Question - Is Inserter item drop position intuitive?

[Jap2.0]

Either way, I vote to not change anything, as it does work in a predictable and consistent way.

Good joke!

No one managed to predict what U, V, W & X would look like. People constantly come up with wrong rules even for cases that they can check against the examples posted in the thread. It seems to be impossible to predict what will happen. You would have failed too if you didn't have the images to check before you posted.

For the record, I predicted W and X.

And yeah, csduff's "If curved, place in the lane that is more often farthest from the inserter" covers what I was attempting to get at fairly well (by the concept of the "far" lane, even in the first post, although at that point I simply generalized that concept as how it appeared in the examples given).

One thing to consider is that U and V, and W and X, are rarely used in real designs (which is used less is debatable), so even those who are familiar with inserter mechanics have little to no past experience with those configurations, and as such have no prior knowledge to use to base these patterns off of. People can predict the mechanics for the belt configurations they regularly use; if these "odd" configurations were used more often, it is likely that some people would have predicted them without the images being posted.

(also, posting 20 belt configurations does carry a bit of the implication that that is all of the possible configurations that need to be tested against.)

[coppercoil]

No one managed to predict what U, V, W & X would look like.

Please, don't forget me . I proposed a rule and submitted an illustration. The corrected rule also works for other cases.
If you think it's wrong, please argue.

[mrvn]

The behavior obviously can be predicted simply by learning all the cases. For me the bigger question is: Do the current rules make sense?

What breaks it for me is that with underground belts 3 inserters place on one side and only one on the other side. That simply isn't balanced and it would be real nice to have an underground belt going under a train stop and unloading with inserters on both sides. Currently that can only fill one side of the belt.

That the behavior also isn't intuitive only is one more reason to change it.

[Qon]

What I meant by prediction was your ability to predict which side of the belt the items would land on for cases you haven't yet seen also. Fewer and simpler rules are more intuitive and require less memorisation and possibly also a more balanxed behaviour.

[Qon]

No one managed to predict what U, V, W & X would look like.

Please, don't forget me . I proposed a rule and submitted an illustration. The corrected rule also works for other cases.
If you think it's wrong, please argue.

Sure, you will be remembered for getting U-X right and some other cases wrong. I will never miss another opportunity to mention your failures ever again, promise!

[coppercoil]

No one managed to predict what U, V, W & X would look like.

Please, don't forget me . I proposed a rule and submitted an illustration. The corrected rule also works for other cases.
If you think it's wrong, please argue.

Sure, you will be remembered for getting U-X right and some other cases wrong. I will never miss another opportunity to mention your failures ever again, promise!

Have you read the SECOND corrected definition? Please comment it too.

[csduff]

Either way, I vote to not change anything, as it does work in a predictable and consistent way.

Good joke!

No one managed to predict what U, V, W & X would look like. People constantly come up with wrong rules even for cases that they can check against the examples posted in the thread. It seems to be impossible to predict what will happen. You would have failed too if you didn't have the images to check before you posted.

Please tell me what part of my description was wrong?

And, when something does work in a consistent way, by definition, is predictable.

[Qon]

Please tell me what part of my description was wrong?

Your description was posted after U-X was posted. Joining the test after the answer sheet has been handed out earns you no credits.

And, when something does work in a consistent way, by definition, is predictable.

Wrong.

What I meant by prediction was your ability to predict which side of the belt the items would land on for cases you haven't yet seen also.

For each letter, the lane that the item lands on can be anything. It will be "consistent" as long as it doesn't randomly switch sides. But if there's no simple rule on what side the items land that works for all cases then you can't predict what side it will land on without testing that specific case.
Consistency in each case doesn't necessarily mean you have predictability from case to case. I defined predictability as the degree to which you can extrapolate.

[Qon]

Have you read the SECOND corrected definition? Please comment it too.

There's not much to say. I corrected the error in the first rule set so after that it has some of that Qon quality sprinkled on it

But the rules are still way more (irreducibly) Qonvoluted than "far right quadrant" would be and which would also give balanced results in the number of ways to output to each side.

[Jap2.0]

I defined predictability as the degree to which you can extrapolate.

A fine definition, but unusable in this case where there's a small number of possible cases. Here you can easily extrapolate to all cases, which makes it difficult to easily judge the "predictability."

[Qon]

The rules are Qonvoluted enough that I probably will forget some cases after some time of not using belts. And then the possibility to recreate that knowledge by extrapolating from the cases that I do remember would be helpful. But as the rules are now which side the items end up on in some cases is something that I would have to react to instead of making plans for when I'm making something non-trivial with belts.

I judge the predictability as "bad". And this thread proves it is so.

[coppercoil]

But the rules are still way more (irreducibly) Qonvoluted than "far right quadrant" would be and which would also give balanced results in the number of ways to output to each side.

Usability is more important than intuitive understanding. The curved belt case is not a rocket science and can be used without understanding, it costs ten seconds to try. I think current rule is useful, it follows the very common for the inserters perpendicular case. The player can change belt direction after row of inserters event without considering there is a special case. Isn’t it “intuitive”?
“Far right quadrant” will break this beautiful layout:

FarRight.png (719.03 KiB) Viewed 4935 times

I think current rule needs good image to explain. Where is the far side of the “horizontal” curved belt?

CurvedFarSide.png (393.98 KiB) Viewed 4935 times

Once you get it, you can’t forget it

[Qon]

Well my problem with the current system is this:


Also, what concrete mod are you using?

[coppercoil]

I'm not sure if I understand your problem. You'd like to fill top-right side of the belt on the corner?

Concrete came from the Dectorio.

[csduff]

Please tell me what part of my description was wrong?

Your description was posted after U-X was posted. Joining the test after the answer sheet has been handed out earns you no credits.

And, when something does work in a consistent way, by definition, is predictable.

Wrong.

What I meant by prediction was your ability to predict which side of the belt the items would land on for cases you haven't yet seen also.

For each letter, the lane that the item lands on can be anything. It will be "consistent" as long as it doesn't randomly switch sides. But if there's no simple rule on what side the items land that works for all cases then you can't predict what side it will land on without testing that specific case.
Consistency in each case doesn't necessarily mean you have predictability from case to case. I defined predictability as the degree to which you can extrapolate.

You failed to clarify what part of my description was wrong.

[Qon]

I'm not sure if I understand your problem. You'd like to fill top-right side of the belt on the corner?

The EC setup has 3 inserter positions that put items on the "left side" (right from the belts perspective) and only 1 that drops items on the other side (stack inserters). That is one problem I have with the current inserters.

And, when something does work in a consistent way, by definition, is predictable.

Wrong.

What I meant by prediction was your ability to predict which side of the belt the items would land on for cases you haven't yet seen also.

For each letter, the lane that the item lands on can be anything. It will be "consistent" as long as it doesn't randomly switch sides. But if there's no simple rule on what side the items land that works for all cases then you can't predict what side it will land on without testing that specific case.
Consistency in each case doesn't necessarily mean you have predictability from case to case. I defined predictability as the degree to which you can extrapolate.

You failed to clarify what part of my description was wrong.

What part of my clarification wasn't clear enough?

[coppercoil]

The EC setup has 3 inserter positions that put items on the "left side" (right from the belts perspective) and only 1 that drops items on the other side (stack inserters). That is one problem I have with the current inserters.

It's because the single rule "left/right" would be oversimplification. The curved belt is one part of the general context, so its rule must match it. Or, we can have a simplicity at the cost of usability. I made my choice.

[dog80]

edit:

ok i was just playing recently and i had just the feeling that after placing it i think when inserters where on the left and right and dropping onto corners of belt it felt un-intuitive...

so yeah maybe there is even more improvement possible

[mmmPI]

Once you get it, you can’t forget it

visual rule.png (204.48 KiB) Viewed 4832 times

JLJL.png (2.38 MiB) Viewed 4832 times

Now i look at the inserter that drop the green science in L8 and in L3, and i try to get it but it seems there something wrong.

In L3 if we consider the 'right' side from the belt perspective before the turn, we are supposed to consider it's a horizontal belt going to the west. It then becomes a vertical belt going north.

In L3 the inserter seems to drop AFTER the turn. So as in L8.

I am still confused, when i try to compare that rules with L6 or L1 for example where the 'visual rules' of before the turn applies.

At least i have this buffer thing sorted out now, even if i would probably fiddle a bit to remake one in a new game, i know i can make one and it's thanks to that topic.

Still found it 'illogical' with character, have not been convinced.

[mrvn]

Once you get it, you can’t forget it

visual rule.pngJLJL.png

Now i look at the inserter that drop the green science in L8 and in L3, and i try to get it but it seems there something wrong.

In L3 if we consider the 'right' side from the belt perspective before the turn, we are supposed to consider it's a horizontal belt going to the west. It then becomes a vertical belt going north.

In L3 the inserter seems to drop AFTER the turn. So as in L8.

I am still confused, when i try to compare that rules with L6 or L1 for example where the 'visual rules' of before the turn applies.

At least i have this buffer thing sorted out now, even if i would probably fiddle a bit to remake one in a new game, i know i can make one and it's thanks to that topic.

Still found it 'illogical' with character, have not been convinced.

The turns parts I think where best described by considering belts going past the inserters and turning before/after the inserters. And then make it one shorter. The placement doesn't change by making it shorter.