C2C - UEM - Ultimate-Earth-Map 100% MOD and SVN update compatible by Pit2015

[Toffer90]

Well; I was inaccurate in my previous post, as our dll is LAA, so I edited it. My point was that the exe is handling RAM allocation, and it is not giving us more than 2.8 GB for whatever reason.

It might have something to do with how BtS is running python natively, maybe 1.2 GB is reserved for python cache.

Also, I'm pretty sure this is correct: https://social.msdn.microsoft.com/F...e-of-large-address-aware-dll?forum=vclanguage
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
Purpose of large address aware dll
Question: My application (EXE) is large address aware and it uses a DLL. This DLL dynamically allocates memory.
...
Interesting point is that non large address aware DLL was also able to dynamically allocate more than 2GB. So, what does large address aware flag signify for a DLL?

Answer: Nothing. Only the flag on the .exe matters.
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
So our dll is just a slave to the "dll within the exe" which is the file with the code that allocates RAM address space for the process, our dll doesn't matter in this regard.

 

[Pit2015]

Really try this, flag the DLL LAA should be not to hard with a LAA patch tool, maybe somehow the used DLL gets loaded into the 4 GB exe range also if its a slave and then not.

Some say dont works other say works, so only a test can tell. Maybe it works only on some old stuff and it is the case here.

Yes, thats the question, where does that 1.2 GB go, at the moment its just thoughts. If its python, maybe you can use it somehow or put more into python.

Otherwise if this test dont give a positive result think about a way to find out where are these missing 1.2 GB exactly go.

 

[Toffer90]

Really try this, flag the DLL LAA

Our dll is compiled with the large address aware flag set.

, as our dll is LAA,

 

[Toffer90]

Yes, thats the question, where does that 1.2 GB go, at the moment its just thoughts. If its python, maybe you can use it somehow or put more into python.

It's a point, I could try caching large amounts in python and see if I can go beyond 3 GB usage that way.
Have an array of arrays, and cache a 1 million big array into the first array every new turn I play should probably do it as a test.

 

[Pit2015]

It's a point, I could try caching large amounts in python and see if I can go beyond 3 GB usage that way.
Have an array of arrays, and cache a 1 million big array into the first array every new turn I play should probably do it as a test.

Beside this i did some RAM testing...

1. I placed 20.000 - 30.000 units on the map, works! Memory usage only 100 MB more. 1.250 MB to 1.350 MB now (After complete game restart). If i move to all regions of the map memory usage is still under 2 GB, as all of the map is loaded with thousands of units outstanding.

2. Closed WB and got this crash: "Failed to allocate video memory..." screenshot RAMTEST1 Line:321 there is a problem but it is fixed if you reload the complete game. Look into there if something can be done there. Increase the outside video memory or something. Tells something about texture, maybe only a bugged texture...

3. After reload i did a game turn, worked! Conclusion 30.000 units only 100 MB more, so 100 civ dll should be no problem.

4. If i restart the game my memory usage is 1.250/1.350 MB also with 30.000 units on map, if i go to every region on world map its 1.900 MB (But the 30.000 units still only take 100 MB more, its the map stuff), so try to fix the unloading of the graphics paging, should clear the memory again if you go to another region, like you restart the game completely, that will save a lot of memory.

Conclusion a long game will be possible now and the CTDs are mostly by game bugs as the memory usage stays far under 2.8 GB.

Interesting if you tab out of the game in fullcreen sometimes the preloaded regions on the minimap disappear, maybe it works to tab out of game and then turn without crashing. Sometimes you need to tab out 3 times before the minimap is not blue anymore everywhere. But ram usage still higher.

Check out and do some testing:

1. Load RAMTEST1 scenario, open WB and close WB, turn = crash maybe.

2. Load RAMTESTSAVE after complete game restart and turn it, should work most times. Watch the combat to ping/view every region. Turn combat animations on. (To be able to start a turn put a production in every placed city)

Looks like someone has to fix the graphics paging to unload the loaded graphics again correctly if you look another region on the map. Game restart does this, in game changing the view to another region its not doing it.

 

[Somebody613]

@Pit2015
I think units affect TURN TIME heavier, which is a much more annoying issue.
MAFs might be less unavoidable now, but 5-minutes-long turns... no, thanks.
On the other hand, it has not DIRECT roots in the number of civs, since I get much slower turns from swarming Barbs alone, which aren't even actual civs to begin with, just units and cities.

 

[bartigol2]

@Pit2015 whats your thought on using Size Matters option on your map?

 

[bartigol2]

@Pit2015 also i see you mention "SVN" - what is it and do i need it to play without issues?

 

[Somebody613]

@bartigol2
Don't necessarily listen to what Pit suggests, lol.
For some weird reason he dislikes several unique features of C2C, so he will tell you to not use them, but that's literally missing a whole lot of FUN.
Definitely use all Combat Mods at the least - these enhance the gameplay like you would never dream of in Vanilla.
Merging, Stealth, Assassination - these are extremely unique features of C2C that make it so hugely distinct from the normative Vanilla Civ4 (or probably ANY Civ, I'm not sure).
And SVN is basically a server-based ongoing updating for the mod, in case you want to be onboard with the very latest updates.
Though I've seen the modders preparing for v.43 soon-ish (no idea when exactly), so it might be less relevant NOW.
While there were certain discernible changes between the previous release and the current SVN, you as a new player wouldn't feel much of it anyways.
Just play what you have - it'll still show you 99% of what C2C is about, and that's what matters.
Enjoy!

 

[Pit2015]

@Pit2015 also i see you mention "SVN" - what is it and do i need it to play without issues?

SVN allows you to update the mod while you game, so you always have the newest mod version and you can continue/update your savegames.

As said before, you can change or add options but some options cause memory leaks or bugs, if you want a stable and working long game its recommended to go with the scenario options i set.

After next UEM update i will add "Hard Conquest" if @Toffer90 can come up with it, otherwise i will go back to standard combat and will take "Always raze cities" out as standard scenario option.

Any suggestions to add a options to the next UEM version? Should be good tested...

EDIT:

@bartigol2 Size Matters has issues, so its not good to use it currently.

 

[Thunderbrd]

@bartigol2 Size Matters has issues, so its not good to use it currently.

I assume you're talking about the worldbuilder interactions with it?

 

[Pit2015]

I assume you're talking about the worldbuilder interactions with it?

Yep, unclear if it affects gameplay also, how can the AI handle it? As i remember @JosEPh_II said AI cant handle it good.

News:

 

[Thunderbrd]

Yep, unclear if it affects gameplay also, how can the AI handle it? As i remember @JosEPh_II said AI cant handle it good.

News:

It does both better and worse in various ways though it's been a bit improved on the AI stuff and more AI fixes are coming that are helping to flat out resolve all sorts of things but none of it is really specific to SM. Some feel the game is more challenging with it and others feel its less so, and a lot of that is based on how the player plays actually. So its a mixed bag either way really.

Interestingly, it helps with memory limit issues to have it on.

 

[Pit2015]

It does both better and worse in various ways though it's been a bit improved on the AI stuff and more AI fixes are coming that are helping to flat out resolve all sorts of things but none of it is really specific to SM. Some feel the game is more challenging with it and others feel its less so, and a lot of that is based on how the player plays actually. So its a mixed bag either way really.

Interestingly, it helps with memory limit issues to have it on.

Ok, looking into to put it maybe into next update.

 

[Pit2015]

@Toffer90 and others, is it not possible to allocate more video memory to C2C somehow? I think that can fix some crash problems.

 

[JosEPh_II]

Keep at it Pit!

 

[Pit2015]

NEWS:

Breakthrough in fusion, after 60 years it looks like the next step is done:

Hopefully the beginning of the end of nuclear plants and a zero carbon future.

https://twitter.com/i/web/status/1602679966163906561

Forschung zu Kernfusion: USA sprechen von Durchbruch

Bei der Kernfusion entsteht je nach Brennstoff wenig oder gar kein Atommüll und es besteht keine Gefahr eines nuklearen Unfalls. Sie gilt auch deshalb als potenzielle Energiequelle der Zukunft. Dem US-Energieministerium zufolge gelang einem Forscherteam nun ein Durchbruch.

www.tagesschau.de

 

[Maltazard]

Physicist on TV said it's gonna take at least 30 years before we see a commercial application after this breakthrough. Fusion power is a joke, it was gonna take 30 years before the breakthrough happened, and again 30 years after the breakthrough, well then it's not a breakthrough at all, it's just an expected milestone. Breakthroughs must cut your ETA shorter.

Also, what kind of a joke is a 50 million dollar investment for fusion power? Pass this research to the Pentagon with its billions, just explain the brass tacks that with fusion they can make energy weapons.

 

[Pit2015]

Physicist on TV said it's gonna take at least 30 years before we see a commercial application after this breakthrough. Fusion power is a joke, it was gonna take 30 years before the breakthrough happened, and again 30 years after the breakthrough, well then it's not a breakthrough at all, it's just an expected milestone. Breakthroughs must cut your ETA shorter.

Also, what kind of a joke is a 50 million dollar investment for fusion power? Pass this research to the Pentagon with its billions, just explain the brass tacks that with fusion they can make energy weapons.

It was done in a military research lab, first breakthrough step, if it works its clean and endless energy, maybe also real space travel and so on, dosent matter how long it takes, if we get it it will be a game changer.

So the C2C fusion techs become real.

2.8 billion was invested in fusion research last year or so.

"Investors like Bill Gates, Jeff Bezos and John Doerr had recently invested a lot of money in companies that are working on the commercialization of fusion energy. According to the Fusion Industry Association, the private sector raised more than $2.8 billion last year alone, reports Reuters."

Btw you dont need always much money to make a breakthrough.

 

[Maltazard]

Fusion has been theorized since 1954 and today they're still facing the exact same challenges as they were back then, all this fuss for this "breaktrhrough" and a "net energy gain" is only referring to the actual reaction, when accounting for the overall power usage of the test facility, it wasn't a net positive. The goal was +1 so not much of a challenge to begin with... usually when you burn something you have a net energy release. If anything, this is ground zero (they were below zero before).

Multiple private companies and private investors looking forward commercial applications of what is supposed to be free energy "too cheap to meter", there's your two problems. It doesn't make sense, it would be against their best interests to actually make this happen.
It could become a reality if it was handled by a single, globally funded public entity. As it is right now, private interests, patent law and all that other nonsense is also involved, further delaying any significant advancements.
Free energy is incompatible with our economic system anyawy. Now, if we think about the military applications, well then that is much more likely to see the light of day.
Will it eventually happen? Hopefully, yeah. Is it being pursued in the best way possible, or in a good enough way? Not even close.