i don't mind stutters or drops in fps in general, would you say ram had an affect on how quickly the months themselves passed?
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
i don't mind stutters or drops in fps in general, would you say ram had an affect on how quickly the months themselves passed?
If the late game is still slow for you I would guess it's not compiling the scripts for you either, you may want to investigate why it takes so long to start. If it can start in 6 seconds on average computers it shouldn't be that slow on others. I don't think I've seen any PDS game take more than 10 seconds to start and my guess is that they all treat scripts the same way. I remember there were some posts on EU4 forums that were mentioning unusually slow startup and that was something specific to some version of Windows. Though I guess improving startup time is not going to make a huge difference...
In the absolute, yes. Like, just a bit. Strictly not worth an upgrade.
I read where running the game without Steam, starting it from the Paradox launcher would speed the end-game.
I tried it and it works.
The only Paradox launcher that I could get to work is the Beta version 1.0-beta-50 at Program Files (x86)\Paradox Interactive\bootstrapper\bootstrapper.exe.
You may not be able to run full-screen. To speed the game further reduce the resolution from High th Medium.
I tried it and it works.
The only Paradox launcher that I could get to work is the Beta version 1.0-beta-50 at Program Files (x86)\Paradox Interactive\bootstrapper\bootstrapper.exe.
You may not be able to run full-screen. To speed the game further reduce the resolution from High th Medium.
We're talking about start up right, from desktop?
Anyway I know the clauswitz engine compiles the scripts, it's how the checksum is made. At the very least it reads from every single txt and mod file it's set to load on start up.
2257. 1000 stars, max advanced AI, 0.25 hw, 11.81 seconds.
2300. 17.68 s.
2352. 19.41 s. (it is 30 Ai at the start, about maybe 27 alive, player population is 450, ais have something like 300-400)
2401. 25.85 s. Pop 983
2425. 32.81 s. Pop 1433. War in Heaven, scourge arrived, all the good stuff.
2459. 30.93 s. Pop 1901
2475. 32 s. Pop 2208
2503. 41.64 s. Pop 2904.
2550. 50.37 s. Pop 4681
2300. 17.68 s.
2352. 19.41 s. (it is 30 Ai at the start, about maybe 27 alive, player population is 450, ais have something like 300-400)
2401. 25.85 s. Pop 983
2425. 32.81 s. Pop 1433. War in Heaven, scourge arrived, all the good stuff.
2459. 30.93 s. Pop 1901
2475. 32 s. Pop 2208
2503. 41.64 s. Pop 2904.
2550. 50.37 s. Pop 4681
Last edited:
I was counting from the launcher screen, but it's almost the same. From the desktop it would probably be an extra 0.5s.
I agree it reads all/most text files (you can see that by checking what files it accesses) - that by itself is enough to create a checksum. It probably parses them too, but I didn't see any evidence of compilation.
1. People quoting star numbers and years should also include global population count and number of AI empires alive. Otherwise data is meaningless. I can have an empty huge galaxy going lightning fast at year 2600.
2. They have ported 2.2 to console. I assume that console has no mod support and thus would be extra-optimised for performace, otherwise I pity the poor console players....
2. They have ported 2.2 to console. I assume that console has no mod support and thus would be extra-optimised for performace, otherwise I pity the poor console players....
I haven't played since 2.3.x BUT I've had decent luck with SMALL galaxies, low habitability, and other settings that would keep the pops, housing, and job numbers lower. I wrote a homebrew mod that further reduced POPs, Housing, Jobs by 1/2 to 2/3 and that also seemed to help. With all of the above even a 7-8 year old system should be able to get you to the standard CRISIS timing without TOO much hassle. A lot of it depends on your personal threshold of "slow but playable" vs. "unplayable".
Note: I heard some elements of performance may have improved since 2.3. If things actually got WORSE when it comes to jobs, pops, and similar then the above statement based on 2.3.x performance may not hold.
I don't know if it helps you BUT you could always try SMALL galaxies with low habitability and no guaranteed starting planets. Seems to work well enough for me with 2.3.x ....
I would THINK (??) that if you snagged a mod that reduced pops, pop growth, pop assembly, jobs, housing, habitats, and similar you could at least INDIRECTLY attack some of the issues.
NOTE: My system was "really good" about 7-8 years ago but is NOTHING compared to current systems. IF you were willing to make HUGE SACRIFICES in the game types you wanted to play I think you could get a game in -- at least until standard crisis.
okay i remembered why i originally believed they're compiled, you can edit the text files while the game is running and changes are not accounted for. the text files are already read and turned into precompiled calls it saves on RAM. it essentially, does exactly what it does for parsing them, but doesn't need to save all the text files onto RAM. there's essentially no reason for them to keep the text files around or access the text files during actual play.
I don't know if this is a confusion of terms, but it's not true compiling (i was using the layman usage of the term, as in it's collecting all of them in one place in an orderly fashion), it's essentially turning all the text files into an object that is a list of what function to call in what order and when with what arguments. Parsing would require the game to access your hard drive, which is how save games are read and the like.
at least i don't see any reason why they wouldn't do this, they could be saving the text files on RAM and parsing them, but why not just parse all the files once.
is stellaris installed on an SSD for you?
all paradox games, on several different computers always take several minutes for me to start, so i doubt there's a hardware thing on my end.
Would it be crazy to overhaul the engine?? Not necessarily. Unfortunately I'd be willing to bet Paradox will NOT do it solely for the benefit of customers. It's going to come down to an issue of what can the do that maximizes medium and possibly long term income.
Specifically some gotchas would be things like:
- Should they do this overhaul or do a Stellaris 2 .... similar to the Anthem issue where the devs more-or-less GAVE UP fixing the train-wreck of the original game and just started working on Anthem 2.
- Is there an ROI for this? I.E. will an overhaul wind up making them MORE money than not doing an overhaul and rolling out a few expansions instead. My guess is that there are WAY too many people buying expansions for Paradox to justify this.
- Should they ditch some of the current systems that are the source of the problems in favor of increasing MICRO even further?? For example if the PLAYER handled 99% of the job assignment / creation that would get rid of checks for job shuffling as an example. While I don't like the performance we have now I "think" I may dislike the perceived MICRO load on the players even more.
What would I do in their place seeing as how I dislike BOTH the 2.3.x MICRO load && Performance:
- I would TRY to keep the spirit of the game we have now
- I would be willing to make concessions in the reality vs. economic events spectrum. For example economic events would likely only be calculated "once per month".
- I would be willing to make concessions when it comes to NUMBERs. For example I'd try to cut the number of POPs, JOBs, and HOUSING dramatically ... say 1/3 the value it is currently. Rebalancing a LOT of things would be required to have this flow properly.
- I would give players the option to turn OFF a lot of the smart automation. [I.E. no job shuffling at the cost of MORE micro]
- I would look at systems to "prune" the number of calculations made. I.E. I would allow the player to create species templates and associate those with "jobs". I would also allow for prioritizing jobs. Combine all of this and automating a "good employee to correct job" becomes a more straightforward calculation.
In a world with unlimited resources AND no concerns about money [silly] I'd certainly remake the engine from scratch. Say "Recommended" system similar to AMD R5-3600 + 32GB RAM + SSD + RX590 [GTX1660] and have it able to use almost every last drop of horsepower & memory. I would try to scale it down but the MINIMUM would likely be equivalent to a good 5-year-old system ... 4 Core 3.5GHz ; 16 GB Ram ; 4 GB video card and hope for the best
FYI: Several revisions ago we didn't see nearly the expected difference between newer systems [1-2 year "power" systems] and older systems like mine that were about 7 years old. I don't think it's SIMPLY a horsepower issue but also intertwined with some hardset timers and a lot of other seemingly "poor" behavior from the engine.
I wouldn't expect night & day difference. I think you might DELAY the point where the game goes "unplayable-for-you". From prior testing Stellaris doesn't use the HP it has even when it's running slower than it should. So my best guess is it will DELAY problems but won't fix them for you.
You could try limiting the "pops" as best you can during game creation ... Tiny / Small galaxies with low habitability + no guaranteed planets can help. Mods that get rid of habitats OR otherwise limit population growth might help indirectly.
I got by with an i5 3570K with those settings ...
There is some confusion of terms. In technical terms parsing converts plain text files containing Paradox scripts into a parse tree stored in the memory of the process (this is a rather fast conversion, it probably can be done while reading the files without incurring any extra time). Compiling is an ambiguous term even within computer science. It is commonly used to indicate translation into machine instruction, but, more generally, it can mean translation into some other language. In the context of this discussion I meant compiling as compiling into machine instructions because that's what is required to have a good performance. This is a fairly expensive conversion.
Yes, it's on SSD. I don't think there is enough hardware spread to explain such difference. My guess would that it's something like active anti-virus or compression on the file system or disk driver or something like that that makes it slow for you.
To remove doubts regarding compilation I've run a small test. I've created a mod with the following event:
Code:
namespace = test
country_event = {
id = test.1
title = "test.1.name"
desc = "test.1.desc"
picture = GFX_evt_synth_organic_relations
is_triggered_only = yes
option = {
immediate = {
set_variable = {
which = total
value = 0
}
log = "test: *** Starting..."
while = {
count = 100000000
if = {
limit = {
check_variable = {
which = total
value < 1000
}
}
change_variable = {
which = total
value = 1
}
}
else = {
change_variable = {
which = total
value = 2
}
}
}
log = "test: [This.total]"
}
}
}The I wrote equivalent in C++
Code:
#include <iostream>
#include <cstdlib>
#include <vector>
#include <chrono>
int main(int argc, char* argv[])
{
auto start = std::chrono::high_resolution_clock::now();
int64_t count = 100000000;
volatile int64_t total = 0;
for (int64_t i = 0; i < count; ++i)
{
if (total < 1000)
total += 1;
else
total += 2;
}
auto end = std::chrono::high_resolution_clock::now();
std::chrono::duration<double> diff = end-start;
std::cout << total << std::endl;
std::cout << "Elapsed time = " << diff.count() << "s" << std::endl;
return 0;
}
What do they compile to darn. A first version of Cobol? I once wrote several puzzle solver algorithms in Cobol just to prove it could be done. It worked but was slow.
But it's a lot of difference between your test and the mod, wow.
But it's a lot of difference between your test and the mod, wow.
It runs fine until it doesn't and then it's really really bad, not just low fps, stutters that last for a second at a time constantly. That's my definition of unplayable. I wouldn't care if it slowed down a bit when everything gets hectic, that would make sense. This is something broken in game.
sorry, yeah, i tend to always forget what things are called, but of course using those libraries will be significantly faster than anything, they've been assembly tweaked for decades. this has just been a semantic debate, sorry for wasting our time, because the parse tree is stored; that to me, seems like compiling, while parsing generally only refers to when it's actually occurring(reading a file or equivalent).
those are the speeds of what i'd expect them to be different when ran through something like stellaris. XD i can hardly think stellaris'd be set up to perform loops very well.
on the load time. I'm not really sure, I don't tend to run with antivirus, but i recently installed one after becoming the head of IT for my company (i'm the only one in IT at my company so it's not as extravagant as it seems) and i do feel i took a hit to the load time after that happened, but it was still at least a minute load time before hand. this applies to all paradox games as well, so it's not a problem with any specific files, any other ideas you have would be greatly appreciated.
Exactly that is what rally baffles me about you people...
This attitude is why never anything gets really fixed. Example is Star Trek Online (which I currently revisit after 7 years): Half of the missions are either broken or massively stacked against the player. But instead of fixing it, they gave the player a "skip" button. But they still pump out new microtransactions stuffs and the people still buy said stuff...
As already mentioned, I have 4.000 hours on CKII but I won't ever touch CKIII - even if PDX would gave it to me for free!
(Let's be honest, that will never happen as I'm only a fan and not a fanboi and that particular straw, why I will never play it anyway, broke back when they "changed" the mediterran portrait pack, cut half of it and sold it as a different DLC)
As long as people, with more money than they actually deserve, blindly buy, politics of Publisher/Developer will never change!
In such cases I always like to give my buddy Steve 1:39 minutes of your time:
Also: I find the lack of Developer in this thread disturbing!
It's a containment thread, developer involvement is pretty much just moving trouble topics into this one.
It's hard to make a guess. I would try to look at where the time is spent (like user vs kernel, possibly in what module, which might point to AV, for example). Also if it's spent in disk IO or not. The choice of the tool would depend on OS, but all of them have tools to monitor that. What OS are you using?