|
Thel posted:Yeah I've gotta say as an TTD vet from way back (anyone remember TTDPatch?), I naturally gravitated toward one-way tracks and loop stations. The nice thing about trains is that it doesn't really matter how fast they go. Only loading, unloading, and acceleration affect throughput. So you can just ignore travel time and put concrete under the tracks to boost pollution.
|
#
?
Aug 20, 2015 06:07
|
|
|
|
| # ? Apr 20, 2024 04:50 |
|
|
Phssthpok posted:The nice thing about trains is that it doesn't really matter how fast they go. Only loading, unloading, and acceleration affect throughput. So you can just ignore travel time and put concrete under the tracks to boost pollution. Are you saying the impact of speed is negligible, or that it doesn't matter at all? It does have an impact, and the significance depends on the proportion of travel time vs. loading time. Sending a train back and forth between two stations is basically like shouting at a wall, waiting for the echo, and then shouting again. If the sound wave travels faster, you have less turnaround time before you can shout again, so you'll get more shouts per second.
|
|
#
?
Aug 20, 2015 06:39
|
|
|
Phssthpok posted:The nice thing about trains is that it doesn't really matter how fast they go. Everything about this sentence is wrong.
|
|
#
?
Aug 20, 2015 06:58
|
|
|
The argument is that you can just send out more trains. If your track is constructed properly, you can add more trains to it until your buffers are overflowing with raw resources.
|
|
#
?
Aug 20, 2015 07:00
|
|
|
Now I want to completely fill the track between 2 stations with train and connect it to itself in a loop. For optimal wear it should be a Mobius loop, but then we'd need wheels on the roof... e: wait, no, that wouldn't actually help. a simpler solution would be a figure-eight track so both rails are the same length. but you can't make that in Factorio either since there are no bridges Nomikos fucked around with this message at 07:49 on Aug 20, 2015 |
|
#
?
Aug 20, 2015 07:19
|
|
|
Nomikos posted:For optimal wear it should be a Mobius loop, but then we'd need wheels on the roof... http://i.imgur.com/y4fCFt7.gifv Dirk the Average posted:The argument is that you can just send out more trains. If your track is constructed properly, you can add more trains to it until your buffers are overflowing with raw resources. True to a point, but train speed still sets ultimate limits on throughput per track.
|
|
#
?
Aug 20, 2015 08:25
|
|
|
Throughput per station always becomes a bottleneck before throughput per track. Any time you have resources coming from multiple places to a central hub, the unloading speed determines how fast resources can flow through that hub. A longer track or a slower loading speed will have a noticeable effect on latency, but a negligible effect on throughput. This breaks down when there is only one train serving an unloading station. In that case you would usually be better off adding a second train or using conveyor belts instead.
|
|
#
?
Aug 20, 2015 11:15
|
|
|
My main issue with double header trains is that if I ever want a train to carry multiple resource types in different cars, I need to add new stops in even increments, since it's going like a slinky from station to station.
|
|
#
?
Aug 20, 2015 13:16
|
|
|
I had no idea people even made arguments against loops. Loops own. All of my railwork is single-header trains with turnaround loops. Factorio is the closest I get to playing TTD these days, since OpenTTD DFW on high-res screens. 
|
|
#
?
Aug 20, 2015 16:00
|
|
|
ToxicFrog posted:I had no idea people even made arguments against loops. Loops own. All of my railwork is single-header trains with turnaround loops. The only arguments against loops I've heard is that it takes more space and limits where you can place them and forces potentially a bit of redesign of your loading stations, and that it takes longer for a train to get on the main line* (you would accelerate faster though, so this may be wrong depending on distance) What happens if you do two trains pointed the same way? Do they both generate thrust or just one?
|
|
#
?
Aug 20, 2015 18:18
|
|
|
Slickdrac posted:The only arguments against loops I've heard is that it takes more space and limits where you can place them and forces potentially a bit of redesign of your loading stations, and that it takes longer for a train to get on the main line* (you would accelerate faster though, so this may be wrong depending on distance) Both pull, I think. The train puts out a limited force and heavier loads will go slower. This is a disadvantage to the non-loop system; you're lugging around a heavy, idle engine all the time.
|
|
#
?
Aug 20, 2015 18:47
|
|
|
Lugging an engine backwards is apparently the equivalent of 2 loaded cars. I like to keep 2 cars per locomotive generally to keep acceleration quick, which means to doublehead I'd have 2 locomotives and no cars... I mean, 2 locomotives and 1 car would be fine, but I guess I'd rather not.
|
|
#
?
Aug 20, 2015 18:53
|
|
|
So now that I've had a chance to take the tank out for a spin and destroy a few bases with it, some thoughts: Machine gun: just as worthless once you're past the very early game as the hand-held version; even loaded with AP rounds it'll take several complete clips to take down even a single biter. Main cannon, AP shells: one-shots anything you point it at, but only hits one target, fires slowly, and misses half the time. It would be good for taking out worms and nests in an alternate universe where fighting worms and nests didn't automatically mean fighting dozens of biters at the same time. Main cannon, HE shells: not as much damage, but reliably takes out groups of biters. Probably my favourite actual gun so far, including the handheld ones. But none of them compare to... Ramming: holy poo poo, this is where it's at. Ram biters, ram nests, ram worms. Grind everything that moves under your steel treads.  I'm probably going to switch to LRPC artillery once I have enough alien artifacts to research it, but in the meantime running everything over with the tank is a decent alternative. Speaking of artillery, is it possible to get logistibots to pick up items lying on the ground, or will I have to manually trek out to each smoking crater to loot it?
|
|
#
?
Aug 21, 2015 15:07
|
|
|
What is "lrpc artillery"? vvvv: Thanks, I was just looking for something longer range than turrets... Truga fucked around with this message at 15:40 on Aug 21, 2015 |
|
#
?
Aug 21, 2015 15:15
|
|
|
Truga posted:What is "lrpc artillery"? It's from Total Annihilation, he's probably referring to one of the artillery pieces in http://www.factorioforums.com/wiki/index.php?title=Mods/SupremeWarfare
|
|
#
?
Aug 21, 2015 15:30
|
|
|
Yeah, those are the ones. Sorry, should have been more clear. "LRPC" in TA or Supreme Commander is Long Range Plasma Cannon, and refers to artillery that can shoot across the entire map (except on the very largest maps). The Supreme Warfare mod looks pretty heavily inspired by TA, although I'm not sure how long the 1000-unit range on the biggest gun actually is in Factorio terms.
|
|
#
?
Aug 21, 2015 16:12
|
|
|
Doing some more random reading on the Factorio forums about loops and double headed trains, it seems to be a somewhat apples and oranges situation. You've actually got two distinct elements to your train network: Train Direction You've got unidirectional trains that will only go forward, or double headed trains that can go in either direction. Intersection Design Here you have junctions where tracks actually cross each other, or roundabouts, where everytime track meets they all merge onto a single loop. Station Layout There's actually a third one that's sort of related to the first, but the design of the stations is really another element. You've got the choice between terminus stations (trains enter and leave the "platform" on the same track), which you need double headed trains (so they can reverse out of the station). Of there are RoRo stations (roll in roll out) which can use either double or single headed trains, but single headed trains will require these. I actually just had a kind of crazy idea if you're using truly enormous trains (like, 4 or 5 locomotives). Put most of your locomotives pointing forward, but point one pointing backwards. Build a terminus station but have a long siding track (as long as the train) with a stop at the end of it. So the 4 or 5 locomotives pull the big heavy train into the terminus station, the single reversed locomotive pulls it out of the station into the siding, and then the 4 or 5 locomotives will be able to pull it back onto the main-line. E: I made this, and found a gif recorder, so you get to see it. It's actually pretty silly, probably don't do it. http://i.imgur.com/9OH2DgO.gifv FISHMANPET fucked around with this message at 18:48 on Aug 21, 2015 |
|
#
?
Aug 21, 2015 16:31
|
|
|
ToxicFrog posted:Yeah, those are the ones. Sorry, should have been more clear. It's pretty far, as in 3 radar squares far and I still have the gun hitting things although the shots take a hell of a long time getting there. It's not entirely intuitive either, you have to run in (or tank in) and throw the 25 range 'tag' capsule - how you deal with the large pack of crap now following you is your problem. The guns (we built a pair of them in our latest save) will totally flatten whatever was at the target site. The supreme warfare mod works well but we found the research costs could do with being balanced and given how many shells it takes to flatten a village with artillery, they could be a bit quicker to make.
|
|
#
?
Aug 22, 2015 11:22
|
|
|
I got kinda bored with the vanilla game in 0.11, so when 0.12 came out, I decided to plug in RSO and see what it was like. I can get in a train, set course for my furthest mine, go make tea, come back, and still be traveling. I can't decide if it's hell or the best thing ever. If you want the game to force you to use vehicles, use RSO, it's awesome.
|
|
#
?
Aug 22, 2015 11:43
|
|
|
Ratzap posted:It's pretty far, as in 3 radar squares far and I still have the gun hitting things although the shots take a hell of a long time getting there. It's not entirely intuitive either, you have to run in (or tank in) and throw the 25 range 'tag' capsule - how you deal with the large pack of crap now following you is your problem. The guns (we built a pair of them in our latest save) will totally flatten whatever was at the target site. The supreme warfare mod works well but we found the research costs could do with being balanced and given how many shells it takes to flatten a village with artillery, they could be a bit quicker to make.  If I had known that I'd have never downloaded the mod, that's awful.
|
|
#
?
Aug 22, 2015 21:34
|
|
|
ToxicFrog posted:
Bit late reply and maybe answered, but you can use the deconstruction blueprint to get your robot army to pick up things on the ground. If you have a personal roboport (which you should if you roam in a tank), they will pick it up for you. You can flag it from far away, as far as you can zoom, then roam past in a car or tank. The bots will catch up eventually if you stop for a bit.
|
|
#
?
Aug 23, 2015 08:07
|
|
|
ToxicFrog posted:
After all the research and huge build costs, it was a little disappointing to say the least yes. There's actually a funny story in it though: We'd built the 2 guns and set them up at the south edge of our turret perimeter. I was ready to go and Dred came running up to watch our pinnacle of destructive power come online. Looking at the tag it has 1000 range, so I thought 'Aha, pull up map and click on nest?' So I pulled the map up, found the nearest nest and clicked a tag on it. There was a ginormous BANG! to which I thought 'wow that was fast' - then Dred started swearing at me on mumble at the same time as I saw the death messages come up... The tag had dropped at my feet and the guns happily annihilated us both. Robots rebuilt the carnage site before we respawned and the guns themselves survived as they have 2500hp. Once we stopped laughing we had another look at the tag. It has 2 ranges: 25 and 1000. The 1000 is how far it can signal the guns to fire at it and the 25 is how close you have to be to throw it (like other capsules). So long range artillery with a huge cost in terms of research and material coupled to a 'spotter' that has to streak into hives to do it's stuff.
|
|
#
?
Aug 23, 2015 21:51
|
|
|
Ratzap posted:After all the research and huge build costs, it was a little disappointing to say the least yes. There's actually a funny story in it though: We'd built the 2 guns and set them up at the south edge of our turret perimeter. I was ready to go and Dred came running up to watch our pinnacle of destructive power come online. Looking at the tag it has 1000 range, so I thought 'Aha, pull up map and click on nest?' So I pulled the map up, found the nearest nest and clicked a tag on it. There was a ginormous BANG! to which I thought 'wow that was fast' - then Dred started swearing at me on mumble at the same time as I saw the death messages come up... The tag had dropped at my feet and the guns happily annihilated us both. Robots rebuilt the carnage site before we respawned and the guns themselves survived as they have 2500hp. If I have time to open the mod up and look around, I might see if I can fix that. Drop the beacon, replace it with a control panel that opens when you access the gun. Enter bearing and range, click fire. The shells may need to be made cheaper, since it'll probably take several ranging shots before you actually start hitting your target with this design, but I like this a lot more, at least on paper.
|
|
#
?
Aug 24, 2015 12:55
|
|
|
Can you make spotting shells that do no damage? I don't know how you'd show the impact on the minimap.
|
|
#
?
Aug 24, 2015 14:44
|
|
|
I've been trying a new high-productivity foundry pattern. I really like how elegant it is, but it's bottlenecked by one belt of output.
|
|
#
?
Aug 24, 2015 14:51
|
|
|
Phssthpok posted:I've been trying a new high-productivity foundry pattern. I really like how elegant it is, but it's bottlenecked by one belt of output. Is having iron, copper, and stone available to smelt by design? If so, how do you determine what gets worked on and when? Looks like a nice, compact, if somewhat inefficient design. E: I started playing around with trains this weekend, which led to me solving my green circuit shortage by starting a new outpost and completely outsourcing manufacturing; bringing them into my all-purpose depot. That got me wondering about rebuilding my bus and integrating logistics support. Have each line fed by a series of requester chests, and have all finished products sent to a central storage. What else am I supposed to do with my 1300 logistic robots? 
Royal W fucked around with this message at 15:05 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 14:57
|
|
|
Phssthpok posted:I've been trying a new high-productivity foundry pattern. I really like how elegant it is, but it's bottlenecked by one belt of output. Well then: is that the right number of smelters to 100% that setup?
|
|
#
?
Aug 24, 2015 14:58
|
|
|
Don't beacons with two L3 speed modules give only +50% speed? With beacons the same size as smelters, You'll get better performance per tile by just using smelters. Right?
|
|
#
?
Aug 24, 2015 15:07
|
|
|
Royal W posted:Is having iron, copper, and stone available to smelt by design? If so, how do you determine what gets worked on and when? Looks like a nice, compact, if somewhat inefficient design. You have to filter the ore inserters to keep iron smelters from grabbing their own product and trying to make steel. So this configuration will work for everything except steel, and the same layout with no filters will work for steel. What do you mean by inefficient? Almost all the furnaces have the maximum number of effect sources. ikanreed posted:Well then: is that the right number of smelters to 100% that setup? I think the right number is 11 with all level-3 modules. 13 if using more underground belts to get past the output bottleneck, which I am still playing with. A lot more with level-1 modules. Is there a good way to work out the 100% capacity? Bhodi posted:Don't beacons with two L3 speed modules give only +50% speed? With beacons the same size as smelters, You'll get better performance per tile by just using smelters. Right? Each speed module is multiplying the effect of 12 productivity modules. So 2 speed modules gives you the equivalent of 6 more productivity modules. For a large build, this is much more cost effective than using only furnaces.
|
|
#
?
Aug 24, 2015 15:11
|
|
|
I just looked again and realized those are smart inserters, not chests. Now I'm imagining one huge central foundry that processes ore in batches based on demand using combinators. I'm surprised that I haven't really seen any kind of absurd wizardry being done with the new logic system yet, it seems like a rather robust system.
|
|
#
?
Aug 24, 2015 15:16
|
|
|
Phssthpok posted:What do you mean by inefficient? Almost all the furnaces have the maximum number of effect sources.
|
|
#
?
Aug 24, 2015 15:21
|
|
|
Royal W posted:Inefficient may have been the wrong word, sorry. I've just never seen a layout like that before, and the ore and plate on the same belt looks really bizarre and frightening. Yeah, it's weird, but it's equivalent to a full blue belt of each. There's just no messing around with getting output onto the inner track, which can be pretty convoluted in the 2 tiles you get within beacon range. Here's what it looks like tiled. Huge central smart foundry here I come. 
Phssthpok fucked around with this message at 15:34 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 15:27
|
|
|
Does that still work where putting a belt sideways onto an underground belt only puts one half of the belt onto the underground belt? I would have assumed that's one of the many things that went away when they redid belt mechanics, but maybe that was also an intentional design.
|
|
#
?
Aug 24, 2015 15:40
|
|
|
yep, that still works. It's a neat trick.
|
|
#
?
Aug 24, 2015 15:41
|
|
|
Wouldn't it make more sense to do like beacons belt furnace beacons furnace belt beacons rather than have each smelter taking from and feeding two belts?
|
|
#
?
Aug 24, 2015 16:04
|
|
|
FISHMANPET posted:Does that still work where putting a belt sideways onto an underground belt only puts one half of the belt onto the underground belt? I would have assumed that's one of the many things that went away when they redid belt mechanics, but maybe that was also an intentional design.
|
|
#
?
Aug 24, 2015 16:05
|
|
|
Dr. Stab posted:Wouldn't it make more sense to do like A high-speed furnace will make full use of 4 inserters, so you lose throughput with less than two belts within reach. With only two tiles between the furnace and beacon, there's room for about 1.5 belts on each side. So you really need belts on both sides (or logistic robots) to make the most of a high-productivity setup.
|
|
#
?
Aug 24, 2015 17:03
|
|
|
I feel like I've got a good handle on most things. Are there any good mods that adds more stuff to build? Preferably without making things more complicated.
|
|
#
?
Aug 25, 2015 22:39
|
|
|
Resource Spawner Overhaul (RSO) basically forces you to make a train network because it places patches of ore pretty far apart but they get more rich/dense as you move away from the starting area. Fully Automated Rail Layer (FARL) makes creating this network much easier. There's another mod that starts you with modular armor and a personal roboport + some basic goodies to get you started.
|
|
#
?
Aug 26, 2015 15:02
|
|
|
|
| # ? Apr 20, 2024 04:50 |
|
|
I know folks were interested in Big Pharma. It's on Steam now, and on sale this weekend.
|
|
#
?
Aug 28, 2015 00:23
|
|