• 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.
Here it is:

If number of pops are used to calculate diplomatic weight, do different pops count for different amounts? Eg ruler pops count more than workers, free workers pops more than slaves, slaves or workers count more than robots or gestalts?
Pops are worth a base of 1 Diplomatic Weight, with those that are subject to happiness adding twice their happiness. Criminals, deviant or corrupt drones, presapients, non-sentient robots, and those undergoing purges or assimilation do not add to diplomatic weight.
 
Both cases are scary, because they haven't commmunicated any progress on any fix while they are communicating new pain for the engine. In my book this is Faux pas - a red card by football terms: as a product manager/director or community manager I would have definitely been posting here - Especially since the thread has reached the milestone of 50 pages.

50 pages and you still haven't realised that this is just a containment thread so they have an excuse to delete complaints posted elsewhere?
 
Just go on reddit, the complaints are more real there. This here is moderated in a positive vibe for paradox of course, its their own forum. When the next major patch doesnt solve the problems with performance and ai at least partly (and mostly) i will not buy any further content for this franchise (federations included). There was a huge outcry a few weeks/months ago and i think this was the reason they delayed the release.
These improvements arent sexy new content which sells, but it is essentially for players like me to play the game again. I have all content from Stellaris negative rated on Steam right now. If you arent pleased with the product tell this to potentially new customers and boycott. There is no excuse for me that the game is in this state for over one year now.
 
Just read the Q&A that came out today.

There's a lot of additions incoming that will require even more CPU work with the new patch. Reading through the features, there will be many new triggers and other concerns for the AI to waste more CPU resources. What caught my mind especially was the calculation of the diplomatic weight: It counts each pop, and depending on classification it calculates a diplomatic weight based on it's effective happines. OUCH! Yay! Another reason to iterrate over all pops and touch all that data!! And happiness changes all the time somewhere in any empire because the sun just came up. So that will be another major thing to calculate and keep up to date. Of course I have no clue how iit will be done. I would assume on a per colony basis, every time happiness is modified - or once per month?

.

This isn't that heavy in CPU cycles. It probably has a speed of n. While checking for open jobs is probably nsquared or log(n) at least. Like doing math isn't difficult for a machine, it's comparing things that is hard, like is x the highest value in a list? Or what pop best fits in this job.
 
This isn't that heavy in CPU cycles. It probably has a speed of n. While checking for open jobs is probably nsquared or log(n) at least. Like doing math isn't difficult for a machine, it's comparing things that is hard, like is x the highest value in a list? Or what pop best fits in this job.

That and most of the relevant values are cycled on already for factions purposes. Certainly diplo weight calcs will be done on the fly while faction calcs are performed, to benefit from the same data access.
 
That and most of the relevant values are cycled on already for factions purposes. Certainly diplo weight calcs will be done on the fly while faction calcs are performed, to benefit from the same data access.

To me this statement seems very confident about their coding. While yes, this would be the ideal implementation, so far it occurs that there are a lot of redundant checks and calculations in the code. (One example would be pops jumping in and out of jobs on a daily basis, even though it only matters one the first day each month. There probably are more examples.)

Anyway, I still remain (cautiously) optimistic about federations - dedicated performance- / AI-development and such. Fingers crossed.
 
In general, this is a wide problem of our time than you would expect. Here is the best vid I know for this: https://youtu.be/owcvg2YZ7Y8
 
To me this statement seems very confident about their coding. While yes, this would be the ideal implementation, so far it occurs that there are a lot of redundant checks and calculations in the code. (One example would be pops jumping in and out of jobs on a daily basis, even though it only matters one the first day each month. There probably are more examples.)

Anyway, I still remain (cautiously) optimistic about federations - dedicated performance- / AI-development and such. Fingers crossed.

it's not redundant, UI philosophy in general is that it needs to update as the player is looking at it, or else people tend to get confused since stuff didn't update immediately.

like say the system was set up so it prioritized food production when it's low (which i believe it does), and so a new player not knowing this, manually moves pops over to food production before the month ticks over, him complaining that the system obviously isn't working and begins to complain that the system isn't working and when he's corrected, that the system is obtuse.

say you enact a policy that increases food profits via increasing amenities and thus you don't need as many food workers. if you only calculated on the month, it would look like you have too many food workers and moving them to other jobs would be tedious when the UI would do it on the month tick, it's just not obvious, hence the UI updated continuously. (you could perhaps do a cooldown on job switching of 1 month, but well, what if you're changing a LOT on a planet and so they'd need to change multiple times, there's no real way to win here)

this is only for food, i hope this becomes obvious that if they went with a method of "only update when something's changed" that you're starting to get triggers all over the place, this will make the code very "spaghetti" and make it harder to update in the future as the use of various triggers must stay consistent or else you'll get very hard to fix bugs.

my opinion is they they need to go from the pop system as we have it to a victorian style pop system, where individual pops aren't kept but instead you have pop entities that say there are x many pops working job x of race y. so instead of having 3 humans working alloys, you have an object for alloy jobs that says 3 humans of the egalitarian faction working here, this is their calculated upkeep, happiness, political weight, and output, etc.

this would have largely no benefit for smaller empire but would become comparatively quicker the longer the galaxy exists.

however it would require a complete rewrite of the pop system, rather disappointing.
 
Last edited:
Stellaris is currently suffering from a lack of competition. Basically if you wanted to play a space strategy game in 2017, 2018 to 2019 Stellaris was certainly the only one around; and a monopoly. Yes, Endless Space 2 is rather new, however its not RTS and also no longer receives updates. But basically because of this monopoly, the development team could add a few shiny things to the game and people would buy the DLC, 'cause it was the only game in town. This is exactly what they did, create a few DLC which superficially advertise the balancing and optimization of the game, but really just add irrelevant extra content without really fixing the underlying problems. Ring Worlds, Titans, World Crackers., Habitats, Megacorps and L Cluster etc. sell good but broke the game in the process.

I think for 2020 the Stellaris Devs will have to buckle up, because the road is gonna get tough. If they don't fix performance, AI, and the abysmally bad balance, new contenders (Distant Worlds 2) will probably cause a spectacular failure of the Stellaris franchise. Because other companies have seen how successful Stellaris was and want a piece of the pie. Currently I see no way for PDX to stop said contenders from taking a big slice.

As it stands, Stellaris is fun for the first 50 hours of gameplay. But once you see the big cracks and faults in the whole game, and start looking to the forums, and then realize that in the last two years (since 2.0) the problem has become constantly worse, you realize the game is basically dead.

I no longer expect real improvements, just cosmetic changes and bling bling to get a few more people started in the game who don't see the cracks yet. Im looking at you Juggernaut.
 
Most of the above comments assume that the new calculations will be done in a hardcoded fashion in the engine. If that were true we would have been capable of playing with
100.000 of pops ages ago but without modding.

What happens is that most of those things are calculated through triggers and throught the modifiers system outlined in the last performance dev diary ages ago. And you bet the engine programmers work only on future engine upgrades or universal changes that apply to all titles, and not exlusively or dedicated on the stellaris DLC.

So no, it won't be O(n) since triggers and events will make it cycle through stuff, as stuff changes in the data graph and go through additional overhead that we don't know since we don't know the details of the engine..

A colony datastructure is not just a C struct in memory - it's a beast. You can't code stellaris/clausewitz with only a CS101 data structures course knowledge.

-edited to correct horrible spelling- I type too fast and don't read the result...
 
Last edited:
We pretty much know. But what can we do?

About the same that was done in any other [high profile] containment thread I guess.


Just go on reddit, the complaints are more real there. This here is moderated in a positive vibe for paradox of course, its their own forum. When the next major patch doesnt solve the problems with performance and ai at least partly (and mostly) i will not buy any further content for this franchise (federations included). There was a huge outcry a few weeks/months ago and i think this was the reason they delayed the release.
These improvements arent sexy new content which sells, but it is essentially for players like me to play the game again. I have all content from Stellaris negative rated on Steam right now. If you arent pleased with the product tell this to potentially new customers and boycott. There is no excuse for me that the game is in this state for over one year now.


I mostly agree with your outlook. I am bummed though that the choice to fix things that I personally care about [better crisis, better awakened empires, performance, lower micro, ...] hasn't had the attention I would have liked to have seen. I just wish enough of my fellow players felt the same way so that Paradox would choose to fix these elements as it would be their best way to ongoing short & long-term profits. Unfortunately, as a business, it seems like making a new shiny widget is more cost effective than going back and fixing the widgets I already bought ... :(


Let's face it, this is the last thread they'd ever dare close.

I think they would dare IF they considered it in their best interests as a business. Note: They might not have to close a thread to take some of the steam out of it. For example I suspect Paradox could keep certain [forum policy breaking] members from posting to certain specific threads. I know that I [intentionally or otherwise] lost the ability to post to another famous containment thread so I would assume the ABILITY for Paradox to do that is there.
 
Just to ask a question of you guys, I am obviously as livid as everyone else that the game has become unplayable after a certain point. I tried 2 playthroughs after megacorp and distant stars both on large galaxies (why not? I have a decent pc and this always ran like butter before) and the grinding slowdown forced me to give up both games (both of which I had probably put about a day of real time into).

My question: if I were to play in a medium galaxy would I have a chance of actually finishing a game in its current state?

Obviously spending money on additions which have made the game unplayable has severally burnt my trust with this game. The lithoids dlc was the first dlc I did not buy and I will not spend any more, of the considerable for a video game, money on this game until I see some wonderfully grateful posts on this forum or elsewhere thanking the developers for fixing the game that they broke. Thank you.

i7 7700 16gb ram 1660ti.... Geez have you seen the system requirements on steam?

End of rant.
 
Most of the above comments assume that the new calculations will be done in a hardcoded fashion in the engine. If that were true we would have been capable of playing with
100.000 of pops ages ago but without modding.

What happens is that most of those things are calculated through triggers and throught the modifiers system outlined in the last performance dev diary ages ago. And you bet the engine programmers work only on future engine upgrades or universal changes that apply to all titles, and not exlusively or dedicated on the stellaris DLC.

So no, it won't be O(n) since triggers and events will make it cycle through stuff, as stuff changes in the data graph and go through additional overhead that we don't know since we don't know the details of the engine..

A colony datastructure is not just a C struct in memory - it's a beast. You can't code stellaris/clausewitz with only a CS101 data structures course knowledge.

-edited to correct horrible spelling- I type too fast and don't read the result...
Good day. You seem to know what you are talking about, certainly more than me. Is the idea of an overhaul of the engine impossible for the most popular space strategy game actually impossible?

If I am not incorrect Sins of a solar empire managed to buff up their ancient engine 5 or 6 years after release.
 
My question: if I were to play in a medium galaxy would I have a chance of actually finishing a game in its current state?
End of rant.

I have a very similar situation than you - and I switched to medium galaxies. Result:
Even on this level the game slows significantly down in lategame: I stopped my last two games during this phase. It's not as ugly as in big galaxies (way better), but after a while...

Unfortunately you can't kill this beast with hardware. I tried: changed my CPU from i7 4990k to Ryzen 3800x (and upgraded the rest properly), and the issue is still there, even on medium.
 
I have a very similar situation than you - and I switched to medium galaxies. Result:
Even on this level the game slows significantly down in lategame: I stopped my last two games during this phase. It's not as ugly as in big galaxies (way better), but after a while...

Unfortunately you can't kill this beast with hardware. I tried: changed my CPU from i7 4990k to Ryzen 3800x (and upgraded the rest properly), and the issue is still there, even on medium.

I suspect and from perusing this thread that the problem isn't the graphics. While they may be taxing in enormous battles the problem is related to the way calculations are made with relation to the post 2.0 population system and the engine not being able to handle it. I was very irritated when I noticed it after megacorp and shocked to see that it wasn't fixed for ancient relics. I'd like to hear from someone who has a normal PC and can run a large game without suffering form stuttering and awful framerate drops. Again I suspect that the only way to fix this is a complete rework of whatever they did in 2.2 or a reworked engine. I hope they do that. It was a good game and in total at launch prices the most I have ever spent on a video game.