Workers Sitting Idle - priority

[Unreal]

I think that the priority system is jacked up. The ability to select a priority level would be cool.

I have a worker sitting idle, next to an idle mine, next to a priority set empty storage facility. Nothing happens.

[mreed2]

This isn't a bug (working as designed), almost certainly.

I can't tell without a screenshot, but my guess is that you have a storage center located next to a resource field with many mines of the same type (e.g. all water, or all chemicals). Assuming your storage center isn't bugged (see the other thread discussing "overfull storage centers"), the storage center will, at most, pull a single good of the type of the resource field. This demand won't be enough to empty a single mine, much less multiple mines.

If you have the same situation but the mine is the only source of that particular good in the area, then that would be a bug.

Prioritizing a storage center doesn't increase the amount of demand it produces -- it simply ensures that goods demanded by the storage center can't be "stolen" while in transit by other sources of demand, and that it will never release the goods stored within to... Well, just about anything.

Assuming that my read on the situation is correct, you can "fix" this problem by setting a minimum greater than the amount of goods of the type in question. This creates a whole new set of problems, but it will ensure that the local drone ferries resources to the storage center.

See my guide for more details -- there is a whole section on how storage centers work:
https://steamcommunity.com/sharedfiles/filedetails/?id=2319632408

The underlying issue is that Per Aspera runs on a "demand driven" logistics model, and lots of players expect it to work on a "supply driven" model.

In a "supply driven" model, once a good is produce it is assigned a destination, with storage centers serving as a "destination of last resort". In a "demand driven" model, goods remain where they are until a source of demand tags them for delivery.

There are advantages and disadvantages to both models:

• Supply Driven
  • Advantage: Goods don't remain in output bins, allowing production to continue even when there isn't much demand for a good.
    
  • Advantage: Worker drones are far less likely to be busy.
    
  • Disadvantage: Large quantities of goods are introduced into the logistics network, many of which are not immediately useful, making it harder for important goods to reach their destination.
    
  • Disadvantage: Building more mines than you need (given your current demand) increases congestion in the logistics network, which has the very counter-intuitive result of reducing production efficiency rather than increasing it.
• Demand driven
  • Advantage: Goods only move in response to demand, which the developers can easily tune to minimize the number of goods in transit at any given time, maximizing throughput and efficiency.
    
  • Advantage: The amount of congestion in the logistics network is driven only by demand -- if the player places a larger than required amount of suppliers of a particular good, then no extra congestion is generated.
    
  • Disadvantage: It leads to situations as described in the initial post -- even if it isn't a bug, it certainly looks like a bug, and it takes quite a bit of effort to figure out what is going on. Not intuitive at all.
    
  • Disadvantage: Storage centers, to serve any purpose at all, have to generate demand, and whatever solution you choose here, it is likely to be counter-intuitive. Certainly what is implemented is counter-intuitive.

For the record, Settlers 2 (which this game is a homage to) uses a supply driven logistics model, and yes, it creates all the problems described as "disadvantages" in the above list. In that game it is mandatory to build storage centers (storehouses in Settlers 2 lingo) every so often as you play, because if you don't, your logistics network will grind to a halt due to the vast amount of logs, stone and (don't get me started) water that you produce but don't immediately need.

Settlers 2 also has the advantage that its storehouse have unlimited capacity for storage, so once you build a storehouse you never have to worry about it filling up. That's unrealistic, obviously, but without this unlimited storage the game would be unbelievably frustrating. This game has more realistic limited capacity (when not bugged) storage centers, so...

[Talon]

While prioritization was maybe not prime.. I had a game prior to patch3, that did not have the issues I did now with a recent game.. and i didnt have no worker hub for only 3-4 building ratio'd out. I did try to do more with less workers in comparison, but the game just had serious issues trying to get materials moved around even to the few things I had set as priority. While I understand your points Mreed, I still think its ridiculous to have a nearly a 1/4 ratio of your buildings have to be worker hubs. Id be more ok with 1/5 or 1/8, unless you got a high density area packed with level 3 mines and factories, and maybe then you need major worker overlap, but not early to mid-game prior to unlocking any of the level 3 facilities.

[mreed2]

Talon, I'm unsure what you are arguing. I see three options:

1. "Prior to patch 3, my logistics worked well with a drone hub to other ratio of less than 1:4, and now a higher ratio is required to get acceptable performance."
   
2. "Based purely on personal preference, I believe that the logistics should work well with a ratio of 1:5 - 1:8."
   
3. "The game currently works well with a ratio of 1:5 - 1:8."

If #1 is what you wanted to say, then all I can say is that I never tried ratios that low, either pre-Patch 3 or afterwards, so I can't say if the performance in such sparse networks works better or worse currently.

If #2 is what you wanted to say, then... This probably isn't the game for you. The logistics model chosen requires, by design, that you build a large number of drone hubs per non-drone hub buildings, and will perform better the higher that ratio is (1:1 is better than 1:2 is better than 1:3, and so forth). There is zero (mechanical) benefit to not building more drone hubs and the marginal cost of adding more is very, very low, so the only reason for not building an enormous number of drone hubs is to increase the difficulty of the game.

If #3 is what you are saying, I disagree. While it is possible (via using the 16x speed extensively, strategic use of priority, and hyperloops) to complete the game with a 1:5 - 1:8 ratio, I cannot say that the game works well with such a sparse network. Honestly, if you need set the game speed above x4 to get construction done in a reasonable time and / or you have to use priority to get your economy to function then.... The game isn't working well, and with such a sparse network both would be required.

[mreed2]

I just checked my current game, and it has a ratio of 68 drones to 218 total buildings, or ~1:3. So I guess I do have some experience with ratios less than 1:1. :)

[BobthenotsoGreat]

Currently running 1-3. Triangles work. ;-)

[Talon]

Yea.. fine.. but my point i think its stupid to have THAT MANY HUBS take up your otherwise limited building limit in this game. Surv Mars doesnt have an arbitrary building limit. Having 1/3 or MORE of your buildings have to be hubs because of the limits of the game..thats what is stupid. If we are going to have lv2 and lv3 buildings.. give us lv2 and lv3 hubs to have 2-3 workers per hub.. and or remove the building limit entirely.

[mreed2]

The purpose of the building limits is performance, I suspect -- in my end games, when I approach the building limit, I have ~10 fps at 2x speed, ~5 fps at 16x. Other users have reported the same, with the results scaled by computer specification.

I'm pretty sure that the reason that the limiting factor on performance is the worker drones: Each worker drone has its own thread and performs its own pathfinding, but the need to synchronize the main UI thread with the worker drones puts an effective cap on performance. Adding 2 or 3 worker drones per hub would still leave you with the same performance issues -- it would just be harder to prevent players from getting to the point where the performance is unacceptable.

Perhaps the developers can come up with a technique to further increase performance, which will allow raising the building limit, but as long as the worker drones work the way that they do, there will be a practical limit on how many worker drones can exist in a single base. And I, for one, don't support changing the behavior of the worker drones (to, for example, point-to-point transport).

[BobthenotsoGreat]

You don't need 1-3 really end game. You can easily ring the planet with 600 or so buildings including production facilities to support them. Also later, better power buildings can have you tear down the hundreds of solar save for connections.

Also, how many mines are enough? You can ring the planet with hubs or you can loop it with around half the buildings.

Excess production only occurs when demand requires it. This isn't saving mars where colonists mean anything after you have researched. Maybe later, but not now.

But if they improve how roads are done and remove the bugs (for roads - and good placement) roads are pretty good which further reduces hub need.

But you need storage to work much better for that. Loop the planet, place production at loops, place storage and hub nest there to. Then have feeder roads that bring the mines in.

But we don't have that ....yet.

[mreed2]

Agreed -- you can get by with a much lower ratio once you have hyperloops.

However, to achieve this means spending lots and lots of time deconstructing worker hubs, solar farms, and so forth -- how many people are likely to invest that much time, when the only benefit is to improve performance / free up buildings in the building limit?

[BobthenotsoGreat]

They wouldn't without a limit. They would report seconds per frame as a bug.

But needs must. To me it's no different to the 50 building limit when you start. You need the colonists down, and fed but you also need to prep for the future with just the 50 buildings. Then its up to you how many you want - upto the limit. Artificial yes but we can see why.