Rail stack size

[Mmmmmmorten]

Is it just me that feel rails suddenly has become the bottleneck on the science build?
Pumping out enough rails to Production science actually was an issue...
As it is now it kind of has to be made on site due to stack size limits on stone and rails...

With 0.17.x - should rails be stack-able in 200 size? (double what it is now)

/old dude

VIDEO MELIORA, PROBOQUE, DETERIORA SEQUOR.

[zOldBulldog]

I didn't notice any issues myself with the science as the supplies for it come on a belt or train, but I do agree that it would be nice to have much bigger rail stack size (same with landfill) as it is a nuisance to have to have multiple wagonfulls with you every time you go out to build rail. I'd be quite happy with rail stack sizes of 500 or even 1000 (no I don't expect it to happen but I can dream, right?)

[Mmmmmmorten]

Yes... deaming, and posting about them here.
Landfill is a good point.
Thankfully not as big an issue this time as I had really bad dreams about landfill from 0.16, and now use the beautiful_bridge_railway mod to run my trains over waters. Only landfill I do now is when there is water where I'm building - and that is more manageable.

But, when we are on dreams - lot's of stuff that could have their stack sizes increased a bit (belts - express as a minimum, robots (both types), all inserters and finally roboports and beacons).

/old dude
ORA ET LABORA

[zOldBulldog]

Yes... deaming, and posting about them here.
Landfill is a good point.
Thankfully not as big an issue this time as I had really bad dreams about landfill from 0.16, and now use the beautiful_bridge_railway mod to run my trains over waters. Only landfill I do now is when there is water where I'm building - and that is more manageable.

But, when we are on dreams - lot's of stuff that could have their stack sizes increased a bit (belts - express as a minimum, robots (both types), all inserters and finally roboports and beacons).

/old dude
ORA ET LABORA

I agree on your new dream additions.

I might add that robots in roboports should probably have their own special slots, not consume regular inventory. Not that it makes a big dent, but it would seem to make more sense logically, plus it would make it easier to release the excess.

BTW... Also an old dude, been gaming and designing software since the days of the original Pong. It is nice to meet a fellow gaming ancient one.

I will have to grab Beautiful bridges. Makes too much sense to ignore.

[Koub]

[Koub] Moved to Balancing.

[eduran]

Agreed on rails needing a larger stack size. Luckily, mods allow you to do that. I suggest ReStack, as it let's you configure rail stack size (and a lot of other items) to whichever values you prefer.

[Mmmmmmorten]

Agreed on rails needing a larger stack size. Luckily, mods allow you to do that. I suggest ReStack, as it let's you configure rail stack size (and a lot of other items) to whichever values you prefer.

LOL!
I was so in the mindset of fixing things in 0.17, that I thought was lacking after changes, that I did not even consider looking for a mod
And that looks really, really interesting...

Thanks Eduran!

/old dude

QUIDQUID LATINE DICTUM SIT, ALTUM VIDETUR.

[MeduSalem]

The rail stack size is fine... what is ridiculous is the amount of rails required for the productivity science pack recipe.

It would be fine if it was only like 10 Rails per pack (or something else entirely), but 30 rails per pack is total overkill. It outdoes everything else the science packs need multiple times over. A true logistics nightmare because even bots have trouble to keep up with it.

Currently it alone accounts for 20-30% (depending on productivity modules etc) of the entire Iron Ore/Plate and Steel Plate requirement for the entire science production (including all packs, also building the Rocket/Space Science)... all other packs don't have such ridiculous requirements.

Then again I was against this stupid change in advance anyway, I already knew it would cause trouble. It almost makes me want to cheat using the Creative Mod and just filling the stupid Assembler from an infinite chest of rails because I don't like it.


TLDR; the Productivity Science Pack recipe is just a steaming pile of unbalanced shit. As simple as that.

[zOldBulldog]

The rail stack size is fine... what is ridiculous is the amount of rails required for the productivity science pack recipe.

It would be fine if it was only like 10 Rails per pack (or something else entirely), but 30 rails per pack is total overkill. It outdoes everything else the science packs need multiple times over. A true logistics nightmare because even bots have trouble to keep up with it.

Currently it alone accounts for 20-30% (depending on productivity modules etc) of the entire Iron Ore/Plate and Steel Plate requirement for the entire science production (including all packs, also building the Rocket/Space Science)... all other packs don't have such ridiculous requirements.

Then again I was against this stupid change in advance anyway, I already knew it would cause trouble. It almost makes me want to cheat using the Creative Mod and just filling the stupid Assembler from an infinite chest of rails because I don't like it.


TLDR; the Productivity Science Pack recipe is just a steaming pile of unbalanced shit. As simple as that.

It does seem strange and unbalanced. On the other hand... I might be wrong but ores feel more abundant (or at least less scattered) in 0.17. That makes up for a lot, unless you are counting pollution.

[MeduSalem]

It does seem strange and unbalanced. On the other hand... I might be wrong but ores feel more abundant (or at least less scattered) in 0.17. That makes up for a lot, unless you are counting pollution.

Well, the ore consumption isn't the problem in my opinion because they are produced and transported around in bulk anyway (using trains etc). If the amount of Iron Ore/Plates would be better distributed over the science packs then why not.

The problem I have with 30 rails per pack is that it bloats the entire consumption artificially for no reason and makes the logistics of supplying the Production Science Pack properly a problem.

I played around with the Creative mod a few days ago, trying to test a 4000SPM base in a sandbox and needing to produce almost 30 000 rails/min is absolutely ridiculous. No other finished product even compares to it remotely (others being located in the 500-2000 Units/min range for a 4000SPM base). The pack consumes Rails as if it was an intermediate product, just that it isn't.

Even the production chart it shows ridiculously far ahead of everything else (except pure Plates/Ore production, but which are obviously also inflated due it stemming from the rails)... making seem everything else a cakewalk in comparison.


Also I am still in the process of figuring out a way to circumvent the logistics bottleneck involved with the rails. Trying different direct-insertion setups with the Iron Sticks, Rails, Science Pack etc... all to no avail, there is not even a good ratio for any of them to make sense either.


When the devs came up with this recipe I don't know how they didn't notice that this feels unbalanced. No matter how you are going to supply the Production science pack... the logistics are bloated, even more so if you are doing by belts, unreasonably bloated in comparison to the other packs.

[Mmmmmmorten]

I played around with the Creative mod a few days ago, trying to test a 4000SPM base in a sandbox and needing to produce almost 30 000 rails/min is absolutely ridiculous. No other finished product even compares to it remotely (others being located in the 500-2000 Units/min range for a 4000SPM base). The pack consumes Rails as if it was an intermediate product, just that it isn't.

Also I am still in the process of figuring out a way to circumvent the logistics bottleneck involved with the rails. Trying different direct-insertion setups with the Iron Sticks, Rails, Science Pack etc... all to no avail, there is not even a good ratio for any of them to make sense either.

When the devs came up with this recipe I don't know how they didn't notice that this feels unbalanced. No matter how you are going to supply the Production science pack... the logistics are bloated, even more so if you are doing by belts, unreasonably bloated in comparison to the other packs.

Totally agree, that is why I brought this topic up. I know there are mods out there to fix stack sizes etc., but this just felt unbalanced in the 0.17.x experimental release...

I'm trying for 4.800 SPPM now and my Production (pink) Science will need 35.000 rails. Fine that it's 2 every 0.5 seconds, but with 10 beacons (dual build with iron stick) it's 40 rails per second per machine!
Getting all this - properly stacked - onto belts and into Production science I have no idea how I will do. We are talking 14 express belts for rails going full out too feed Production science machines! I had trouble enough in my starter city.

As to dev's. Have a feeling this was the solution to all the cuts done in Iron plate consumption overall. They beefed up Copper consumption in Space science (Low Density structure), but needed something to use Iron plate. Spitting Steel plate at Production science certainly keeps the Iron plate consumption up (even though it's total down in 0.17).

/old dude

[leadraven]

TLDR; the Productivity Science Pack recipe is just a steaming pile of unbalanced shit. As simple as that.

1. 30 rails per !3! science packs.
2. Here is detailed comparison of science packs prices : https://www.factorio.com/blog/post/fff-275
3. I also don't like this recipe very much, but completely for other reasons : viewtopic.php?f=6&t=65846

[MeduSalem]

TLDR; the Productivity Science Pack recipe is just a steaming pile of unbalanced shit. As simple as that.

1. 30 rails per !3! science packs.
2. Here is detailed comparison of science packs prices : https://www.factorio.com/blog/post/fff-275
3. I also don't like this recipe very much, but completely for other reasons : viewtopic.php?f=6&t=65846

Doesn't matter at all if it is 30 rails per 3 packs. The recipe still sucks big time. They should use a resource item that condenses the amount of iron down way more instead of using a low-cost item like rails and blowing the amount required up to extreme proportions this late in the crafting cascade.

I know they needed some stupid recipe that requires stone just for the sake of it, but it renders the logistics for Production Science absurd.

Like Mmmmmmorten wrote above... for 4000+SPM you need already 10+ full express belts just of stupid rails to the center of the base where you produce the science packs. That is what you normally need for second-level intermediate stuff.

My entire base is bottlenecked like hell because of this stupid rails and I can't do anything about it... There is no way I am going to cram 10 belts of rails in there when it should be only like 1-2 belts just like everything else this far up the crafting cascade.

I habe more than 10000 bots dedicaded hauling just stupid rails. That is almost half of what ALL other science packs require combined.


[Edit] But then again forget about it.

I didn't play for more than a year (last time when 0.16 came out) ... and now I am pretty much done with 0.17 too... and all I can say it is that it's the worst release so far with tons of abritrary changes and other weird kinks or bugs I can't wrap my head around since I am too used to the older versions... and all that is left is nitpicking on stuff that is pretty much just pissing me off, like the unbalanced recipe for example or that my newly generated map has become virtually unplayable within a few days with all the drops in UPS since everything that was perfectly calculated is slowly going out of sync and everywhere arise bottlenecks where there where there theoretically shouldn't be any because there are way more assemblers and stuff than necessary, but still bottlenecks out of nowhere due to lag and the game probably skipping update cycles or whatever and everything becoming more and more inaccurate. Either that or it's a bug or I don't know, at least the game is starting to behave weird and I don't feel like dealing with it any further.

Guess I will call it a day and maybe check back when the final version is out.

[Zavian]

10+ full express belts just of stupid rails to the center of the base where you produce the science packs.

You don't need to put the rails on belts.

I habe more than 10000 bots dedicaded hauling just stupid rails. That is almost half of what ALL other science packs require combined.

10000 bots hauling high volume stuff isn't really a solution. Just thinking about all the roboports needed to charge that many bots makes me cringe. Especially since it sounds like you already have another 20000 bots also hauling stuff.

My entire base is bottlenecked like hell because of this stupid rails and I can't do anything about it...

So design a different solution. I find that direct insertion works well for production science. eg something like

beaconed.prod.science.jpg (831.7 KiB) Viewed 6956 times

For me solving those sort of production/logistics issues is a large part of the core gameplay loop of Factorio. If your existing approaches aren't working, then sometimes you just need to find a different approach.

[Mmmmmmorten]

Been sitting with my nose down in excel, building test modules in my Inner City Test site (I test concept and math in game only) and think I have come up with a design I can live with. Will try to build it in the coming days and see how it does in live test.
Only use calculators, Excel and Max Rate Calculator and a lot of hands on testing to see if my ideas will work or not.

Now I think I have a 12 beacon Production science (pink stuff) design that I think might work. 6 assemblers in each column with Electric furnace & Productivity module 1 fed on same belt from the side, and two full belts (only 40+ rails) fed from bottom. Should give 560 science out per min so I need 9 of these for 4800 SPPM build I'm doing.

Production Science (pink) column design

Factorio 0.17.28 Production (pink) science build - test2.png (617.04 KiB) Viewed 6949 times

Too support this with Rail's I build it on site with a 14 beacon build that has Iron sticks direct feed into each Rail assembler. The setup produce 40 rails pr second, and is enough for one belt feeding the Production (pink) columns I designed.
Build as block with 8 of these modules per block with Rail belts (40) going out on each side. Feeding from the bottom 4 full belts of Steel plate, 4 full belts of Stone and 2 full belts of Iron plate. (Does not actually use full belts - ratio is more Sp 3,556 belts, Stone 3,556 belts and Iron plate 1,511 belts, but I can't be bothered with splitting stuff when it comes directly off trains and will limit the input anyway).
I block can support 4 of the Production (pink) Science builds showed above. So for my 4800 build I need two of these blocks + 2 extra modules (total of 18 belts of Rail to support the 9 columns of Pink science).

Production Science (pink) - Rail production block

Factorio 0.17.28 Production (pink) science build - test1.png (2.49 MiB) Viewed 6949 times

Be aware - this is not tested out in production yet! I have only done the math and limited testing in game at my BUS test site. So no idea if this works or not or if there are errors, bottlenecks or plain stupid stuff in it If it works as hoped I might post blueprints later if any want.

And if any has a design suggestion or ideas for improvement please give me!

When designing this I got a feel that the outputs are more 0.16 scaled when it comes to rails?

/old dude

[MeduSalem]

10+ full express belts just of stupid rails to the center of the base where you produce the science packs.

You don't need to put the rails on belts.

I habe more than 10000 bots dedicaded hauling just stupid rails. That is almost half of what ALL other science packs require combined.

10000 bots hauling high volume stuff isn't really a solution. Just thinking about all the roboports needed to charge that many bots makes me cringe. Especially since it sounds like you already have another 20000 bots also hauling stuff.

My entire base is bottlenecked like hell because of this stupid rails and I can't do anything about it...

So design a different solution. I find that direct insertion works well for production science. eg something like
beaconed.prod.science.jpg

For me solving those sort of production/logistics issues is a large part of the core gameplay loop of Factorio. If your existing approaches aren't working, then sometimes you just need to find a different approach.

Honestly... I don't mind the amount of roboports... I have a modular base layout where I place roboports evenly spaced in a "ring" at max-roboport distance and then fill out the area in between with 5 rows of each 12 machines, for a total of 60 machines and 6 rows of beacons to speed it up. The modules are 2x2 chunk-aligned, leaving also some space for streets to walk/drive around. Easy to copy&paste/blueprint the stuff around to create bigger bases. Never had an issue with the roboports at all, even less once the robot speed is up. Easily managed 50000 bots under the 4k SPM stress-test with no bot recharge queuing at all and still had less than 3ms update time for logistics, something a belt-setup would never be able to beat.

Been trying various similar designs with direct insertion already the night before I wrote the post... and pretty much ended up hating everything I came up with and even your design doesn't appeal to me because it messes up the alignment of the beacons due to the gap between the assemblers. I already had something better without the alignment issues, but before I could find a proper way to place the inserters/chests I gave up on Vanilla Factorio entirely.

I was trying to build a test 4k SPM base using creative mod to test a few new layout strategies before eventually going the real deal (because I hate rebuilding everything after the initial science is done) and before I got to build the furnaces it already ended up with 45 UPS and worse slow-downs due to the amount of entities on the map (even if I go full-retard PM3/SM3 beacon-setup). Even if the rails weren't an additional issue the map became virtually unplayable without a single furnace on the map... now imagine if there's 1000 Iron/Steel furnaces.

The 1000 SPM base ran pretty good... but 4000 is way too much for my computer to handle. Maybe a 2000 SPM base might be more realistically possible with the game update times I had left over after studying the 1000SPM base. Total game update time was somewhere around 8-10ms for 1000SPM, so with 16.66ms needed for constant 60UPS a 2000SPM base seems more like the limit, but then after that that's definitely it... can't go further with Vanilla. Yeah, my computer is 10 years old too. Even though it's a Quad Core and still runs pretty much every newer game still fine (or otherwise I would have bought a new computer already) it is obviously limited by the single-threaded performance of Factorio.

So in the end I just started playing with Mods for the first time since 2012 or whenever I started playing Factorio to be able to push beyond the limits of my computer.

Namingly Bob's Mods to cut down on the amount of entities even though some of his stuff is way too OP to be in league with the standard game in my opinion. Especially the modules resulting in 240% total productivity in machines resulting in a 6k SPM base in my first test build (didn't even intend that high, it's just the modules are so ridiculous) that is about the size of about 4x2 chunks while my previous 1k SPM Vanilla base required almost 25 2x2 chunks, also using only 4 Mk4 roboports instead of hundreds of vanilla's. Ridiculous. Cheeses the game big time... that way I could probably go 20k SPM easily. Also using his changed Science recipes which I find more in line with what it used to be.

Also trying out Angel's Mods in a parallel savegame for the same reason... maybe in the end going for a combination of Bobs+Angel's, though I have yet to determine in which fashion because I know Angel doesn't allow Productivity Modules in most of his production chains (even though I like his chains a lot more because they are more complex than just slapping down a furnace and PM the hell out of it), effectively increasing the amount of machines again and there seems to be no setting to enable productivity modules there either.

Maybe I will abandon the idea of having a huge mega-base entirely and settle for something with lower SPM but rather something more complex (like Angel's) and also looking more pretty and even being belt-based instead of bot-based for once.

I mean it's about trying to find new goals and ways to play... because if I am also honest... I have played Vanilla Factorio to death already. There is not a single thing I haven't tried or thought about already when it comes to designs and the one I used for testing I mentioned above is the ultimate limit. There is nothing better anymore (that is if the devs don't decide to implement new production chains or something else screwing up the base, like f.e. the rails, though they are a minor annoyance and I agree that it can be done using brute force) and all I have been doing the past 3-4 years is pretty much just fiddling around figuring out 1-2 tile differences in layouting to find the best module design... ending up being burned out and not playing for a year in between major updates because it got boring and annoying that there's nothing left to explore.

[Zavian]

even your design doesn't appeal to me because it messes up the alignment of the beacons due to the gap between the assemblers.

Just wanted to point out that although not all the assemblers in my design above have 8 speed beacons, they all have enough speed that the bottleneck is the purple science assemblers. Sure I could redo the design so that everything got 8 speed beacons, but since everything is keeping up, and I can't add more science assemblers without either some starving some of the time, or a radically different design, the only purpose would be changed aesthetics. The limit on the purple science assemblers in that design is the shared belt for stone + iron sticks. Adding another 2 science assemblers would push the demand for iron sticks and stone to 23.5 of each, more than the shared blue belt can supply. Theoretical production with 6 science assemblers is 6.6 flasks per second. If I was willing to settle for the science assemblers producing 6.3 flask per second, then that would consume a full lane of both iron sticks and stone, but actual throughput would probably be closer to 6.2, unless I tweaked the iron stick output inserters, since that output inserter design will stutter. Completely draining the stone + iron sticks belt might also be awkward from an inserter throughput standpoint. (I think it will be ok, but I would need to build and test it to be sure).

Regarding mods, I tend to agree with you. I find modded gameplay more interesting than building a vanilla megabase. If you like Bobs + Angels, then you might enjoy Seablock, which includes both Bobs + Angels.