Please Help - Possible Bug

[ColonelSandersLite]

Note the difference between consumption measured in the game's production/consumption window and that which was measured directly by the circuit network. Probably a bug. Will wait for others to confirm before reporting.

The math I have done on the research points getting generated strongly suggests that the circuit network output is correct, and the game's consumption display is incorrect.

Note - I would greatly appreciate it if someone that is more of a Factorio veteran would check my circuit network to ensure that it is correct. I'm pretty sure it is, but not 100%.

Update - This is caused by the game merging partially consumed flasks. The merge does not count as "item consumed". The circuit network is showing the true rate of science consumption.

20180520035918_1.jpg (495.03 KiB) Viewed 3533 times

Save (0.16.36):

Consumption Tabulation Bug.zip

Savegame

(4.39 MiB) Downloaded 79 times

Reproduction:
Delete the construction.
Place the blueprint.
Delete the combinators.
Wait a few minutes for the lab inputs to stabilize.
Place the blueprint over top the rest of the setup to start the circuit network tabulating.

Wait 15 minutes (or an hour) for the averages to stabilize and compare the results.





Edit:
The below is the original post, containing extraneous information. Preserved for archival reasons.


Alright, this is probably either a blindingly obvious problem that will make me feel really dumb for not spotting it, or a bug.

I've been working out the stuff I need to meet or exceed my goal of producing and consuming 100 science per minute.

At the production end of things, I have overbuilt by about 10% to account for "shit happens". The lowest producer is a tie between red and black science at 109.2/min.

A train loads 1000 science packs at each of 7 loading stations where that science type is produced. It then delivers all 7000 packs, 1000 of each, to a research station.

I don't remember the exact consumption numbers, but I know for sure that the labs are capable of eating the science packs faster than they're being produced.

The train can unload all 7000 science packs and make it back to the first science loading station (red) before 1000 science packs are produced and must wait there.


By my thinking, in the system described the bottleneck is the red and/or black science packs and the whole system should consume science packs at a rate of 109.2/min.

These are the results the game is telling me:

20180519182043_1.jpg (608 KiB) Viewed 3586 times

That's an 18% discrepancy...



I've tried chasing this down from every direction I can think of.

I calculate that with a production rate of 109.2 items/sec, the train should be on a 9m 9.45s loop. A quick look with a watch showed me that it was actually 9m 14s. Assuming that this is *every* trip and real time is perfectly synched with game time, that's a delivery rate of 108.3/min for each science pack.

As my last resort I finally set up a little circuit to actually measure and compute the amount of science/min moving down the belts into my labs and found:

20180519200639_1.jpg (372.88 KiB) Viewed 3586 times

The circuit network shows the expected 109/min for each science pack, yet the production/consumption screen still shows 89-90/min.


So, did I just find a bug, or am I just missing something?

Edit: Typo

[Zavian]

The way the lab consumption rate drops every time the silo works makes me think that the labs are running short on something every time the silo works. (Are you waiting until you are all out of space science flasks before starting to make a rocket? )

Any chance that you are accumulating science flasks somewhere? Maybe at the train stations? Because unless science flasks are accumulating somewhere, I would have expected the production rate to equal consumption rate.

If you really think you have found a bug, I'm going to suggest uploading the save.

[ColonelSandersLite]

As pointed out, the expected bottleneck is the red/black production at 109.2/min

The labs eat science a bit faster than I can make it. That way, if there's a backlog of built up science packs, it will eventually be cleared. For example, if I disable research for a while, or forget to press T to start a new research.

There's no accumulation of the reds and blacks. There is a very slight accumulation of the others, just because they produce a bit faster. The purple and yellows are producing at 114/min so that's an accumulation of 288 of those over the course of a whole hour.

Will upload save in a bit. Just wanted to know if anybody would spot an obvious problem before I do.

[Lav]

Your science consumption graph shows sharp drops every 12 minutes or so. Your pic shows the situation during the "normal" period, but what happens during the drops?

It's also suspicious that frequency of those drops matches the frequency of white science production. Could your labs experience periodic deficit? Afrer all, 89/m is the average for the chosen period, which includes those drops, so your current consumption speed must be higher than that - maybe even higher than calculated.

Maybe your labs simply consume packs faster than you calculated? Have you researched research speed?

[ColonelSandersLite]

An additional test -

As shown earlier, a circuit network setup reads that the labs are getting handed 109.2/min for each pack, but the consumption graph disagrees. So which is correct? Perhaps they are both correct and science packs are getting lost at the labs somehow?



My understanding is that when a lab consumes 1 of each required pack, 1*Productivity research points are gained.

Follower Count 52 and 53
Cost 5500 + 5600 = 11100 research points

If the labs are consuming 90/min as the game says:
consumption rate = 90/min
+20% productivity
90*1.2 = 108 research points/min
11100/131.04 = 102.78 mins to complete

If the labs are consuming 109.2/min as my logic says it must be, and is confirmed by my circuit network:
consumption rate = 109.2/min
+20% productivity
109.2*1.2 = 131.04 research points/min
11100/131.04 = 84.70 mins to complete


I rigged up a simple in game clock and measured the time from the research start to end.
Result - 84 mins

That confirms for me that the labs *must* be consuming 109.2/minute to produce 131.04 research points/min. I have no idea where the game is getting the 90/min number. My best guess ATM is that it's some sort of tabulation bug.



Will upload a save soon.

[ColonelSandersLite]

Your science consumption graph shows sharp drops every 12 minutes or so. Your pic shows the situation during the "normal" period, but what happens during the drops?

It's also suspicious that frequency of those drops matches the frequency of white science production. Could your labs experience periodic deficit? Afrer all, 89/m is the average for the chosen period, which includes those drops, so your current consumption speed must be higher than that - maybe even higher than calculated.

Maybe your labs simply consume packs faster than you calculated? Have you researched research speed?

My labs are able to consume more research than I can produce. That's not unexpected.

The interval of those drops is actually every 9 minutes and change. To be more explicit:

This is a 100 science / minute system. This means 1 rocket launch every 10 minutes. This means a delivery of 1000 of every science pack to the labs every 10 minutes. Delivery is actually a bit faster (9 mins 9 seconds) as I overproduce a bit. The labs eat 1000 of each science pack every delivery cycle.

The labs are designed to be able to eat something like 120 of each pack per minute to clear backlogs of science.

You're looking at the problem from the perspective of - Why aren't my labs getting 100 science per minute. That's backwards. The problem is - How is it possible that a system that delivers science at 109.2/min and *is* fully consumed getting reported as consuming 90/min?

Edit: Clarification

[DaveMcW]

The labs are designed to be able to eat something like 12000 of each pack per minute to clear backlogs of science.

I don't see any evidence of that.

36 labs * 350% research speed * 70% module speed / 60 research time = 1.47 packs per second = 88.2 packs per minute.

That is a lot lower than 12000 per minute, and not even enough for 100 per minute.

[ColonelSandersLite]

Easy to check!

20180520031641_1.jpg (710.51 KiB) Viewed 3545 times

The game's consumption screen reports 111/m

Measuring from the feed belts shows 150/m


Where I said 12000/min before was a typo btw. My numpad 0 key likes to stick sometimes and I just didn't notice it. I meant 120/min.


That being said, neither the game's consumption screen or my direct measurements agree with your numbers. I don't know where they actually come from. Do you?

I'm glad you brought it up though, as now I can create a save that shows the problem and *only* the problem with relative ease.

[ColonelSandersLite]

Initial post updated. Further clarified and simplified and a savegame included.

[Zavian]

There definitely is something wrong with the lab flask consumption stats. Below is the production stats window of a belt based lab setup where I carefully fed 8 labs 6x25 science flasks. Note the production stats shows I consumed 197 of each flask.

Test2A.png (1.12 MiB) Viewed 3529 times

Labs where flasks are moved from lab to lab are even worse. (Below)

Test2B.jpg (335.17 KiB) Viewed 3528 times

I'll prepare a bug report shortly.

[Hedning1390]

I have not read through this thread, but flasks are unlike any other intermediate item in that they have durability. An almost used up flask doesn't count as used until it is completely gone.

[Zavian]

Bug report at viewtopic.php?f=7&t=60492 .

I have not read through this thread, but flasks are unlike any other intermediate item in that they have durability. An almost used up flask doesn't count as used until it is completely gone.

In my testing, I made sure that all the science flasks (except space science, I used excess space science flasks) were completely used before those screenshots from my post above.

[ColonelSandersLite]

The labs are designed to be able to eat something like 12000 of each pack per minute to clear backlogs of science.

I don't see any evidence of that.

36 labs * 350% research speed * 70% module speed / 60 research time = 1.47 packs per second = 88.2 packs per minute.

That is a lot lower than 12000 per minute, and not even enough for 100 per minute.

I got back to thinking on this and I've spotted the problem there. It's that you used 60 research time. I was doing Follower Count research (because it's the only infinite 7 pack science). This has a research time of 30. That means my labs are theoretically capable of taking 176.4/min but are losing efficiency due to labs taking from other labs.

It also means that I need to add more labs so I can chew through the slower researches faster...