Jorpho wrote:Oh, you can run whatever you want and I can hardly tell you otherwise, but when you come along with questions about such combinations, you're pretty much inviting them to become someone else's business. Also, language like "I consider it to be the only 'right' way" kind of comes across as dismissing any other conceivable alternative as the "wrong" way.
I suppose I see where you are coming from there, but on the other hand I don't make a habit of asking questions without doing a good deal of experimenting myself. I've been using rloew's RAM patch for 6 years now and have never had one single problem with it. Granted I have not used it in this exact hardware combination, but I know from experience that it is stable and certainly not the first thing to suspect when a problem arises. Maybe other methods are not "wrong" per se, but, IMO, other methods are "workarounds" where rloew's patch is a "bugfix."
At any rate, I will continue to experiment both with my original configuration and with only 512MB of RAM and no RAM patch for a control.
Jorpho wrote:Jorpho wrote:But how can you say it's not a bug specifically triggered by the combination of the X850 and the Limitation Patch?
Because the exact same problem still persists even if I uninstall the RAM Limitation Patch and reduce the RAM to 512MB.
Jorpho wrote:Fair enough; that is a much better starting point. I'd be curious though if less than 512 MB makes a difference.
I seriously doubt that less than 512MB would change it, but if it comes down to it I will attempt that configuration to rule it out.
Also though, keep in mind that the older WarCraft I is working properly, so that seems to indicate it is a very specific difference or "bug" of some kind, rather than being a more general "too much memory" or other "too much/too fast hardware" type issue.
Jorpho wrote:Alternatively: isn't this a DOS program? What happens if you restart in MS-DOS mode and run the game? If it's using DOS4GW, then DOS32a is also a possibility. UniVBE might also be worth looking into.
chinny22 wrote:I agree with dropping back real dos, at least for testing. System as "new" as yours may be bit of a pain to get going with sound but just a generic Win98 bootdisk will be enough to get the game running with no sound which is enough for testing.
On a side note I also prefer the DOS version of the game but moving around the map is WAY too fast on my P3 1Ghz so ended up using battle net version anyway which doesn't have the same problem
Now we're getting somewhere. Running the game in pure DOS printed an error to the screen that never gets displayed in Windows.
DOS4GW Error wrote:DOS/4GW Professional error (2001): exception 0Dh (general protection fault) at 3E58:000024D7
TSF32: prev_tsf32 6B14
SS 180 […]
Show full quote
DOS/4GW Professional error (2001): exception 0Dh (general protection fault) at 3E58:000024D7
TSF32: prev_tsf32 6B14
SS 180 DS 3E68 ES 198 FS 0 GS 0
EAX 0 EBX 1 ECX 0 EDX 1
ESI 40 EDI 4010 EBP 378 ESP 33C
CS:IP 3E58:000024D7 ID 0D COD 0 FLG 10202
CS= 3E58, USE16, byte granular, limit CA8F, base 115820, acc 9B
SS= 180, USE16, byte granular, limit 1FFF, base 150F80, acc 93
DS= 3E68, USE16, byte granular, limit 886F, base 1222B0, acc 93
ES= 198, USE16, byte granular, limit 3F, base 154000, acc 93
FS= 0, USE16, byte granular, limit 0, base 16, acc 0
GS= 0, USE16, byte granular, limit 0, base 16, acc 0
CR0: PG:0 ET:1 TS:0 EM:0 MP:0 PE:1 CR2: 0 CR3: 0
Yeah, the DOS version does scroll ridiculously fast on another P4 3GHz machine of mine. I basically adapted to clicking directly wherever I wanted to look on the minimap rather than attempting to scroll... 🤣
Sammy wrote:I have the same card and tested it under win98 and dos7.1: […]
Show full quote
I have the same card and tested it under win98 and dos7.1:
Under win98: see dos4/gw then back to desktop.
under plain dos7.1: dos4/gw crashes with general protection fault. 0Dh
This machine has 3gb ram (reduced to 512 via bs_ram9x in autoexec.bat)
CPU= AthlonXP 3200 (reduded via bios to 650 Mhz)
Next machine i tryed:
Athlon Thunderbird 1300 Mhz
Geforce Ti 4200
1512 MB ram
Dos7.1 and Win98
the game runs fine in win98.
in dos it hangs after war2.exe (enter) because i have not mscdex installed.
But there is no Protection fault.
Maybe it is really GFX Card related?
Sammy wrote:Update: […]
Show full quote
Update:
-reduced the ram to 512 mb (removed other memory sticks)
- changed ATI X850 XT PE with Geforce256DDR
Still getting general protection fault
maybe the geforce1 card makes also problems in Warcarft II ?
or it is motherboard related (asus a7n8X-X)
Sammy wrote:update 2: now tested the geforce4ti (which works in other pc) in Dos and there it still crashed. […]
Show full quote
update 2: now tested the geforce4ti (which works in other pc) in Dos and there it still crashed.
but in windows98 it runs:
Geforce256DDR with 512MB in win98 = run
Geforce4TI4200 with 512MB in win98 = run
Geforce4TI4200 with 2048 mb (reduced to 512 via bs_ram9x.exe) in win98 = run
X850XT PE with 2048 mb (reduced to 512 via bs_ram9x.exe) in Win98 = crash
x850XT PE with 512 mb in Win98 = crash
it seems the X850xt pe does not support the resolution of the blizzard intro video.
I see you came to the same conclusions as I did. As I mentioned above, WarCraft I works though. Do you have it to test on your X850 to see if you can reproduce my results?
Joey_sw wrote:iirc there was dos program that could gives list what resolutions that supported by graphic card.
is there any certain resolution that X850pt doesn't support but it was supported by other cards?
or the warcraft 2 using non-bios method to set the video mode?
Sammy wrote:i don't know... but the game crash at that moment when the Video should be played.
I think the best for the Threadstarter is to use another Graphics Card, or to try Dosbox to play Warcraft II
Jorpho wrote:That's sort of why I was thinking UniVBE might be a good idea, but then again I don't think it supports a card as new as the X850.
Reverting to a 9800 XT is always an option, but I plan to keep trying to sort this out for now. I prefer not to run DOSBox on a 9x system that shouldn't need to do so, but I guess it's a viable "last option."
Maybe the error I posted above will shed some light on the problem, if anyone knows how to interpret it...