Talk:Inserters: Difference between revisions

From Official Factorio Wiki
Jump to navigation Jump to search
← Older edit
Tag: Reverted
→Splitter to Chest: new section
 
(5 intermediate revisions by 3 users not shown)
Line 28: Line 28:
:I'd describe extension speed as how fast the inserter "extends", so in this case how fast its arm gets longer/shorter to be able to reach an item on the far or near side of the belt. It's expressed in tiles per tick because that is a unit of speed, just like meters per second :) An inserter that has an extension speed of 1 tile per tick needs 1 tick to make its arm longer by a whole tile. I hope this helps. -- [[User:Bilka|Bilka]] ([[User talk:Bilka|talk]]) - <span style="color:#FF0000">Admin</span> 11:11, 6 March 2018 (UTC)
:I'd describe extension speed as how fast the inserter "extends", so in this case how fast its arm gets longer/shorter to be able to reach an item on the far or near side of the belt. It's expressed in tiles per tick because that is a unit of speed, just like meters per second :) An inserter that has an extension speed of 1 tile per tick needs 1 tick to make its arm longer by a whole tile. I hope this helps. -- [[User:Bilka|Bilka]] ([[User talk:Bilka|talk]]) - <span style="color:#FF0000">Admin</span> 11:11, 6 March 2018 (UTC)


== Recent Power Usage Edit ==
== Multiplication sign notation ==


The Power Usage section was recently edited by user ''Queithe1E'' in a way I find to be incorrect and misleading. I'll break down some of the problems I have with the edit and why I think they're problematic.
To Cardinal, why are you changing all the multiplication signs from " * " to " x "? This is less proper notation mathematically, and can cause confusion algebraically which is why the * notation exists. I don't see any benefit to explicitly make this change which is arguably less correct.{{Unsigned|GregFirehawk|09:08, 10 May 2025}}


Firstly large swaths of context were removed and branded as "speculation". Those are actually very important contextually as inserter behavior is unusual and not what you'd expect. I'm okay with the section being rewritten for the sake of brevity (in fact I'd probably recommend it as I was perhaps too verbose originally), but only if it's concise and accurate.
: The wiki generally prefers × and ÷ for mathematical formulas, so these edits are correct. For completely proper mathematical notation, you can use latex instead, there is a proposed style guide: [[Factorio_talk:Style_guide#Draft:_equation_style_guide]]. -- [[User:Bilka|Bilka]] ([[User talk:Bilka|talk]]) - <span style="color:#FF0000">Admin</span> 09:19, 10 May 2025 (UTC)


Which segways directly into my second point, which is that the new information that took its place seems to show a misunderstanding of the original information, and is incorrect at times. For example at one point it says "The maximum power consumption as shown in an inserter's tooltip assumes that both types of movement take place at the same time", but that's not actually true, and the previous explanation explicitly communicated a completely different idea. Inserters maximum power usage is absolutely not the sum of rotational and lateral movement costs being expended simultaneously, and there's absolutely nothing to even suggest that might be the case anywhere, neither in the games data or the empirical data that was gathered.
:: Well, the char I've used was "×", not "x". Those edits are basically for 2 reasons: Firstly, as Bilka said, to maintain the consistency of wiki contents. Secondly, to avoid ambiguity from the "*" representing the annotation, you can see them in the tables on this page. Of coures, sometimes "x" can also be confused with "×", maybe it's a better idea to reform the equation with LaTeX syntax. Sadly I'm not good at it, so if you can deal with this, I'll glad to apply it in my translation. [[User:Cardinal|Cardinal]] ([[User talk:Cardinal|talk]]) 09:55, 10 May 2025 (UTC)


And that once again segways neatly into my next problem which is I find the inclusion of the games lua data to be misleading and problematic, with the reason being that it's not actually accurate. Empirical testing shows the real values are very different from what the game claims in its lua data, which is something I went to great length to identify, explain, and quantify. The inclusion of the games lua data in such a position of prominence right at the top of the section, and with absolutely no disclaimer, is very misleading due to the discrepancy with actual measured data.
== Splitter to Chest ==


There's also some other information that I don't really know what to make of. They mention part of the inserters initial power draw is used filling an energy buffer. I'm not aware that electrical inserters even have an energy buffer, as far as I'm aware that's just the burners, and how he can confidently attribute this power draw to this buffer is anyone's guess. It's really just idk, it could be right but I have no clue what he's basing this on. Also later he does an equation seemingly based off the graph and the lua data which does seem to check out, but it's completely at odds with the empirical data that was collected. This actually made me question my sanity so I went back and reviewed my original data and confirmed that my findings ultimately also agreed with the games power panel when averaged over time (this is cited in my forum post as 8.8kw of operational cost for red inserters during testing), and combined with the fact that it's a larger data set with more stringent analysis I'm more inclined to believe my tests accuracy over a low detail in game graph and some numbers I know to be inaccurate.
I noticed that there is Chest to Splitter, but no Splitter to Chest. I clocked it at ~20 item/s using blue splitters and green inserters, and at over 17 item/s with red splitters. That's pretty significant, much more so than the sideloaded tunnel strat thats prominently featured. For reference, I did this by doing 1 inserter per splitter, and loading it with both belts. There is a drop in throughput when using only a single belt, and another smaller drop when using 2 inserters per splitter. I'm pretty sure I'm not the first to come across this as I think I've heard people discuss it before, so it should be added to the page. I have no plans to do that right now, so if anyone else would like to then please do!
 
Also as I was reviewing this before hitting publish having already written everything further down I realized even the subject of the above paragraph is in self contradiction, because if you apply the equation he gives to extension/retraction using the value from lower down on the wiki you end up with ''0.0457 * 5 * 60'' which equals just 13.71. If you add drain that becomes 14.11, so it seems like the internal buffer he alleges is the remaining 0.89. But then he turns around and I guess to get that 21 max power equals these two numbers he takes 15 and adds the 6 he calculated, meaning he essentially adds this buffer as a second source of drain or something during extensions or during peaks? It just doesn't add up, it's very strange. I pulled that 0.0457 from further down on the wiki so it's possible that number is actually approximately 0.003 too low, in which case his formula for max power would add up properly, but of course then his buffer theory would disappear. Anyway remember the paragraph below this is meant to follow the paragraph directly above this one.
 
And this is another important problem with the change, because now the entire section feels incoherent and self contradictory. It was specifically written to show these empirically derived numbers that differ from what the game claims, but now the original game numbers are present and conflicting with and undermining the empirical data, which most importantly is more accurate.
 
And I could keep nitpicking about little stuff or how the little context around the burner was removed, but ultimately I think I made my point. The main problems with the change are that it says certain things at are provably wrong, removes certain important context, and ultimately confuses the issue by prominently asserting a separate data set that was tested to be inaccurate, and is in direct contradiction with the data set directly below it. I'm gonna be rolling back the change for these reasons, since I believe the prior version was more accurate and less confusing. If you disagree or have questions please reach out as I'm happy to discuss the topic further
 
--[[User:GregFirehawk|GregFirehawk]] ([[User talk:GregFirehawk|talk]]) 01:01, 13 April 2025 (UTC)

Latest revision as of 01:42, 5 August 2025

[1]

Removed. But eventually helpful. Ssilk (talk) 02:45, 31 December 2013 (CET)

Changes for 0.17

The faster belt speeds probably change the results for some of the entries in these tables. It looks like they haven't been updated yet. --Omnifarious (talk) 14:59, 27 May 2019 (UTC)

They have been updated by Bilka in This revision. --Swan (talk) 17:19, 27 May 2019 (UTC)

Inserter speed

I updated this wiki to show ticks per turn in the inserter speed section. My results seem to clash with the previous information about rotation speed (102 ticks = 1.7 seconds, there is a discrepancy between the rotation time of burner inserters most notably). The experiment I performed was done on vanilla factorio 0.14.22. I used two steel chests with varying inserters for each experiment. The inserters were fully powered at all times. The first chest is wired to a combinator which tests for any items in the chest, and outputs 1 signal X. This signal is added to a tallying arithmetic combinator. This arithmetic combinator is wired to itself, and outputs T + X. The value of T at setup is 0. I begin this experiment by placing large stacks of items into the first chest. I place only 201 transfers worth of material into the chest. This will stop the timer as soon as the last item is removed, which occurs after 200 transfers. With this setup I obtained the results as I have reported them. The timer always stopped on a multiple of 200, indicating that the number of ticks per full turn is an integer number. This is why I decided to add a separate column for this value. I may in the future update the other rotation speed metrics to match my results. --Reububble (talk) 18:19, 8 March 2017 (UTC)

Which orientation were the inserters? Inserters facing north have lower throughput, might mean they have slower rotation speed. --Artorp (talk) 20:47, 5 April 2017 (UTC)

put the "Inserters And Transport Belt Interaction" examples in a table so they are side by side

i see why that was a bad idea, but i would rather see them side by side if it were possible to make a table that, when the horizontal space it has to display is smaller than all 4 makes it so there are 3 and on another "line" another 1 or 2x2 or all underneath one another like it is now.

I put them in a 2x2 square; I think that's the best way to see all of them at once. -- Bilka (talk) - Admin 13:17, 14 August 2017 (UTC)

Problem with translation

Could someone explain to me what it is Extension Speed and why it is expressed in Tiles per Tick.

I'm trying to make a translation and can not find a word - without knowing what it is ;)— Preceding unsigned comment added by JakubSTR (talk • contribs) 08:43, 6 March 2018‎

I'd describe extension speed as how fast the inserter "extends", so in this case how fast its arm gets longer/shorter to be able to reach an item on the far or near side of the belt. It's expressed in tiles per tick because that is a unit of speed, just like meters per second :) An inserter that has an extension speed of 1 tile per tick needs 1 tick to make its arm longer by a whole tile. I hope this helps. -- Bilka (talk) - Admin 11:11, 6 March 2018 (UTC)

Multiplication sign notation

To Cardinal, why are you changing all the multiplication signs from " * " to " x "? This is less proper notation mathematically, and can cause confusion algebraically which is why the * notation exists. I don't see any benefit to explicitly make this change which is arguably less correct.— Preceding unsigned comment added by GregFirehawk (talk • contribs) 09:08, 10 May 2025

The wiki generally prefers × and ÷ for mathematical formulas, so these edits are correct. For completely proper mathematical notation, you can use latex instead, there is a proposed style guide: Factorio_talk:Style_guide#Draft:_equation_style_guide. -- Bilka (talk) - Admin 09:19, 10 May 2025 (UTC)
Well, the char I've used was "×", not "x". Those edits are basically for 2 reasons: Firstly, as Bilka said, to maintain the consistency of wiki contents. Secondly, to avoid ambiguity from the "*" representing the annotation, you can see them in the tables on this page. Of coures, sometimes "x" can also be confused with "×", maybe it's a better idea to reform the equation with LaTeX syntax. Sadly I'm not good at it, so if you can deal with this, I'll glad to apply it in my translation. Cardinal (talk) 09:55, 10 May 2025 (UTC)

Splitter to Chest

I noticed that there is Chest to Splitter, but no Splitter to Chest. I clocked it at ~20 item/s using blue splitters and green inserters, and at over 17 item/s with red splitters. That's pretty significant, much more so than the sideloaded tunnel strat thats prominently featured. For reference, I did this by doing 1 inserter per splitter, and loading it with both belts. There is a drop in throughput when using only a single belt, and another smaller drop when using 2 inserters per splitter. I'm pretty sure I'm not the first to come across this as I think I've heard people discuss it before, so it should be added to the page. I have no plans to do that right now, so if anyone else would like to then please do!

Retrieved from "https://wiki.factorio.com/index.php?title=Talk:Inserters&oldid=215218"