Factorio Forums

Hi,
Hi,

TL;DR: Logistic requests from space are invalid (way too high, like ten-folds too high) and inconsistent (difference in requests for the same item(s) over 2 sets off rocket-silo sets within the same logistics network)

We have a save that we've been working on for a bit (multiplayer, but sometimes this save is loaded in singleplayer too), as everything has grown we're now noticing something odd with logistic requests from space, for a while, this even caused the entire logistics system to stop working (all rockets were loaded with rockets (the ones for the turret, that is) and none of them were firing, even though all but 2 space platforms were at nauvis and were ready to receive there loads).
This issue had to be resolved 'manually' by simply clicking through all 224 main rocket silo's and clicking on 'deliver cargo' for a few cycles, at some point the entire thing started to just work again, it's still unclear why this happened.

However, this is where we found something odd: the logistics network said that was a request for 7.2 million rockets from space (no I unfortunately don't have screenshots of that anymore). The strange thing is, all space ships together would only request a max of 92k (92 thousand) rockets, definitely not 7.2 million.
This behavior held true for other resources as well.
After some more manual "sending rockets up", this sort of resolved itself, but numbers have since been climbing again.
It seems as though the space platforms keep "adding" to the request(s) each cycle or something like that, this is the only explanation I could come up with why the request numbers would be so high.

What's more, we have a set of separate rocket silo's for some more requested parts, I connected a wire to one of those silo's and to my surprise, the logistic requests were NOT the same. The same items were 'there' as being requested, but the number of items being requested was/is significantly lower.

Now, this is problematic for 2 reasons: 1) it makes these numbers unreliable and so you can't really 'act' on them, 2) it can (apparently) grind the entire space logistics system to a complete halt

Here's some screenshots of the logistic requests for both sets of rocket-silo's, note the stark difference in requests for the very same item(s): (and the other set, notice how the request for rocket turrets is there, but the number is WAY lower): Note that both sets of silo's are connected to one and the same logistics network. Also, both request numbers for the rocket turret are incorrect, at most this would be roughly 9000 of them being requested.
Here's a screenshot of the logistics network(s) (there's only one) on the planet:
To us this seems like a bug. If not we'd love to know what exactly we're doing wrong here

Save file and logs can be found here -> https://drive.google.com/drive/folders/ ... drive_link

Post a save where the issue can be observed please and describe what values are erroneous and what values you would expect to instead, see 3638.

My apologies @Loewchen, I should've attached those from the get-go.
I've put the save file and the log files in a google drive dir (save is too large for direct upload), they can be found here -> https://drive.google.com/drive/folders/ ... drive_link

In the example in this post, I'd expect the number of "rocket turrets" being requested to be at most 5000 (should be even less at the time the save was made, some of it was already satisfied), but instead it's requesting a whopping 900000 (900k). These can be observed in the save file when looking at the rocket silo's on nauvis, or in the screenshot in the post (top right).

I believe the issue is you have 224 rocket silos, all reading orbital requests and transmitting it to the same green circuit wire.
So if SS BoomBoom starts requesting its 31k rockets, the number gets multiplied to 6.9 million.

Regarding the perceived unresponsiveness of rocket launches and the need to launch them manually, that sounds like 122321 Not all rockets launching when ready (fixed in 2.0.26).

hmm, yes, the summing of the number by 224 makes 'sense', although I'd still say this is a bug, it's to read orbital requests, not the sum of all silo's, right?

As for the unresponsiveness of the rockets, if it was "fixed" in 2.0.26, then the fix might not have caught all possible scenario's as this occurred on a version higher than 2.0.26

It's how circuit network wires work - they sum signals from all entities they are connected to that want to output something. They don't care that the values are ultimately coming from the same place.
It would be the same if you join two constant combinators configured the same. The values on the wire will be twice the values in the logistic group the combinators are using.

If you encounter some idle rockets not launching, post a save so it could be examined.

Ah, ok.
If/when the idle-rocket-issue pops-up again, I'll post a separate bug report for that again with a save

Thnx for the replies