Friday Facts #137 - The release scarecrow

[Fatmice]

Hold off on making that graph until we have 0.13. =)

[RobertTerwilliger]

This just seems a bit strange to me:

graph

Fast inserters will be slower than burner inserters are now.

Fast inserter will be slower than burner was before only in direct inserting.
The idea of R.Inserter is it will be used in high throughput item flows no matter belt-based, or direct. It doesn't mean to be something new with uber-high throughput, it is designed to make all inserters more specialized. Nothing strange in other inserters becoming slower - you will just use rapid one instead, where you need it. As Adil have wrote before, you don't really use stack bonus in your whole factory, only in some blocks, which can use rapids so easily, that I can't even tell)

[ratchetfreak]

This just seems a bit strange to me:

graph

Fast inserters will be slower than burner inserters are now.

Fast inserter will be slower than burner was before only in direct inserting.
The idea of R.Inserter is it will be used in high throughput item flows no matter belt-based, or direct. It doesn't mean to be something new with uber-high throughput, it is designed to make all inserters more specialized. Nothing strange in other inserters becoming slower - you will just use rapid one instead, where you need it. As Adil have wrote before, you don't really use stack bonus in your whole factory, only in some blocks, which can use rapids so easily, that I can't even tell)

indeed, nearly everywhere except bot builds inserters are used to pick up from and put on to a belt where the fast inserter second fastest way (the fastest way takes up a ton of space and involves several chests in a row where the items will get hopped across over and several inserters putting on the belt).

[AutoMcD]

i can't think about a rapid inserter would be a bad addition

It's not, just most of us don't want to see the other inserters nerfed so badly.

[Qon]

This just seems a bit strange to me:

graph

Fast inserters will be slower than burner inserters are now.

Fast inserter will be slower than burner was before only in direct inserting.
The idea of R.Inserter is it will be used in high throughput item flows no matter belt-based, or direct. It doesn't mean to be something new with uber-high throughput, it is designed to make all inserters more specialized. Nothing strange in other inserters becoming slower - you will just use rapid one instead, where you need it. As Adil have wrote before, you don't really use stack bonus in your whole factory, only in some blocks, which can use rapids so easily, that I can't even tell)

Of course I can replace the fast ones with rapids. But I use direct insertion a lot, not just in a few parts of the factory. The point wasn't that there would be no fast insertion method, but that the inserter upgrade curve looks silly. If you use direct insertion everywhere then fast inserters will be slower than burners are now. The throughput gap between a 1 stack inserter and 12 stack inserter is huge. To me it would make rapid the only choice. Which is ok, no matter the cost my factory will be able to mass produce them quickly at the point I am at now. But I would prefer if fast and long continued to work exactly as they do now. If the stack bonus is removed from basic and burner inserter then yeah that is fine with me.

I'm not saying the the graph shows that something is wrong, just that it shows something strange and that a bit more thought should be put into it before it is decided. If you can motivate why the red curve is superior to the blue curve then I'm ok with it. But otherwise it seems like someone just tried to balance it without doing the math and ended up with something unintentionally skewed.

indeed, nearly everywhere except bot builds inserters are used to pick up from and put on to a belt where the fast inserter second fastest way (the fastest way takes up a ton of space and involves several chests in a row where the items will get hopped across over and several inserters putting on the belt).

I use a lot of bots. But I use direct insertion where I don't use bots also.

[RobertTerwilliger]

inserter upgrade curve looks silly.

Upgrade curve? No upgrade, just different tasks. If there were few medium throughput inserter stages - would you use them? I very doubt so.
- you use burner inserter where you have no power
- you use regular inserters for economy in slow machines
- you use long inserters for reach
- you use fast inserters in throughputs where regular is not enough (very roughly x3)
--- do you need more stages between?
- you will use rapid in extra-high throughputs (nearly x10 from fast)

[Neotix]

inserter upgrade curve looks silly.

Upgrade curve? No upgrade, just different tasks. If there were few medium throughput inserter stages - would you use them? I very doubt so.
- you use burner inserter where you have no power
- you use regular inserters for economy in slow machines
- you use long inserters for reach
- you use fast inserters in throughputs where regular is not enough (very roughly x3)
--- do you need more stages between?
- you will use rapid in extra-high throughputs (nearly x10 from fast)

It will turn into:
- you use burner inserter where you have no power
- you use long inserters for reach (but you will still avoid situations where you need them)
- you will use rapid inserter everywhere else because other inserters are such inferior that are not worth using.

For me there shouldn't be superion/inferion inserter. All inserters should be equal if some way but with different features like burner don't need electricity, long can reach further, smart can be connect with logic network, rapid can pick and place stack items on belts. Right now normal and fast are inferior version of rapid without any feature. Devs could change them into one inserter that have feature like faster direct insertion. In that way we don't have to choose between slower/faster but choose based on theirs functions.

[eloy2030]

inserter upgrade curve looks silly.

Upgrade curve? No upgrade, just different tasks. If there were few medium throughput inserter stages - would you use them? I very doubt so.
- you use burner inserter where you have no power
- you use regular inserters for economy in slow machines
- you use long inserters for reach
- you use fast inserters in throughputs where regular is not enough (very roughly x3)
--- do you need more stages between?
- you will use rapid in extra-high throughputs (nearly x10 from fast)

It will turn into:
- you use burner inserter where you have no power
- you use long inserters for reach (but you will still avoid situations where you need them)
- you will use rapid inserter everywhere else because other inserters are such inferior that are not worth using.

For me there shouldn't be superion/inferion inserter. All inserters should be equal if some way but with different features like burner don't need electricity, long can reach further, smart can be connect with logic network, rapid can pick and place stack items on belts. Right now normal and fast are inferior version of rapid without any feature. Devs could change them into one inserter that have feature like faster direct insertion. In that way we don't have to choose between slower/faster but choose based on theirs functions.

Thats an excellent way of thinking! Thats what makes (for me) the difference between a good game (chess) and other games (not chess) Thats the rock paper scissors concept and for me its the only way to go...

cheers!

edit: without that its just brute force. can i build bigger?... then i build bigger/faster/stronger... i like smarter!

[Nemoricus]

Regular inserters also have a niche in this system, for those situations where throughput doesn't matter and you have power.

[HulkingUnicorn]

It will turn into:
- you use burner inserter where you have no power
- you use long inserters for reach (but you will still avoid situations where you need them)
- you will use rapid inserter everywhere else because other inserters are such inferior that are not worth using.

Actually, with the rapid inserters waiting for full capacity in its hand (max stack bonus), you'd prefer slower inserters to maintain throughput in low-thoughput situations (unless there's something in today's friday facts where this is changed).

[Neotix]

Thats why i mentioned about changing normal/fast inserter by one inserter that don't have feature like rapid inserter but different feature like fast direct insertion (double stack bonus). It can be used with belts like normal inserter (it don't wait for full stack).

[RobertTerwilliger]

Thats why i mentioned about changing normal/fast inserter by one inserter that don't have feature like rapid inserter but different feature like fast direct insertion (double stack bonus). It can be used with belts like normal inserter (it don't wait for full stack).

It's completely legit suggeation. Probably fast inserter may be just an upgrade for regular one. Or may not, considering we still use all 3 belt tiers. For example, unmoduled smelter uses regular inserters, also production of long-crafting items doesn't need fast inserters, so regular inserters will still have it's niche.

[iccle]

Railtrack additions look great it would be nice as suggested, if the signals also showed an outline of carrages on the track so that you can space them out appropriately for your size of trains.

I don't really understand or see the need for why the introduction of this new inserter should conincide with a nerf to all the others. It looks like a very niche (though neccesary) use inspired it's creation (fast unloading from a container onto a belt in a more condensed fashion and vice versa) yet the rest of the game is being altered to make this the only option for container to container inserts that relied on the previous inserter stack bonus (feeding 20 green chips to purple chip assembler springs to mind). I would also suggest that 'Buffered Inserter' would be a more apt name for it and avoids similarity with the 'Fast inserter'.

If i understand it correctly we will now have several inserters that can only insert or extract 1 item, and 1 inserter that will always wait around for 12 items (or wherever you've researched the stack bonus up to?) before inserting? When it inserts items will it always insert 12 or will it insert partial amounts then stall to pick enough items to make it up to 12 again, or will it overfill machines with 12 items at a time ?

Given that smart inserters are being dropped and the circuit/logistics network is being unified and applied to all chests/inserters/belts etc, will this new inserter also be able to be 'programmed' by a circuit network condition to limit the amount of items it waits for or inserts? If yes then why would we choose any other inserter? If no then why reduce the capacity of the others down to one item only?

The number 12 bothers me because the most common stack sizes (50,100,200) are not divisible by it without remainder meaning up to 11 items for each 'Rapid Inserter' left hanging in the air instead of delivered. In game I would be likely to choose to use this new inserter for unloading trains or buffers to belts and over many deliveries the withheld items would eventually get there, however with the stalling 'feature' and stack size of 12 the inserter will always stall as it gets near to the bottom of the last stack remaining, this complicates doing things in circuit networks like checking the total number of items in the inital buffer and comparing that to how much has been unloaded since you would have to account for how many inserters will stall per N stacks of product in the container.

So in the use case where the objective is to completely empty one container into another or onto a belt eg train unload, these new inserters if used will always stall on the last few items in a stack unless all stacks are a multiple of 12 (or the researched stack bonus), the old inserters will be able to completely empty and deliver the first container but will take up to 5x longer to do so or 5 times as many inserters than it did previously, really cant grasp the why of this change.

If the new inserter were introduced on its own with no changes to the previous inserters I wouldn't really mind since I could take it or leave it, I would prefer it had a timeout though so that delivery of held items is always guaranteed if there is space in the target container.

[ratchetfreak]

I don't really understand or see the need for why the introduction of this new inserter should conincide with a nerf to all the others. It looks like a very niche (though neccesary) use inspired it's creation (fast unloading from a container onto a belt in a more condensed fashion and vice versa) yet the rest of the game is being altered to make this the only option for container to container inserts that relied on the previous inserter stack bonus (feeding 20 green chips to purple chip assembler springs to mind). I would also suggest that 'Buffered Inserter' would be a more apt name for it and avoids similarity with the 'Fast inserter'.

Putting on belts and taking off them is all you do if you play factorio as advertised. That the stack size bonus does not work with belts but only from inventory to inventory is often a big surprise for new players. This surprise is a bad thing when talking about the game design; it's not inuitive, it's poorly explained, you are artificially limited in what you can do with it.

now we get a inserter that will use the bonus on every operation it can do.

[AutoMcD]

makes sense from that perspective

[iccle]

I don't really understand or see the need for why the introduction of this new inserter should conincide with a nerf to all the others. It looks like a very niche (though neccesary) use inspired it's creation (fast unloading from a container onto a belt in a more condensed fashion and vice versa) yet the rest of the game is being altered to make this the only option for container to container inserts that relied on the previous inserter stack bonus (feeding 20 green chips to purple chip assembler springs to mind). I would also suggest that 'Buffered Inserter' would be a more apt name for it and avoids similarity with the 'Fast inserter'.

Putting on belts and taking off them is all you do if you play factorio as advertised. That the stack size bonus does not work with belts but only from inventory to inventory is often a big surprise for new players. This surprise is a bad thing when talking about the game design; it's not inuitive, it's poorly explained, you are artificially limited in what you can do with it.

now we get a inserter that will use the bonus on every operation it can do.

I get that the stack bonus did not apply to belts and also that it may have been confusing for some, however I dont see how having this change eases anything its essentially swapping one caveat for another since it behaves differently to all other inserters, would it not have been simpler to make the existing inserters transfer up to 5 items to/from belts then the stack bonus is applied uniformly?

This inserter is fundamentally a different kind from the other inserters because it's buffered and stalls until it's buffer is full. If it did not stall (say a timeout after 0.5s of no items for pickup) then I would have no issue and would welcome the change to all the inserters but as it stands It is not a direct equivalent replacement for the functionality we had before, so now the inserter stack bonus doesn't actually apply to any inserters except this one which behaves differently from all other inserters which I think is just as likely to be unclear to new players as the previous situation.

Also taking things on and off belts is not all you do, you also transfer things from carrages to chests, from chests to machines from machines to chests, bot factories tend to use very few belts at all, now whenever you want to do these things and you must transfer more than 1 item at a time you have an overhead of the buffer inside the inserter itself for each inserter used, consider making purple chips you could serve green chips using two 'rapid inserters' if each inserter only receives 11 chips they both stall, yet combined they hold enough to complete the process with two to spare, this is not an issue for machines first in line, but further down the belt and the more imbalanced it becomes the worse it gets.

[RobertTerwilliger]

it's not inuitive, it's poorly explained

Disagree. It is explained good enough.
Dunno in your localization - I play English version, living in non-english speaking country. The only thing in whole game, that was not completely clear for me is difference between active and passive provider chests, but after a short test it also became 100% clear.
This is kind of engineering game - reed manual (aka tooltips) carefully, and you'll be okay.

Intuitive? Never had problems with inserters working not like expected - see above :/

[ratchetfreak]

If there is a name change to the research like "rapid inserter stacksize bonus" and have the tooltip on the rapid inserter say that it will pick up multiple items and initially have a stack size of 2 (or more) then there will need to be no additional explanation.

Everything is clear from the get go; rapid inserter is the one marked to pick up multiple items per movement; research only mentions rapid inserter; normal inserters stay the same.