure out what is hardware related glitches and what are newest versions of the software glitches. As a last resort I did clean installs in different folders with different features turned on or off. After experiencing all these various glitches I transferred those folders to my old computer to try them out. No problems of any kind on the old XP computer except the outdated nvidia card does not display the polymer lighting.

So it seems all of my crashes,lockups and many texture glitches have something to do with my new computer's hardware or it has something to do with the Vista operating system. AAARRRGGG!!!! If I wasn't already going bald I would be tearing my hair out right about now.

[Stabs]

are u adding new tiles?

should just try making them pal40 variants of similar scaled textures

ive found a bug, masked walls will only accept sizing and a pal value on 1 side

[TerminX]

Looks like you have an ATI card and you're running into their problem/"optimization" where they sample from lower resolution mipmaps than they're supposed to. This only happens with newer versions of the ATI drivers, but I think "newer" is anything from the past year at this point. I think the only fix at the moment is r_texturemode 0 in the console. A good question to ask is why is the ATI driver sampling lower resolution versions of the mipmaps when higher resolution versions are available, why do previous versions work and if its an EDuke32 problem, why do other vendors drivers work fine?

[Mark]

Its a new computer and I am running a Nvidia 9800 gtx with the latest drivers, which are not always the best. There are no earlier ones on the computer to roll back to.Everything else runs flawlessly so far. Just Eduke and Mapster issues. I hate to go backwards in drivers but I just might have to try.

EDIT. I have been able to confirm by repeated tests that changing rendmode=3 to rendmode=4 in the Mapster32.cfg file is when the skybox and lo-res texture problems appear. So for now I just have to forget about viewing polymer lighting in Mapster. And by the way, the incorrect display of the skybox in mapster under polymer is the same incorrect display that shows up in the game. So for some unknown reason polymer is enabling that skybox glitch in both programs on my computer.

[Helixhorned]

Nfelli64, on Oct 5 2009, 12:16 AM, said:

I have a minor issue. When you are in mapster and the new feature is (T)est level but it doesnt seem to work? is it just me or is this common...

This is probably caused by mapster32 not finding the game executable. I always run mapster from the same directory as eduke32 and everything works... dunno for other configurations.

Marked, on Oct 5 2009, 03:13 AM, said:

(...) The closest answer I could find was someone saying "look at the end of the mapster config file". I did and don't know what to make of it.

I just meant that you can remap some key (like KP0) to act as KP5, and the problem will go away because it was blocked at the keyboard (i.e. hardware) level.

Gambini, on Oct 5 2009, 04:26 AM, said:

oh! wait a minute! there is ONE thing: I always use ¨map shade preview enabled¨ because it transmits better the final product´s looking when mapping but there is one thing that doesn´t work with it. sprites take the sector´s shade and palette except in cases where the palette is not 0, 1, 2, 4, 7 or 8. That means that if i have a sector´s palette of 17 sprites look different from the ingame result.

You probably mean when the sky is parallaxed? It's fixed now.

DanM, on Oct 11 2009, 12:15 AM, said:

ive found a bug, masked walls will only accept sizing and a pal value on 1 side

This appears to be due to a known Polymer bug (incorrect normal computation).

[XThX2]

About the one DanM said, it jumps to the wall behind it if you try to set things besides those. For example, I was trying to darken my forcefield, but it was changing the shade of the wall behind it.

[Stabs]

alot of my rendermode 4 crashes seem to be related to holding shift to move the textures, its a crash that has happened alot to me latley and thats what iam doing when it crashes.

[Helixhorned]

This is very odd considering those two things are completely unrelated (editor vs. Polymer code). The most helpful thing you could do is to try to reproduce the crash in a debugger. That is, in the Windows console from the eduke32 directory, "make veryclean", "RELEASE=0 make mapster32.exe", then "gdb ./mapster32.exe" (and "r" in the gdb prompt). When the crash happens, go to the console you started gdb from and enter "bt". This will print valuable information which could be used to track down the bug.
(Note: Of course, these instructions assume that you have the GNU toolchain installed.)

[Stabs]

is that all the compiling stuff for eduke32?

i have mingw and GCC stuff, so do i run and do this all from the dir i keep the source files in?

[Stabs]

i got to the gdb part and downloaded 7.0 , where does this go in relation to mingw?

[Helixhorned]

Well, executables compiled with GCC (which is part of MinGW) should be debuggable using a native GDB. I got mine with the MinGW package, however (this was years ago). Have you checked for a GDB in MinGW/bin? You should be able to run mapster like usual (from the command prompt), just with "gdb" prepended. For this, the gdb executable should be in your %PATH% (It probably already is. If not: google for "path windows").

[Forge]

Using snapshot 20091017.
Mapping in 32 bit polymost.
After about fifteen to twenty minutes mapster starts randomly freezing and unfreezing. In 3d mode mainly when shading or texturing. In 2d mode mainly when adding or moving vertices. The freezes can last anywhere from a couple of seconds to half a minute.

(T) test map still doesn't work unless mapster is in windowed mode. For some reason the program is having difficulty minimizing mapster and bringing eduke topmost. Even with mapster windowed, after eduke is exited, mapster wont reinitialize and become topmost.

Weird sector number report. I'm at around 1074, but it's reporting it as something like 107423.

[Sobek]

Forge, on Oct 25 2009, 02:25 PM, said:

Using snapshot 20091017.
Mapping in 32 bit polymost.
After about fifteen to twenty minutes mapster starts randomly freezing and unfreezing. In 3d mode mainly when shading or texturing. In 2d mode mainly when adding or moving vertices. The freezes can last anywhere from a couple of seconds to half a minute.

(T) test map still doesn't work unless mapster is in windowed mode. For some reason the program is having difficulty minimizing mapster and bringing eduke topmost. Even with mapster windowed, after eduke is exited, mapster wont reinitialize and become topmost.

Weird sector number report. I'm at around 1074, but it's reporting it as something like 107423.

I am experiencing something similar, only it occurs much much sooner.

After about 5 minutes of using Mapster32, it just starts to 'crawl'. Every 10 seconds or so, it will bog down from ~80fps to, say, 20 for a moment, then back to normal. It keeps doing this until it either crashes or I quit. Mostly though, it seems to just get slower and slower the more you use it... After about 4 or 5 switches between 2d/3d modes, it starts to take absolutely AGES to make the changeover. Sometimes, I can end up waiting over 30 seconds just to get back into 3d mode.

It doesn't appear to be a memory leak (at least nothing obvious), as it never really uses up more than about 100mb of ram. It makes no difference if Polymer or Polymost is used either, but Polymost does tend to last a little longer.

Also, fullscreen 3d is still not possible for me in Mapster32. It fires up in fullscreen 2d mode just fine, but crashes out with no errors in the log when I try and switch to 3d mode. If I start Mapster32 and it loads directly into 3d mode on backup.map or whatnot that's fine, and I can switch back to 2d mode, but can't go back to 3d mode again after that.

I'm basically mapping in 2 or 3 minute blocks. If I use it for much longer than that, it seems to bring my whole system to its knees, even though it's barely using any resources. It can take upwards of a whole minute just to quit and return me to my desktop, even which after that it takes absolutely ages to do ANYTHING with my PC (like, open my computer). I have to restart to restore overall performance to what it should be.

[Forge]

Not that I have the latest, state of the art computer, but it is a 2.0 g dual with 4 g ram and mapster is doing something to the system resources. The more I add to the map, the worse it gets. I'm at about the same point as Sobek. Around five minutes into mapping the system resources for the program jump from around 3 - 6% up to 50 - 60 % then it fluctuates up and down causing mapster to "freeze" and "unfreeze".

I'm not having any problems switching back and forth from 2d to 3d mode in either full screen or windowed yet. The only real crashes I've experienced was trying to launch "(T) Test" from mapster while in full screen mode which froze up my computer.

[Sobek]

Oh, yours actually eats up lots of RAM and such? Mine never really uses up too much ram or cpu power, but it sure as heck acts as if it were.