The Great Apple
Big Cheese
Hmmm - except that the patcher might check for .dll consistancy when it patches.Harkonnen said:
Memory-fix by Harkonnen is out!
- Thread starter Harkonnen
- Start date
warpstorm
Yumbo? Yumbo!
Are you going to release source code? Without the ability to see what you changed (and recompile it myself) I will not download and execute dlls.
phalzyr
Warlord
Can't wait to try it MP, SP I have no issues but last night my secodary began to have moments of slow response...
Hark: BTW it looks like there is going to be a meeting that I have to go to at my usual leave time so I'll be here a bit later than norm to put the file up on my mirror though you could always do fileplanet or such with the final release. I'll give it a go in pirates and see if it effects anything, is there anything special I have to do? like rename files or such... prob not...
Hark: BTW it looks like there is going to be a meeting that I have to go to at my usual leave time so I'll be here a bit later than norm to put the file up on my mirror though you could always do fileplanet or such with the final release. I'll give it a go in pirates and see if it effects anything, is there anything special I have to do? like rename files or such... prob not...
CyberChrist
You caught my attention
Well done! 
Reduction of slow down is significant. However it seems that the leaderheads get corrupted after a while (with insane_mode=1 in full screen), but that may be because I loaded a few saves without restarting CIV.
Reduction of slow down is significant. However it seems that the leaderheads get corrupted after a while (with insane_mode=1 in full screen), but that may be because I loaded a few saves without restarting CIV.
sabrewolf
Demidemigod
warpstorm said:
paranoid... like me
but honestly: you should submit your stuff to firaxis. i'm sure they would appreciate you fixing one of the memory leaks and would surely add you in the credits. this would also reduce the concerns of us paranoid people
warpstorm
I understand your worries, but I've spent around a month with this patch, all my free time and even more. I mean I will make an offer to Firaxis as soon as I see that this patch is really helpful. Currently I see it is.
If Firaxis refuses to deal with me, I will publicly release it, probably even with an article on codeproject or something like that. I have released binaries, and I will release the source as soon as making it public won't hurt my interests in my deal with Firaxis if it's ever gonna happen. I wouldn't think about any kind of "deal" with Firaxis if it would be very easy, or even very hard, but easy in the last instance - this last version.
It is very complex code with very complex approach and a lot of rewriting and profiling. Heck, it contains dynamic assembly code allocations near template specialization next line... so it's gonna cost.
I am a bit different man, as compared to those who would spread trojans, viruses and alike through this most-wanted patch (also I doubt that kind of people would pursue their goals this way). Yes, I know ASM pretty well, but that's becauses I'm ~12 years with computer in my 22. Also I have attended ACM programming contest world finals twice. I don't want to boast, just people at these contests are sooner "awkward eggheads" rather than so-called "russian mafia" spreading s**t on the web.
P.S: If you look into the very begining of 'memory leak' thread, you will see that I HAVE released source when I tried some easy effort of reporting less physical RAM. After all it's 174Kb of source code in size, and the code isn't yet well-organized. Wanna read it?
I understand your worries, but I've spent around a month with this patch, all my free time and even more. I mean I will make an offer to Firaxis as soon as I see that this patch is really helpful. Currently I see it is.
If Firaxis refuses to deal with me, I will publicly release it, probably even with an article on codeproject or something like that. I have released binaries, and I will release the source as soon as making it public won't hurt my interests in my deal with Firaxis if it's ever gonna happen. I wouldn't think about any kind of "deal" with Firaxis if it would be very easy, or even very hard, but easy in the last instance - this last version.
It is very complex code with very complex approach and a lot of rewriting and profiling. Heck, it contains dynamic assembly code allocations near template specialization next line... so it's gonna cost.
I am a bit different man, as compared to those who would spread trojans, viruses and alike through this most-wanted patch (also I doubt that kind of people would pursue their goals this way). Yes, I know ASM pretty well, but that's becauses I'm ~12 years with computer in my 22. Also I have attended ACM programming contest world finals twice. I don't want to boast, just people at these contests are sooner "awkward eggheads" rather than so-called "russian mafia" spreading s**t on the web.
P.S: If you look into the very begining of 'memory leak' thread, you will see that I HAVE released source when I tried some easy effort of reporting less physical RAM. After all it's 174Kb of source code in size, and the code isn't yet well-organized. Wanna read it?
phalzyr
Warlord
Harkonnen said:
Couple hours? Cool I'll still be here for about 3 more. Any more than that though...
CyberChrist and all
If 'insane_mode=1' starts well, plays well, but some time later gives visual bugs WITHOUT alt-tabbing, then decreasing 'vid_mem_usage' should help leaving 'insane_mode=1'. Try ~30-40 Mb if you experience visual bugs even without alt-tabbing in insane mode.
The effect of vid_mem_usage should be kind of such:
slow-slow-avg-avg-best-avg-slow-slow-bugs-crash
It's up to you to find that golden "best"... It depends on video memory and AGP apperture size (the latter is set in BIOS to 32, 64, 128 or 256Mb, depending on your motherboard capabilities). But if that would be everything it depends on, I would set it automatically. The nasty thing is that it also depends on graphics quality you select, and that is hard to track from my code.
If 'insane_mode=1' starts well, plays well, but some time later gives visual bugs WITHOUT alt-tabbing, then decreasing 'vid_mem_usage' should help leaving 'insane_mode=1'. Try ~30-40 Mb if you experience visual bugs even without alt-tabbing in insane mode.
The effect of vid_mem_usage should be kind of such:
slow-slow-avg-avg-best-avg-slow-slow-bugs-crash
It's up to you to find that golden "best"... It depends on video memory and AGP apperture size (the latter is set in BIOS to 32, 64, 128 or 256Mb, depending on your motherboard capabilities). But if that would be everything it depends on, I would set it automatically. The nasty thing is that it also depends on graphics quality you select, and that is hard to track from my code.
gunnergoz
Cat Herder
Great work, Harkonnen, thank you!
warpstorm
Yumbo? Yumbo!
Harkonnen said:Wanna read it?
Sure, if the comments are in English.
NateDawgNY
Just Chillin'
Outstanding! I will definitely be trying this out when I get home.
BlueLikeYou
Chieftain
- Joined
- Nov 2, 2005
- Messages
- 53
I will give this a shot tonight after work and see how it goes.
Joebasalt13
Warlord
Need Help From Harkonnen or anyone who has used his patch! When I try to enter the Configure settings file it says access denied how do I get around this?
phalzyr
Warlord
What program are you using to try to open the ini file? Is civ 4 running at same time? Oh and where did you download them to, it sounds like a NTFS file security issue could be the cause also...
Edit: One more thing... did you try right clicking the ini and chose "open with" and select notepad?
OH and Hark: The meeting got canceled so I'll be here another hour rather than two, otherwise I'll miror first thing tomorrow.
Edit: One more thing... did you try right clicking the ini and chose "open with" and select notepad?
OH and Hark: The meeting got canceled so I'll be here another hour rather than two, otherwise I'll miror first thing tomorrow.
I'm afraid your patch crashes my Civ IV during the "Init Engine" stage. Here's what I get in the event log:
Faulting application civilization4.exe, version 1.0.9.1722, faulting module patchbyharkonnen.dll, version 0.0.0.0, fault address 0x000017f6.
As soon as I set patch_on to 0 it starts working again, so it's probably not the python DLL. Using the patch with insane_mode off doesn't fix it, either.
I'm using a ATI Radeon 9800 Pro with 128 MB and the 5.11 Catalysts, so I guess it's not the graphics hardware...
Is there anything I can supply you with, like a Dr. Watson log or the memory dump or something?
(I'd break out the debugger, but debugging without source is a bit beyond what I'd want to do...)
Faulting application civilization4.exe, version 1.0.9.1722, faulting module patchbyharkonnen.dll, version 0.0.0.0, fault address 0x000017f6.
As soon as I set patch_on to 0 it starts working again, so it's probably not the python DLL. Using the patch with insane_mode off doesn't fix it, either.
I'm using a ATI Radeon 9800 Pro with 128 MB and the 5.11 Catalysts, so I guess it's not the graphics hardware...
Is there anything I can supply you with, like a Dr. Watson log or the memory dump or something?
(I'd break out the debugger, but debugging without source is a bit beyond what I'd want to do...)
Similar threads
