Purple and pink screen of death.

Something happened to my first attempt to post this and it vanished. But, I got it again.

The first time it happened it was as if my display unsynced - it had narrow horizontal rows of groups of mostly purple and pink
lines of perhaps 2-3 pel vertially by about 160 pel horizontally. There were 8 distinct vertical rows. There were also a lot of
yellow lines and some orange lines.

The second time the lines were almost entirely blue with white speckles.

In both cases the system froze. Only the brute force power off had any affect (a steady push of the power button for 5-10 seconds).

This looks like what would happen if the program got a bad pointer and started storing data in the memory reserved for the display
adapter card.

The first time the debug file had an error message in it that SmartExceptioln was unable to open the file into which it stores the
exception report. The second time the debug file had no indication of errors whatsoever, not even any indications of memory management
errors. However, there was this message:
--------
Debug Message: No AutoSave files in current directory.
-----------

Just in case something can be done with the autosave files I am sending them and the debug file to Cari.

The first time it happened I was left clicking on the main screen to clear either the planet list or ship list (no longer remember
which). This time I was just looking at the screen contemplating my next action (for about 3-5 seconds).

Cheers,
Jack

596 views 28 replies
Reply #1 Top
I tried starting from the autosave and it did this again before I could finish the current turn (this time giving me a pink and purple out of sync display of death).
I noticed this time that the "open space" music continued for a couple of seconds after the screen went berzerk, then quit. The autosave files may be useable to
recreate the problem.

I have my display set to 1280*1024*fullcolor, if it matters. I have more than ample on board and swapper memory.

Cheers,
Jack
Reply #2 Top
Can you email me your saved game?

[email protected]
Reply #3 Top
Brad,
Cari should already have it in her reader. If not, I can send it again.

Cheers,
Jack
Reply #4 Top
I have another saved game with the same problem. This time I counted how many planets I had colonized and how many
ships I had (78 planets, 25+ ships). This does not count the planets and ships my enemies had which could have pushed
the total planets to well over 100, but the number of ships the enemies had may have been near zero.

This is, of course, on a gigantic galaxy. I have copies of the saved game saved on the turn just prior to the wipe-out.

Do you want a copy of this saved game?

Cheers,
Jack

Reply #5 Top
After installing todays build (are we calling it "Alpha 5"?) I have restarted a saved game that was giving me this problem and have
been playing it now for some time without it crashing.

With Alpha 4 I was not getting the CNN for the completion of planet projects. Now I am. Could there be an associative affect?

Cheers,
Jack
Reply #6 Top
How about calling it 0.2.005 since that's what it says in the Drengin Game Manager. Oh wait, in the windows task manager, it's called v.0.01.

Maybe we could call it the LGV for the latest, greatest version.

;)

Thanks!
Reply #7 Top
When all else fails, I just default to the build date on the Galciv.exe, which now is 4/2/02. But I would much rather call it
what everyone else expects it to be called.

Cheers,
Jack
Reply #8 Top
This failed again on the Alpha 5 build. I got a blue patterned screen with lots of white dots this time. I checked
first to see if it was a solid failure and the game restarted fine. Hence, I lost all of the file data that might
have been useful. Sorry.

Jack
Reply #9 Top
There is something very strange going on here. When we find out what is causing this we may be
truly amased. The screen colors I get seem to be dependent on the last research project I
selected or defaulted to for the few I have tried. My current research window gives me a choice
of these projects: Cloaking, Battleship 2, Star democracy, Organic piecing, 2D phasing, Turbo
phasers, Nano-metal composition, and Brain Mapping.

With Cloaking I get the Purple and pink screen of death. With battleship 2 it is a blue pattern
with white dots. I am going to try the others to see what they give.

Cheers,
Jack
Reply #10 Top
Wow. Strange indeed.
Sadly I cant help with input... I dont have that error here :( Hmm I normally, for testing, use small maps. I'll try the biggest map after the weekend.
Reply #11 Top
This may be a bit premature, but I have been doing all of my testing in true color (32 bit). I just changed to high color (16 bit) to test something
else for Cari and where I was able to get the screen of death very quickly on restarting the game, I am now able to go to the nest turn. More testing
on my part is needed, but is there anyone else out there testing with 32bit color?

Cheers,
Jack
Reply #13 Top
This problem only seems to happen when I get into the higher research projects. I may be the only oen getting there
because of my tendency to test on a grander scale (gigantic galaxies for the most part).

Or, it could have something to do with a difference in display adapter card. I am using a NVIDIA GForce2.

In any case, just as soon as I switched back to 32 bit color the saved game I have that gives me these screens of death
failed very predictably and quickly.

Now for the big questions to all.

1). How many of you are testing in 32 bit color? 16 bit color? What display adapter are you using?

2). How many of you have gotten down the research tree as far as I have (see a previous post on this thread - 04/05/2002 at 6:30 PM)?

3). What galaxy sizes have you been using for your testing?

All of these responses to these questions will help to develop a direction of examination in solving this problem.

Thanks in advance for everyones response.

Cheers,
Jack
Reply #14 Top
I spoke too soon. It crashed in 16bit color mode. It just took longer.

I am now trying to figure a way of getting all of the galciv files and SmartException over to an older PC to try it
there. It has an ATI mach64 PCI display addapter and an MWave sound card, so that will let me test the theories about
my hardware being at the root of the problem. So far, however, no luck with the transfer. I may have to hook the other
PC up to the net, load DGM on it and get it that way. Unfortunately, the modem is a V34, not a V90 (28K verses 56K).

If anyone has any advice, I would love to here it, either here or by e-mail.

Cheers,
Jack
Reply #15 Top
Well, I have given up trying to load GalCiv onto my older PC. Some of the files are too big to move across via diskettes, the old PC doesn't have the support to read the CDs I burn on this PC,
and I can't get a solid enough connection to download Galciv from Stardock. I did manage to get the DGM up and running, however.

Any new ideas on how to get something meaningful on these screens of death?

Cheers,
Jack
Reply #16 Top
Jack,
I also test with 32 bit colour and have reached advanced stages of the research. As you already know there seem to be some problems with the tech tree but I have never had this screen of death problem yet.
I use a voodoo 3 graphics card.
Will test this to exhaustion tonight.

Paul.
Reply #17 Top
Jack, your old PC can't read them even if you format them to be read on normal CD drives?
Reply #18 Top
Cari,
I have tried what options I can find and understand in AdaptecCD. None of them result in a CD readable on the older PC. Is
there something I am missing?

Cheers,
Jack
Reply #19 Top
I have downloaded and installed drivers for the nVidea display adaptor (no luch) and the sound card. Believe it or not, I got a
few more turns - then got the purple and pink screen of death. Are one of these smily face thingies beside the entry space I am
entering this post in a face of frustration? My hair seems to be scattered all over my keyboard and display (not to mention the
desk, floor, and chair).

If anyone has any ideas of how to find out what is causing this and how to trap it, please jump in.

Cheers,
Jack
Reply #20 Top
Jack,
I player the entire research tree out 4 times last night (reduced the research costs) and never once got the pink screen of death. If you send me your saved game I can try that? AS for frustration, I'd go with the fifth down on the left.


Paul.
Reply #21 Top
Paul,

Thanks for you offer. The saved game is on the way.

I am not convinced it has anything to do with the tech tree. It may have more to do with something related to
the size of things, such as a gigantic galaxy, the number of planets, the amount of space I have explored, the
screen size or resolution (1280*1024*32bit), etc.

Cheers,
Jack
Reply #22 Top
Well, I've received the saved game and played it four times on my machine. Twice I had a hard freeze (CAD does not bring up task manager) but no interesting color effects -- just the cursor stopped moving around on the screen. Twice I was dumped to desktop and got an exception report -- I include them both below:

First exception report, second game

Stardock SmartException Logging Information 04/09/02 10:17:37
-------------------------------------------------------
Versions:
Microsoft Windows 98
Internet Explorer 6.0.2600.0000

Video Card : Matrox Millennium G200 AGP
Crash Information:
Application Exception code: C0000005 ACCESS_VIOLATION
Crash Addr: 004C4B38 01:000C3B38 C:\PROGRAM FILES\DRENGIN.NET\GALCIV\GALCIV.EXE

Processor Registers (x86):
EAX:11091109
EBX:02F65BC8
ECX:006CB14E
EDX:000000DD
ESI:02FA2EB1
EDI:CC8DDAA2
CS:EIP:0177:004C4B38
SS:ESP:017F:00A6F22C EBP:FFC9A759
DS:017F ES:017F FS:1C17 GS:0000
Flags:00010286


Current Application Call Stack:

Addr Frame Logical addr Module Name
---------------------------------------------
004C4B38 FFC9A759 0001:000C3B38 C:\PROGRAM FILES\DRENGIN.NET\GALCIV\GALCIV.EXE

***
Hey that was short. Here is the second exception from the third game

Stardock SmartException Logging Information 04/09/02 10:19:33
-------------------------------------------------------
Versions:
Microsoft Windows 98
Internet Explorer 6.0.2600.0000

Video Card : Matrox Millennium G200 AGP
Crash Information:
Application Exception code: C0000005 ACCESS_VIOLATION
Crash Addr: 004C4B38 01:000C3B38 C:\PROGRAM FILES\DRENGIN.NET\GALCIV\GALCIV.EXE

Processor Registers (x86):
EAX:00000000
EBX:02F65BC8
ECX:006EE14E
EDX:00000000
ESI:02FA2EB1
EDI:CC83DA20
CS:EIP:0177:004C4B38
SS:ESP:017F:00A6F22C EBP:FFC88F59
DS:017F ES:017F FS:0D1F GS:0000
Flags:00010286


Current Application Call Stack:

Addr Frame Logical addr Module Name
---------------------------------------------
004C4B38 FFC88F59 0001:000C3B38 C:\PROGRAM FILES\DRENGIN.NET\GALCIV\GALCIV.EXE

***

On the first, second, and fourth crash, I had just clicked on the turn button. For the third crash I believe I just loaded the game and it immediately crashed -- no, I had to hunt for the place to close the research screen even though the ship display had covered almost all of the done button. So it crashed when I clicked on the done button.

I'm not sure this will help much, but I thought it was worth a try.

Thanks!
Reply #23 Top
Frank,

Thanks for your try. Hopefully Cari will get something out of the report that will lend a possible explanation to what I was getting.
I have been trying for some time to rationalize some suspicions I have about the source of these crashes. One thing that keeps coming
back to raise doubts about it being a problem with my PC is that I was able to get much further that this before build Alpha 4. I know
that this doesn't really mean anything about the cause, so I have been trying everything I can to try to eliminate possible problems
on my PC. I have many newly updated drivers on now that have had little or no affect on the problem.

The fact that you have Windows 98 and got an exception report might be significant (I have ME).

Thanks,
Jack
Reply #24 Top
Jack,
I have tried your saved game and can play it again and agin without pink screens of death. I have however had some crashes. These are usually for a variety of reasons and never the same one. For example, I had it crash when I changed the autoturn settings. I changed the settings (turned it off), returned to the game and it crashed on returning? I didn't actually do anything! I have however successfully played another few hundred turns a few times now, so the problem does not seem to be with the saved game itself. These tests are also with your copy of the tech files.

I'll send any of the interesting exception reports to Cari this evening along with a description of what caused the crash,

Paul.
Reply #25 Top
Paul,
Thanks,
Jack