Today in 0.17 news, belts are gaining a very slightly higher throughput, with the basic yellow belts going from 13-and-a-third items per second to an even 15 items/second.
The level of detail in that post is why I love the Factorio dev blog, where it breaks things down to the individual pixel, and how these numbers are an emergent property of simple rules, and very small changes can be the difference between things being simple or complex down the line.
This is a great change in that it greatly simplifies the ability to make circuits that can detect whether a belt is fully compressed: A fully-compressed belt segment now has exactly 8 items on it, compared to the old behavior in which it might have anywhere between 6 and 8 items.
Mind you, I don't exactly have an application for this in mind, but I'm confident the community will use it to do something clever.
Though it definitely got annoying when I was forced to expand.
Yeah. Clearing nests is just straight-up bad gameplay in my opinion. I have to work pretty hard to find bad things to say about Factorio, but this counts.
This is a great change in that it greatly simplifies the ability to make circuits that can detect whether a belt is fully compressed
My Kovarex rig is so happy right now!
EDIT: Just reading the FFF now. Apparently the new mod API will allow the creation of arbitrary lights. I want to make "Lantern Bots" that will follow the player and provide light. It might be time to learn some Lua.
MightyMooquack wrote:This is a great change in that it greatly simplifies the ability to make circuits that can detect whether a belt is fully compressed
My Kovarex rig is so happy right now!
I never did get the need for circuits. Would somebody like to give me some "for examples" that would encourage me to dig into these some more?
Here are some simple things you can do with circuit networks. They are pretty optional in vanilla IMO but fun to mess around with.
I use them for a couple of things:
- Control how many bots get created and added to my main base' robot network
- Control the flow of uranium fuel cells into and out of my reactors
- Remote sensing of how much remote bases have in various containers (like acid for uranium mining)
They really become mandatory with some mods like LTN, especially if you want to control how much "stuff" is needed/available for transport by trains beyond "full" and "empty".
So far, the first use that tboon mentioned is the one that I have used them for, but will probably branch out to uranium stuff at some point.
The top use for circuits is to regulate oil cracking. I usually do it like this:
Hook up a tank to your heavy oil and your light oil. Set up a pump so that it sits between the tank and the relevant oil cracking plants. Connect the tank to the pump with a wire, and configure the pump to only activate when the tank has more than (say) 20,000 units of fluid in it.
The result is that you should always have a supply of heavy and light oil for making the things which require it, and any excess will cascade over into petroleum gas.
Another standard use is regulating the number of bots in a logistic network. I often set it up so that it adds bots as long as the number of "available" bots is less than some number, usually 100. (In very large factories that make extremely heavy use of logistic bots, I might set this margin to 1,000 bots.) This allows my supply of bots to grow automatically with demand.
Another: Once I have a proper mall set up in my factory, I like to set up a "construction train," which is basically a portable mall that I can summon to anywhere on the rail network, which allows me to build things out without having to keep going back to the mall to restock on items.
Put another way, this is a train with a huge variety of items that need to be automatically restocked. There are a couple of ways of doing this, but one solution I've used was to design this unholy mass of combinators, to handle the train loading.
A lot of the time, you can simply use a bunch of requester chests instead, but after designing the blueprint, I figure there's no reason not to use it.
Another: Eventually I'll set up a train that supplies my outposts with defensive consumables: Repair packs, bullets, fuel for flamethrower turrets, and so forth. I use a circuit (in a blueprinted train station) to enable the station only when it has run out of any one item.
Another: If I'm using solar power, I'll usually wire up my old steam power plant such that it will only activate itself when my accumulators run out, and stay active until the accumulators are close to fully-charged again. An example for how to do this may be found on the wiki.
Once you get used to circuit networks, you'll start seeing opportunities to use them everywhere.
Put another way, this is a train with a huge variety of items that need to be automatically restocked. There are a couple of ways of doing this, but one solution I've used was to design this unholy mass of combinators, to handle the train loading.
I don't think it was added until a fairly late iteration, but you can set filter limits on what types and amounts of material any given train car can take. I think all you'd need in 0.16 would be a string of belts feeding different materials into chests, and then filters on your rail cars so that only as many resources as you want will get loaded. Fill times wouldn't be very fast, and the feeding belt network might be a little complex, but it wouldn't likely take much logic.
With logistic bots and requester chests, there wouldn't even be any special feed complexity.
Oh, one useful trick I found was that you can connect belts to circuit networks. This can function as a choke point, like a pump. If the enabling signal is on, the belt flows, but if not, it turns off.
It's been a number of months since I fired up the game, and my memory's a little blurry, so I'm not certain if it was the actual belt that got disabled or the belt under the squarish things you can add. (sensors, I think.) But there is *definitely* a way to turn belts on and off easily, which can be handy in the same way as overflow pumps in your oil industry.
I was specifically finding this useful for Kovarex uranium enrichment; the feeding belt wouldn't allow any ore through until the target belt was empty. This meant that the reactor wouldn't get backed up, because its own output would get priority and recirculate back into the reactor again. Only when that ran low would the input belt add more.
edit to explain: Normally, the ore would go around and around the reactor, 40 getting loaded and 37 or 38 getting unloaded each cycle. If the output was blocked, that process would stop working, so there always needed to be enough room on the output belt to empty the reactor completely. But then there also needed to be enough to fill the reactor completely, and each cycle had a few less ore. So, the sensor ensured that new ore was added when there was room, but only when the supply was exhausted far enough. I think the circuit was just a sensor block, a pole, a normal vanilla feeding belt, and two wires to connect sensor->pole->input belt.
Put another way, this is a train with a huge variety of items that need to be automatically restocked. There are a couple of ways of doing this, but one solution I've used was to design this unholy mass of combinators, to handle the train loading.I don't think it was added until a fairly late iteration, but you can set filter limits on what types and amounts of material any given train car can take. I think all you'd need in 0.16 would be a string of belts feeding different materials into chests, and then filters on your rail cars so that only as many resources as you want will get loaded. Fill times wouldn't be very fast, and the feeding belt network might be a little complex, but it wouldn't likely take much logic.
With logistic bots and requester chests, there wouldn't even be any special feed complexity.
One limitation is that, using either of these approaches with long inserters, the maximum number of distinct items you can have in a train wagon is 24 (two rows of six inputs on each side). My crazy combinator approach would allow filling each of the 40 slots in a wagon with a different item (though you'd need four copies of the blueprint to do it, since each one is limited to 12 items, the maximum number of requests that a requester chest can make).
So, yeah, that's better than just belts, but as soon as you go to requester chests, that's not a problem anymore... four chests would allow for 48 unique item requests. And then you could make up volume with the remaining chests on the same side, or all the chests on the other, no?
I would think that if you're up to the point of genuinely needing more than 12 items in any given car on a train, you're probably advanced enough to have logistics bots?
Another possibility would be having two stations, but getting the train to move out of the first one reliably might be hard.
If you just naively request multiple items in a single chest, then you would need to set the inserter's stack size to 1, to avoid the problem of the inserter attempting to over-fill the cargo wagon, and the loading process deadlocking when the inserter can only fit some, but not all, of the items it is currently holding.
The bulk of what that mess of combinators is doing is conditionally changing the hand size of the inserter so as to avoid this problem.
Been playing again and I always hit the wall. I just find getting enough basic materials to be rather grindy. Maybe it would better with friends but I always lose interest by blue or yellow science packs.
Still the game is amazing and makes me want to go back again and again....
I just restarted with the Whistle Stop Factories mod. It creates a scattering of super-furnaces and assemblers with 30-50x throughput that you're intended to connect with trains.
I've gotten a starter base running so now I'm just scouting the map in a car looking for some of these furnaces and addemblers and some big deposits to feed them. I'm still thinking of a bus design, using the big factories just for the belt supplies, but I'll see what I find. That plan might be too small.
Finally finished and launched a satellite. I was griping before but turning in high resources helped a lot. Still a lot I haven't done but this seems good for now.
Many rocketey congratulations to you!
In today's update, we get some information about the GUI update and the wave defense scenario, but we also have this tidbit:
You can see, that there is still a lot of to do, but the work tends to accelerate as more and more of the GUI layouts/tilesets/standards are being finalized and reused. The conclusion is that 0.17 experimental in January is possible, but it might be February as well :).
So the January release thing may not be set in stone. Oh well!
And now in today's update, we get confirmation that 0.17 won't be out in January:
Looking over what is left to be done, It is clear to me that the release won't be ready in January.
When we are ready to release 0.17, its launch definitely won't be a surprise. We will announce the exact date in the FFF at least the week before.
Ah well. But the new toolbar is nice, eh? I'm going to have to retrain a lot of muscle memory. I'll need to learn to hit 6 instead of shift-1. And the ability to drop a ghost at-will is something I've wanted for a long, long time.
I just restarted with the Whistle Stop Factories mod. It creates a scattering of super-furnaces and assemblers with 30-50x throughput that you're intended to connect with trains.
Launched a rocket for this playthrough- I could probably have done it several hours earlier than I did, but I was busy just adding new outposts and a big nuclear plant. Then I had to spend a while untangling contaminated fluid stations (caused by LTN issues with stuck trains). But at this point I can build a rocket in less than 2 minutes and I'm getting white science faster than I can use it.
LTN did save me a lot of trouble; I still have a crapload of trains but not as many as if I had to dedicate them for each pair of stations, and because of depots I don't need parking lots or fueling at every station. It's sure nice to be able to just plunk down a station that demands X and Y and provides Z, independently of everything else.
Just recently I found the LTN Tracker mod which adds a helpful GUI to LTN, listing all of the depots and trains and all of the materials in the LTN network. It's new and a little crashy though.
I ended up not using the big factories for space production as it just seemed easier to set it up from a blueprint, and I didn't need that much stuff. If I keep going with this base I'll probably build the stations though so I can have more rockets. But to do that I'd need to up several big trees of production and I'm getting tired of building stations.
I'll wait for 0.17 before trying again, and maybe next time I'll use a different train manager.
Probably have to since I doubt something as complex as LTN would be ported to 0.17 for a while (hopefully not - hopefully Wube has been getting modders of at least the biggest mods early acces for them to be getting their mods updated close to day 1).
I keep plunking away at my vanilla game when I need to play me some Factorio, waiting for them to release 0.17. Got a lot of cheevos out of this one, I guess I play modded most of the time. 
I started up a new game a couple of weeks ago. I actually misread something here and thought that the new version had just dropped, not that it was about to drop, but oh well. It's been two years since I've played, so I definitely have new stuff to toy around with. (Splitters filter now!!)
I'm way further than I got at my first playthrough, which I believe stalled out trying to get blue science working well. Now I've got robust Red, Green, Blue, and Black science production. I've got a tiny little yellow science line squirreled away as well, but I have to manually move the inputs and outputs.
I'm coming to the end of a big spring cleaning effort, however, which I think will help me push forward and progress into full purple and yellow production. My Copper has been mined and smelted off of the mine base for awhile, but now I am also coiling it right at the edge of my base. Another more distant base mines and smelts iron, steel, and crafts iron gears, shipping these three products in by train. The offsite Steel and Gear section is new, and yesterday I finished pulling up those sections of my main base, then doing another round of cleaning where I could now more directly route belts that used to need to meander around.
That gave me the room to expand circuit production, and there is still a bit of surplus breathing room (which I think I'll hold on to as possible sites for train stations). I also set up another distant manufacturing site, although rather than producing raw goods this base creates my most advanced products: importing the three circuits and creating lvl 1 Speed Modules and lvl 3 of all modules (Without Purple Science I don't have lvl 3 Productivity modules yet, but the factories are waiting).
.17 is out.
List of changes/updates: https://forums.factorio.com/viewtopi...
So .17 is only kind of out. It's 'experimental' and you have to opt in to the .17.x beta on Steam to get it.
The tutorial campaign is in, it's apparently incredibly difficult, and the devs would like you to try to break it.
I'm about to give it a crack.
Yes, this is how they've done new releases since antiquity: They start as "experimental" releases, and once they're confident with it, they will switch it to being "stable."
So far it's pretty great. I think the changes to the toolbelt are quite good, but the keyboard shortcuts are different, and will require some getting used to.
It's been a while since I jumped into this game, and it's nice to get back into it.
Is .17 going to break save game compatibility?
I believe it does. Blueprints seem to come across OK, for the most part, though.
I like the changes in 0.17 so far. I like that the hot bar is now an alias instead of a separate inventory slot. It seems like they have increased reach (I did not see that in the patch notes nor in the dev blogs) which is good IMO. Manually putting down a smelting area last night was much smoother than usual.
Tier 2 factories now take steel which I think is good (steel was generally utilized IMO) but it has messed up my early bootstrap base design.
Also, want to thank Mooquack for his calculator and his efforts to keep it up to date with the patches. Much appreciated!
Save games should still load just fine. It's just that some parts of your factory will stop working correctly, due to the changes in recipes.
The increase in reach was mentioned in the changelog, though it is one tiny line among many. The devs are highly receptive to player feedback, and have pulled a number of popular mods into the base game: Increased reach, the upgrade planner, and the research queue come immediately to mind. Neat!
0.17 is now up as an Experimental build, they're at 0.17.6 already, as they're patching bugs like mad.
They had an issue where their patch notes weren't being posted correctly on Steam... because it exceeded the 40K maximum message size! There are a lot of updates this time around.
Not directly related in anything except timeframe, check this out:
It took me awhile to realize what I was seeing, but once I did.... dear lord, I can't even imagine how to do that.
He apparently needed a mod to make it work, one that would build all the completing belt segments instantly. It's so timing-sensitive that apparently even electrical signals aren't good enough. All the belts have to start on the exact same tick for this to come out right.
I have been making mods.
Mod the first: Toolbar Hotkeys. This is a very simple mod that adds hotkeys for adding the item in your hand to the toolbar. CONTROL + Q will add it to the first empty slot (modulo a bug with blueprints which I'm hoping they fix soon). CONTROL + a number will add it to the corresponding slot. These binds may be changed in the game settings, and there are some other ones I'm thinking of adding.
Mod the second: Time Series Graphs. This is still somewhat experimental, but this mod adds a combinator which can record and plot time series data received from circuit networks. There are still a number of features I want to add, and I still need to write up some proper documentation (e.g. to describe the difference between the two types of combinator in the mod), but I'm really pleased with how it's turned out so far.
I have been making mods.
Mod the second: Time Series Graphs.
Nice! I was thinking about building something to export metrics to Elasticsearch, like this one for Prometheus. Being able to do custom metrics with your mod would be a perfect combination.
Pages