Recommendations on your IC code: * There's VERY little reason to ever use the `sleep` command. IC10 scripts are very cheap on performance anyway, since they never run more than 128 commands per tick, and sleep can easily cause your program to miss things it should catch. For example, if you turn on your pump and then sleep for 3 seconds, that's 6 ticks that your script is not checking the pressure and could overpressure. Unlikely given the volumes involved, but it does mean that your script will never pump less than 6 ticks of gas into the pipe. If you're pulling from, say, 30 MPa, and pushing to a pipe network of ~600 L, that's 30 MPa of pressure you pumped in over those 6 ticks (or 3 MPA, if you're using a regular volume pump). If your output pipe were only, say, 300 L, oops, now you've hit 60 MPa in those 6 ticks. Using `yield` and evaluating pump state on every tick is almost always safer, and I guarantee you you won't notice a performance hit from it (I currently have at least 30 IC chips running every tick in my base and my FPS hasn't dipped at all, and your rig is rather better than mine). * I recommend placing your `yield` directly after `start:`, rather before `j start`. Main reason is, that single yield now covers every single conditional branch back to start, rather than having to have a `yield` before every single one, which saves a lot of lines. In addition, it makes sure you don't _forget_ to yield, which can cause VERY weird errors in your script otherwise. The only difference it makes is that your script won't run until the tick after you turn on the IC housing, which makes basically zero difference. * For "stateful" programs like your pump sequence, where the program is in one of several states and transitions between them (in this case, 3 states: pumping in, waiting for the gas to cool, pumping out), I find it's more robust to not try to keep state through the program, but instead _identify_ the state on each tick. For example, you can figure out which of those three states you are currently in by _reading_ whether the pumps are _currently_ on, and what the pressure in the cooling pipe is. If the input pump is currently on and the pressure is less than your target, you're inputting, so keep doing that (ie. jump back to start). If the input pump is on and your pressure is equal or greater than your target, time to turn the pump off and start cooling. If neither pump is on, you're waiting for cooling, so check temp, and if it's at or below the target, turn on the output pump. And then, if the output pump is on and the pressure is above the minimum, you're outputting, keep going (jump back to start). If output pump is on and you're at or below minimum pressure, time to start inputting again, so turn off output and turn off input. Now your program can handle whatever state the system happens to be in when it is turned on. It also means you can _manually_ flip one of the pumps to force the script to move to the next state (for example, flipping the output forces the script to consider itself outputting and check for minimum pressure to swap to inputting). Since the state of the code itself gets reset if you pull the IC10 chip to tweak the code, that also avoids the script being left in a weird state like having pressure in the pipe cooling and being reset to considering input (and thus never checking the temp or outputting). I don't _believe_ your program would run into issues when pulled mid-sequence, but avoiding specific states in the code itself helps avoid issues like that, especially as the sequence of things you're doing gets larger.
Rough lesson to learn. I’ve dealt with similar in the past with other games and my solution? After I get past my initial setup I leave that in place and build identical replacements in a facility nearby/another room/etc as soon as able/practable. Worst case, I can revert back to the original equipment. Remember: “One is none. Two is one.” Always have a backup.
It might be worth putting an IC chip in all the cold gas filtration units so they automatically stop filtering if the tanks are getting too full. This would mostly benefit the CO2 tank, which will eventually fill up, even if the gas is being used for the greenhouse. The hot gas side will probably need to have it integrated into the existing IC-housing code, since that will already be trying to turn the filtration units on and off.
And of distinct note, when using the onboard slot on a filter, disable filtering using the Mode var, NOT the On var. If you turn the unit off, the IC code can't run, so it also can't turn itself back on once pressure stops again. Setting Mode 0 has the same effect, the unit stops processing gas, but leaves the unit powered on so the IC code keeps running.
@3:57:00 - Name hashing is...complicated. Basically, you can either interact with a device _directly,_ using one of the device pins (on the IC housing or filtration or w/e), interact with all devices of that type globally (using the type hash from the Stationpedia), or interact with all devices of that type _with a specific name._ That third part is what chat was recommending. Using the name style, you need _two_ hashes, the device type hash (again, from the Stationpedia) and then a hash of the name. Now, the hash of the name is pretty obscure, so they included a handy function in the IC10 code for it: HASH("name"), which takes a string and converts it into the equivalent hash. So the idea is, you declare using `define` the _hashes_ of the _names_ at the start of the program. Example: define TankName HASH("Tank Waste Phase 2") define FilterName HASH("Filter Pollutants") That stores the integer hash equivalent of the name string you gave it into that define. You then access it using the named version of the batch commands, `lbn` (load batch named) and `sbn` (set batch named). Unlike `lb` and `sb`, which just take a device type hash (ex. `1280378227` for a large insulated gas tank), you use _both_ the device type hash and that name hash var. Note that just like `lb`, you need to include a batch _mode_ (ex. Minimum, Maximum, Average) for `lbn`, even if it's just one device, because the game doesn't know if only one device or multiple will match that type and name combo. Example: lbn r0 1280378227 TankName Pressure Average 1280378227 here is the type hash for large insulated tanks, from the Stationpedia. So that loads the average pressure across all large insulated tanks that have the name "Tank Waste Phase 2". sbn -348054045 FilterName Mode 0 As above, -348054045 here is the type hash for Filtration units, from the Stationpedia. This sets the mode var of all Filtration units with the name "Filter Pollutants" to 0, which has the effect of disabling it from processing gas while leaving it powered on (this is actually important for on-board IC10s, because if the device is off, the IC10 can't run, so it also couldn't, for example, turn the Filtration unit back on if the output tank dropped in pressure. Using `Mode` instead allows you to shut off the gas processing/output without powering off the unit, which allows it to turn gas processing/output back on later). The fundamental downside to coding IC10s like this is that you have to modify the code to point it at a different device. For example, you'd need to use a different version of the code for the Pollutants filter versus the Nitrous filter, since their output tank names are different. This is really the main benefit to using device pins, you can re-use the same code with different combinations of devices. On the flip side, named batch commands let you use batch commands on a more limited subset of devices. For example, put a pump on both outputs of the Filtration unit, name them the same, and then you can use `sbn` with that name (and their type hash) to turn _both_ on and off at once, without affecting any other volume pumps on that network, and without having to assign them to the limited 2 pins on that Filtration housing.
The bit that chat wasn't explaining was that I just needed to define the name at the top, then to do things, define the device type, then it's name from that definition later. At no point did I read anything that explained WHERE I was going to tell the thing that this named item is actually X type of device. This makes a little more sense now, thanks :)
@@Flipsie Ya, I suspected the issue was the confusion over "hash", since it's used in two completely difference senses with named hash functions. You're used to normal batch calls, so "hash" to you just meant the _type_ hash, and chat clearly wasn't explaining sufficiently that the names are _also_ turned into hashes (and that that is, in fact, the entire point of the HASH("name") function) >.
I would guess that if you went back and read chat at the time, you would be able to get more out of it than you were in the middle of streaming, it can always be a challenge to learn something fairly complicated with 10 different people trying to tell you how something works, all with limited time to try and write out a concise way of explaining how to do it
Also, anytime you use define, it needs to be on the initialized part of the program, it will give you an error if you try to define the same variable twice
And to clarify why you need to use the HASH("name"), is because ic10 programs do not store characters in any of the variables, it has to be a number, so the hash command will take a string and convert it to a number, so for your tank you would do define WasteTankPh3 HASH("TankWastePhase3") So now WasteTankPh3 would be the variable you would use as the device name. If you were to do this s dB Setting WasteTankPh3 When you point towards the ic housing it would show -275719632 That number is TankWastePhase3 converted to a HASH
I know that Splitsie has decided to do all of his cold stage filtering in parallel. But in case its efficiency needs to be improved, I think there's a very simple hands-off way to filter stuff in sequence (as long as you're fine with it not being a flow process). You set up all filtering units in a loop, with the "mixed" output of one piped to the input of the next, and the "mixed" output of the last piped to the input of the first. ("Pure" outputs from the filters obviously to their filtered tanks, and you can have the efficiency boosting pumps right after the filter outputs.) Then to each of those loop sections, you connect 2 things (on the output side of the efficiency boosting pump): A pump which feeds the mix from the mixed tank into each loop section. And an over-pressure valve (the one confusingly named "back-pressure regulator" in game) back to the mixed tank, to keep the pressure levels in all pipes safe.
You don't, and in fact _shouldn't,_ filter in sequence. As counter-intuitive as it sounds, you actually filter _faster_ from a common input pipe than in series, _and_ you avoid the risk of blowouts if one backs up and the need for intermediary tanks to buffer. The key with Filtration units is that their filtration speed is based on the pressure _difference_ between the input and the _higher_ pressure of the two outputs. So connecting the waste output _directly_ back to the input sets the filtration speed at the floor rate, because there's no pressure differential. The solution is to use pumps to keep the outputs at 0 pressure. And conveniently, the maximum pump rate of a normal Volume Pump is the same as the volume of a single pipe segment, 10 L, so if you build a single pipe segment off each output, then a normal Volume pump set to maximum, it will completely empty that pipe segment every tick, resulting in the maximum filtration rate. From there, it's just about keeping the input pressurized as high as you're comfortable (I generally keep mine pressurized to 45 MPa). The result is very rapid filtration, _even for trace gases!_ And since you're not running anything in series, it doesn't matter if a filter backs up or shuts off because its destination tank is full, because it won't affect any of the other filter units. And since the filter canisters in the unit are only used up when they remove the target gas to the output (not when they cycle gas through to the waste output), it also doesn't use up the filters any faster. For that matter, it doesn't even use more power. Also, on the note of backpressure regulators, keep in mind that their pump rate is _also_ based on pressure differential. They have a floor rate, but it's quite slow. If, for example, your mixed tank was quite high pressure, it's _very_ possible for your backpressure regulator to just flat out not be able to keep up with the waste output rate of filter it's supposed to be protecting. That's specifically why series filtering just doesn't work well. If one filter slows down or has to shut off because it's output is full, everything upstream also has to shut off, or one or more of the pipes are going to burst. In parallel, any filters shutting off doesn't affect any of the others.
One change to make bootstrapping possible would be to give you a tablet module/chip/[whatever it's called] that lets you call in a delivery of specific things. Obviously having it custom delivered would come at a premium (so that it's still cheaper to wait for a trader). This would let you get an autolathe to get you out of an effective softlock; but it could also be used if you want to buy a specific thing, like seeds or oxygen and don't want to wait for the respective trader.
I was thinking more of an "emergency 3d printer" (in the tablet or in the suit) that could craft the autolathe (ad few other things) consuming 10x the resources.
@@hubertnnn The thing that makes the emergency thing running on money more forgiving is that you can go into debt. If you are missing at least one of the materials too, you'd have lost the run. But in the end it probably comes down to what loss condition the developers like better.
I _suspect_ the reason the devs haven't is because making the autolathe itself isn't _technically_ enough. You also need 2 iron sheets, 2 cable coils, and 2 plastic sheets, none of which can be made without an autolathe, plus a screwdriver and a welding torch. That's why the sole other way of obtaining an autolathe, a package from a trader, include all of those things except the welding torch (it literally does include a screwdriver, though). But imo, it's completely silly, because it's too black-and-white. If you lose your autolathe, even if it's really _just_ your autolathe you lose, like in Splisie's case (ok, he also lost the pipe bender, but the point is, he still had at least one production facility, and basically all of his storage and (non-gas) resources survived), you're completely beholden to getting an appliance trader or you're dead. If you lose your autolathe before building the parts for a landing pad, you're also straight hardlock dead, no recovery possible. That's, imo, just silly. No other building is that make-or-break. Realistically, depending on the world, you could lose _everything_ except your autolathe and a power source (battery, solar, w/e), and either some metals (maybe in the autolathe) or an arc furnace for making more, and you could still conceivably survive. Yet you can be actively launching rockets into space and if a canister explodes in an unfortunate spot and takes out your autolathe, you're softlocked, and completely hardlocked if you don't have trading setup yet. There's no reason that building should be that pass/fail. Having them buildable from the other "lathe" type buildings solves that nicely without really removing the existence of a lose condition. If your power infrastructure goes out, you're still pretty well toast. If all your oxygen explodes and you have no reserves (and no access to oxite), you're toast. If all your water goes and you don't have reserves or ice, you're probably toast. If all your food and plants go (possibly _just_ if your plants go), you're probably toast (though that one takes long enough you could conceivably find a food or seed trader and get some food growing again). Still plenty of ways to hard lose, it just isn't such a pass/fail on a single building.
Bit overkill, imo, given how much other stuff got destroyed anyway. And the extra pipe bender mods he created. Those cost him far more resources than the autolathe would have, even once recycled. An autolathe is just 20 iron, 10 copper, and 2 gold, it's super cheap.
Flipsie I don’t know yet if you caught it, but both the pump in & pump out were on during the emptying phase of your new mixed gas cooling loop. Check at about 3:00:00.
You should create a storage room, imagine shoot coming from your production printer going directly into said room (without stackers) and the whole room would be filled with the item produced spread on the floor
I you use pumps at the exits of the filters, they should be linked by a single pipe between the filter and the pump to ensure the pumps empties the pipe every ticks. Also, if you know the pipe is always empty, you could, after turning off the filter, yield and turn off the "efficiency pumps".
In one of my playthroughs the other day, I noticed some of the plants require nitrogen (wheat specifically) where as I dont recall them requiring it before, so you may want verify (using your plant analyzer tablet is probably the easiest) the need, then figure a method to bleed in a bit until your filtration system is complete and can be tapped off of.
5:11:50 It varies from location and field, some places an antique would be 50+ years and some other places you wouldn't call anything younger than a 100 an antique.
Wish i were watching this live rather than a VOD so I could remind splitsie his greenhouse is chock full of O2, giving him ample time to build the proper system, instead of panicking at the start
Splitsie, you are better at programming than you think you are (which is true for most people). The pump control program is not very far off from an Arduino sketch, it's just that the loops are BASIC style rather than the C/C++ style used in Arduino. And like any other skill, the main reason why coders are fast at this is because they've practiced a lot and they recognize common patterns.
There is a nice simple grow light script...nothing fancy, it just turns them on for 10 mins then off 10. Works well. Soybeans seem to kak it if you do the slightest thing wrong.
Technically, you only need them off for 5 minutes, not 10. Genes can shift that a bit, but if you sorta split the difference, turning them on for ~12.5 minutes (1500 ticks) and off for ~7.5 minutes (900 ticks), you get better growth rate (light above the minimum required raises the Light Efficiency and thus the overall Growth Efficiency of the plant) without hurting the crops. I _believe_ genes can only shift that by 25%, though, so 6.25 minutes of darkness might be enough to handle even plants that manage to accidentally mutate to maximum darkness required.
Easier yes, but seldom better. When Volume pumps are controlled by IC, they are usually switched off most of the time or set to very low volumes. Since they use energy proportional to their volume setting, they can be very energy efficient. The pumps on my atmo system consume less the 1W. Also the volume of the regulator is dependent on how your pipe network is configured. The volume would be good in some places, but bad in others. Sure, you can manipulate that with one-way valves and such, but it's counterintuitive and easy to mess up.
I'm trying to remember why I put that there and I honestly don't recall if it was there for the basic furnace and then not replaced, or if I'd had some other idea :P
@3:42:00 Going to recommend again that you create a Configuration cartridge for your tablet. You can point it at any device that has a logic port and see every single logic variable it accepts or exposes, and their current values. It's _incredible_ for writing any sort of IC10 code, because you can very easily see what values are available, which can be written to, and what they do when you do things. For example, you could see what the names of the vars are for the input and two output pressures on the Filtration unit, and what their current values are, which also tells you what unit their in (kPa, in this case) by comparison to your Atmo cartridge (same for temperature, actually. Very obvious it's in Kelvin from that cartridge).
Almost completely agree with this advice, but it will not show you slot variable, that's why I will make a slot reader while writing code for an unfamiliar piece of equipment, it's an easy way to help write code when your trying to see what is available, since slot information is not very well documented.
Ya, the fact that it doesn't show slot vars is a serious annoyance mine. That said, Splitsie isn't likely to interact with slots vars until he tries automating Harvies, and the config card is still ridiculously handy regardless.
@@timothyloveridge4051 ah, ya, he did talk about using a script to read those. Though I think that was partly driven by his belief, prior to chat correcting him, that filter units with no or exhausted filters would pass everything to the filtered output, rather than everything to waste like they actually do. So reading the filter slots is really just a convenience thing to indicate exhausted filters, rather than a necessary thing. And given that heavy filters last for-freakin-ever, I personally don't see it as worth the effort. That said, an onboard script that auto-detects the gas being filtered by the unit based on the filter cans inserted can be very handy.
Not only should you stash a backup Autolathe somewhere, anywhere, you should build a second one to sacrifice to the gods of hacked save files. Seems only fair.
Hey Splitsie, With running your base air through the storage and filter system, couldn't you have a gas sensor on the pipe to detect nitrous and pollutant and the moment it does detect it, it turns off the normal base system and turns on a pump to direct it to the pollutant and nitrous filtration so it runs through that and once the pollutant and nitrous is no longer detected it automatically switches back to bypass base storage
Make another autolathe kit and put it somewhere safe, like on the shelves next to your seeds. You got caught by surprise this time, but now you know the problem exists, you have time, and you have the resources to avoid having to cheat in a solution a second time.
Not just the autolathe, also the parts you need to build the autolathe, so 2 iron sheets, 4 minimum cable coil, 2 plastic sheets, since those items also need to be created by the autolathe.
at about 30 ish minutes I realised that the heavy cable that was disconnected in the middle of the fabrication room was the grow lights for the greenhouse and I'm 50 mins in thinking that Splitsie is never going realise and the plants are doomed.
Should the cold filters not have pumps on their outputs? for both filtered and returned mixed gas? This will improve efficiencies all around. Basically what you did for your "in series" filters. Just because the others are in parallel and already having an efficiency hit, shouldn't mean they need to take a 2nd punch to the gut.
Oh, good point about the hot gas, I should really replace it with cold CO2 as soon as I can. I wasn't planning on locking and vacuuming out that room, but equally, I shouldn't need to go in there too often so I guess we'll see how I feel about it when I get to that point :)
If it helps with name batching from lbn or sbn commands, you need two hash IDs. One for the device itself and one for the exact name you gave the device. something like a volume pump you named "NOSPump": define VolumePump 12345678 #I dont know what a volume pump hash is and cant look it up but it is a simple copy from stationpedia and pasting. define NitrousPump HASH("NOSPump") sbn VolumePump NitrousPump ON 1 would reference every volume pump named "NOSPump" on the network and turn them on. Hopefully it helps for later so that you wont need as many pins. With a proper system, would next be water making? and personally I dont see a problem with an h2 combustor if you regulate the output and not have it attached to a blammo canister. But understandable if you want it in a separate room... just curious if you will use the filtration for cooling and getting water. I will say though that when you make water you will get some very hot water and gases that could double as furnace hot gases and could potentially replace your heater setup.
my pentium I5 4 core struggles with Stationeers and firefox open at the same time. calculations for pressure changes are noticible in airlocks taking much longer to do and videos on YT freeze but audio continues until i go into menus or pause the game. tabbing takes a while too. runs allright without a browser open. my CPU maxes out before my 16GB or RAM
While I disagree other lathes should be able to build an autolathe (there needs to be some kind of lose condition in a sandbox game), I was surprised to learn when I check just now that it isn't available from any of the appliance vendors either. Cheating in the autolathe kit seems legit in this case. If you want to "balance" it maybe run the credit card through the furnace to simulate an emergency premium humanitarian delivery or some such thing. Edit: Follow up thought, since this game is based on SS13 and you build specific lathes from a base machine frame in SS13 by adding a specific motherboard to them, maybe there's a place for the electronics lathe to be the replacement with a manually buildable machine frame.
@@chriswoodend2036 But now the question might become: If you only have the autolathe, how do you get the motherboard to replace an electronics printer?
Would not the "lose condition" be if _all_ of your crafting machines got blown up? It just seems super silly to me that losing your autolathe is a effectively a death sentence. Only Appliance Trader sell the kits to make new ones, and if you don't have a trading setup already built and running, you're just straight dead, period. I don't see any reason a sandbox game needs to have a complete "well I guess I just die" condition just from losing _one_ device. Like, if your entire crafting setup explodes, that's one thing, but right now, if you lose your autolathe for any reason, you're completely beholden to finding an Appliance Trader to be able to restart at all, and if you don't have trading set up yet, your save is dead, full stop, no recovery possible. It's utterly silly that the autolathe is _that_ make-or-break of a building. Adding a recipe for it to the other crafting buildings (probably not the oven, but certainly the pipe bender, electronics, rocket, tools, and security) seems just fine to me. You still have a distinct hard-lock lose condition, it just requires _all_ of your crafting facilities to explode, not just one specific one. Actually, the first thing I did after seeing the start of this stream was go mod in a copy of the autolathe recipe to the other crafting facilities for my local game.
@@KaedysKor The appliance trader will buy autolathes but they aren't on the trader sell tables. You literally cannot get a replacement autolathe without an autolathe.
@@Flipsie Interesting. I think it just looks a bit that way. When i was watching TFE, Nev, You, and the destructo 9000, Before that series went away, i remember seeing something that reminded me of watching someone in VR playing a FPS, or something similar. I don't Think i am describing this the way that makes sense. so forgive me.
@Splitsie I still think the right choice for recovery on this one is to roll back to a save before you placed the combustor and pretend it was all a nightmare. That said, you did learn to make a spare auto lathe and store it away from your active equipment.
Something I found about this game is that it has quite a few situations where recovery is impossible: - loosing the autolathe - loosing your suit - running out of oxygen/water/food - loosing one of your tools (only early game before you make the tool crafter)
I find it a pity that you couldn't rename your Channel 'Crapsie' just for this Episode...as general 'Aww crap' Moments go, this is Monumentally bad...Just as an Aside, I only play Permadeath Survival Where Possible...This happening to me would mean a Complete Restart...Sorry Buddy, But thems the Dice...:)...
The difference between "Mode" and "On" on filters only matters if using the internal IC slot. If you turn yourself off via "On", you can't turn yourself on again. But with "Mode" you can toggle between Processing/"Active" and "Idle" though.
![BOWKYLION - วิงวอน (ex-change) [Official MV]](http://i.ytimg.com/vi/cLdnZWDaO-w/mqdefault.jpg)
Cool guys, the reverse physcology has worked. We have now trained splitsie to avoid stackers and throw thongs on the floor muahahaha!
With the explosion blowing out the airlock door, it's interesting that the game didn't turn Splitsie into a grease stain inside your suit.
Sending IC audio alerts to your suit works in vacuum too. Your suit have internal speaker so you'll always hear it.
Recommendations on your IC code:
* There's VERY little reason to ever use the `sleep` command. IC10 scripts are very cheap on performance anyway, since they never run more than 128 commands per tick, and sleep can easily cause your program to miss things it should catch. For example, if you turn on your pump and then sleep for 3 seconds, that's 6 ticks that your script is not checking the pressure and could overpressure. Unlikely given the volumes involved, but it does mean that your script will never pump less than 6 ticks of gas into the pipe. If you're pulling from, say, 30 MPa, and pushing to a pipe network of ~600 L, that's 30 MPa of pressure you pumped in over those 6 ticks (or 3 MPA, if you're using a regular volume pump). If your output pipe were only, say, 300 L, oops, now you've hit 60 MPa in those 6 ticks. Using `yield` and evaluating pump state on every tick is almost always safer, and I guarantee you you won't notice a performance hit from it (I currently have at least 30 IC chips running every tick in my base and my FPS hasn't dipped at all, and your rig is rather better than mine).
* I recommend placing your `yield` directly after `start:`, rather before `j start`. Main reason is, that single yield now covers every single conditional branch back to start, rather than having to have a `yield` before every single one, which saves a lot of lines. In addition, it makes sure you don't _forget_ to yield, which can cause VERY weird errors in your script otherwise. The only difference it makes is that your script won't run until the tick after you turn on the IC housing, which makes basically zero difference.
* For "stateful" programs like your pump sequence, where the program is in one of several states and transitions between them (in this case, 3 states: pumping in, waiting for the gas to cool, pumping out), I find it's more robust to not try to keep state through the program, but instead _identify_ the state on each tick.
For example, you can figure out which of those three states you are currently in by _reading_ whether the pumps are _currently_ on, and what the pressure in the cooling pipe is. If the input pump is currently on and the pressure is less than your target, you're inputting, so keep doing that (ie. jump back to start). If the input pump is on and your pressure is equal or greater than your target, time to turn the pump off and start cooling. If neither pump is on, you're waiting for cooling, so check temp, and if it's at or below the target, turn on the output pump. And then, if the output pump is on and the pressure is above the minimum, you're outputting, keep going (jump back to start). If output pump is on and you're at or below minimum pressure, time to start inputting again, so turn off output and turn off input.
Now your program can handle whatever state the system happens to be in when it is turned on. It also means you can _manually_ flip one of the pumps to force the script to move to the next state (for example, flipping the output forces the script to consider itself outputting and check for minimum pressure to swap to inputting). Since the state of the code itself gets reset if you pull the IC10 chip to tweak the code, that also avoids the script being left in a weird state like having pressure in the pipe cooling and being reset to considering input (and thus never checking the temp or outputting). I don't _believe_ your program would run into issues when pulled mid-sequence, but avoiding specific states in the code itself helps avoid issues like that, especially as the sequence of things you're doing gets larger.
This is why when I start...I try and make a backup auto lathe either stored somewhere or setup at a different location
Rough lesson to learn. I’ve dealt with similar in the past with other games and my solution? After I get past my initial setup I leave that in place and build identical replacements in a facility nearby/another room/etc as soon as able/practable. Worst case, I can revert back to the original equipment.
Remember: “One is none. Two is one.” Always have a backup.
It might be worth putting an IC chip in all the cold gas filtration units so they automatically stop filtering if the tanks are getting too full. This would mostly benefit the CO2 tank, which will eventually fill up, even if the gas is being used for the greenhouse.
The hot gas side will probably need to have it integrated into the existing IC-housing code, since that will already be trying to turn the filtration units on and off.
And of distinct note, when using the onboard slot on a filter, disable filtering using the Mode var, NOT the On var. If you turn the unit off, the IC code can't run, so it also can't turn itself back on once pressure stops again. Setting Mode 0 has the same effect, the unit stops processing gas, but leaves the unit powered on so the IC code keeps running.
@3:57:00 - Name hashing is...complicated. Basically, you can either interact with a device _directly,_ using one of the device pins (on the IC housing or filtration or w/e), interact with all devices of that type globally (using the type hash from the Stationpedia), or interact with all devices of that type _with a specific name._ That third part is what chat was recommending. Using the name style, you need _two_ hashes, the device type hash (again, from the Stationpedia) and then a hash of the name. Now, the hash of the name is pretty obscure, so they included a handy function in the IC10 code for it: HASH("name"), which takes a string and converts it into the equivalent hash. So the idea is, you declare using `define` the _hashes_ of the _names_ at the start of the program. Example:
define TankName HASH("Tank Waste Phase 2")
define FilterName HASH("Filter Pollutants")
That stores the integer hash equivalent of the name string you gave it into that define. You then access it using the named version of the batch commands, `lbn` (load batch named) and `sbn` (set batch named). Unlike `lb` and `sb`, which just take a device type hash (ex. `1280378227` for a large insulated gas tank), you use _both_ the device type hash and that name hash var. Note that just like `lb`, you need to include a batch _mode_ (ex. Minimum, Maximum, Average) for `lbn`, even if it's just one device, because the game doesn't know if only one device or multiple will match that type and name combo. Example:
lbn r0 1280378227 TankName Pressure Average
1280378227 here is the type hash for large insulated tanks, from the Stationpedia. So that loads the average pressure across all large insulated tanks that have the name "Tank Waste Phase 2".
sbn -348054045 FilterName Mode 0
As above, -348054045 here is the type hash for Filtration units, from the Stationpedia. This sets the mode var of all Filtration units with the name "Filter Pollutants" to 0, which has the effect of disabling it from processing gas while leaving it powered on (this is actually important for on-board IC10s, because if the device is off, the IC10 can't run, so it also couldn't, for example, turn the Filtration unit back on if the output tank dropped in pressure. Using `Mode` instead allows you to shut off the gas processing/output without powering off the unit, which allows it to turn gas processing/output back on later).
The fundamental downside to coding IC10s like this is that you have to modify the code to point it at a different device. For example, you'd need to use a different version of the code for the Pollutants filter versus the Nitrous filter, since their output tank names are different. This is really the main benefit to using device pins, you can re-use the same code with different combinations of devices. On the flip side, named batch commands let you use batch commands on a more limited subset of devices. For example, put a pump on both outputs of the Filtration unit, name them the same, and then you can use `sbn` with that name (and their type hash) to turn _both_ on and off at once, without affecting any other volume pumps on that network, and without having to assign them to the limited 2 pins on that Filtration housing.
The bit that chat wasn't explaining was that I just needed to define the name at the top, then to do things, define the device type, then it's name from that definition later. At no point did I read anything that explained WHERE I was going to tell the thing that this named item is actually X type of device. This makes a little more sense now, thanks :)
@@Flipsie Ya, I suspected the issue was the confusion over "hash", since it's used in two completely difference senses with named hash functions. You're used to normal batch calls, so "hash" to you just meant the _type_ hash, and chat clearly wasn't explaining sufficiently that the names are _also_ turned into hashes (and that that is, in fact, the entire point of the HASH("name") function) >.
I would guess that if you went back and read chat at the time, you would be able to get more out of it than you were in the middle of streaming, it can always be a challenge to learn something fairly complicated with 10 different people trying to tell you how something works, all with limited time to try and write out a concise way of explaining how to do it
Also, anytime you use define, it needs to be on the initialized part of the program, it will give you an error if you try to define the same variable twice
And to clarify why you need to use the HASH("name"), is because ic10 programs do not store characters in any of the variables, it has to be a number, so the hash command will take a string and convert it to a number, so for your tank you would do
define WasteTankPh3 HASH("TankWastePhase3")
So now WasteTankPh3 would be the variable you would use as the device name.
If you were to do this
s dB Setting WasteTankPh3
When you point towards the ic housing it would show -275719632
That number is TankWastePhase3 converted to a HASH
I know that Splitsie has decided to do all of his cold stage filtering in parallel. But in case its efficiency needs to be improved, I think there's a very simple hands-off way to filter stuff in sequence (as long as you're fine with it not being a flow process). You set up all filtering units in a loop, with the "mixed" output of one piped to the input of the next, and the "mixed" output of the last piped to the input of the first. ("Pure" outputs from the filters obviously to their filtered tanks, and you can have the efficiency boosting pumps right after the filter outputs.) Then to each of those loop sections, you connect 2 things (on the output side of the efficiency boosting pump): A pump which feeds the mix from the mixed tank into each loop section. And an over-pressure valve (the one confusingly named "back-pressure regulator" in game) back to the mixed tank, to keep the pressure levels in all pipes safe.
You don't, and in fact _shouldn't,_ filter in sequence. As counter-intuitive as it sounds, you actually filter _faster_ from a common input pipe than in series, _and_ you avoid the risk of blowouts if one backs up and the need for intermediary tanks to buffer.
The key with Filtration units is that their filtration speed is based on the pressure _difference_ between the input and the _higher_ pressure of the two outputs. So connecting the waste output _directly_ back to the input sets the filtration speed at the floor rate, because there's no pressure differential. The solution is to use pumps to keep the outputs at 0 pressure.
And conveniently, the maximum pump rate of a normal Volume Pump is the same as the volume of a single pipe segment, 10 L, so if you build a single pipe segment off each output, then a normal Volume pump set to maximum, it will completely empty that pipe segment every tick, resulting in the maximum filtration rate. From there, it's just about keeping the input pressurized as high as you're comfortable (I generally keep mine pressurized to 45 MPa).
The result is very rapid filtration, _even for trace gases!_ And since you're not running anything in series, it doesn't matter if a filter backs up or shuts off because its destination tank is full, because it won't affect any of the other filter units. And since the filter canisters in the unit are only used up when they remove the target gas to the output (not when they cycle gas through to the waste output), it also doesn't use up the filters any faster. For that matter, it doesn't even use more power.
Also, on the note of backpressure regulators, keep in mind that their pump rate is _also_ based on pressure differential. They have a floor rate, but it's quite slow. If, for example, your mixed tank was quite high pressure, it's _very_ possible for your backpressure regulator to just flat out not be able to keep up with the waste output rate of filter it's supposed to be protecting. That's specifically why series filtering just doesn't work well. If one filter slows down or has to shut off because it's output is full, everything upstream also has to shut off, or one or more of the pipes are going to burst. In parallel, any filters shutting off doesn't affect any of the others.
One change to make bootstrapping possible would be to give you a tablet module/chip/[whatever it's called] that lets you call in a delivery of specific things. Obviously having it custom delivered would come at a premium (so that it's still cheaper to wait for a trader). This would let you get an autolathe to get you out of an effective softlock; but it could also be used if you want to buy a specific thing, like seeds or oxygen and don't want to wait for the respective trader.
Yeah, that would be a nice alternative
I was thinking more of an "emergency 3d printer" (in the tablet or in the suit) that could craft the autolathe (ad few other things) consuming 10x the resources.
@@hubertnnn The thing that makes the emergency thing running on money more forgiving is that you can go into debt. If you are missing at least one of the materials too, you'd have lost the run.
But in the end it probably comes down to what loss condition the developers like better.
I _suspect_ the reason the devs haven't is because making the autolathe itself isn't _technically_ enough. You also need 2 iron sheets, 2 cable coils, and 2 plastic sheets, none of which can be made without an autolathe, plus a screwdriver and a welding torch. That's why the sole other way of obtaining an autolathe, a package from a trader, include all of those things except the welding torch (it literally does include a screwdriver, though).
But imo, it's completely silly, because it's too black-and-white. If you lose your autolathe, even if it's really _just_ your autolathe you lose, like in Splisie's case (ok, he also lost the pipe bender, but the point is, he still had at least one production facility, and basically all of his storage and (non-gas) resources survived), you're completely beholden to getting an appliance trader or you're dead. If you lose your autolathe before building the parts for a landing pad, you're also straight hardlock dead, no recovery possible.
That's, imo, just silly. No other building is that make-or-break. Realistically, depending on the world, you could lose _everything_ except your autolathe and a power source (battery, solar, w/e), and either some metals (maybe in the autolathe) or an arc furnace for making more, and you could still conceivably survive. Yet you can be actively launching rockets into space and if a canister explodes in an unfortunate spot and takes out your autolathe, you're softlocked, and completely hardlocked if you don't have trading setup yet. There's no reason that building should be that pass/fail.
Having them buildable from the other "lathe" type buildings solves that nicely without really removing the existence of a lose condition. If your power infrastructure goes out, you're still pretty well toast. If all your oxygen explodes and you have no reserves (and no access to oxite), you're toast. If all your water goes and you don't have reserves or ice, you're probably toast. If all your food and plants go (possibly _just_ if your plants go), you're probably toast (though that one takes long enough you could conceivably find a food or seed trader and get some food growing again). Still plenty of ways to hard lose, it just isn't such a pass/fail on a single building.
Cheat in the autoleathe if needed, just remember to destroy an appropriate amount of supplies in exchange.
pretty sure he did that last part already.
Bit overkill, imo, given how much other stuff got destroyed anyway. And the extra pipe bender mods he created. Those cost him far more resources than the autolathe would have, even once recycled. An autolathe is just 20 iron, 10 copper, and 2 gold, it's super cheap.
@@Rikard_Nilsson I shouldn't have been taking a sip of my water while reading this pfffffttt 😂
Flipsie I don’t know yet if you caught it, but both the pump in & pump out were on during the emptying phase of your new mixed gas cooling loop. Check at about 3:00:00.
You should create a storage room, imagine shoot coming from your production printer going directly into said room (without stackers) and the whole room would be filled with the item produced spread on the floor
I you use pumps at the exits of the filters, they should be linked by a single pipe between the filter and the pump to ensure the pumps empties the pipe every ticks. Also, if you know the pipe is always empty, you could, after turning off the filter, yield and turn off the "efficiency pumps".
In one of my playthroughs the other day, I noticed some of the plants require nitrogen (wheat specifically) where as I dont recall them requiring it before, so you may want verify (using your plant analyzer tablet is probably the easiest) the need, then figure a method to bleed in a bit until your filtration system is complete and can be tapped off of.
Hey Splitise. Weekly tip: You can send audio warnings to your suit, the same way as a klaxton.
5:11:50 It varies from location and field, some places an antique would be 50+ years and some other places you wouldn't call anything younger than a 100 an antique.
Maybe add some pressure regulators between the stages of the cooling system so you can extend the loops without pressure going down
Wish i were watching this live rather than a VOD so I could remind splitsie his greenhouse is chock full of O2, giving him ample time to build the proper system, instead of panicking at the start
Splitsie, you are better at programming than you think you are (which is true for most people). The pump control program is not very far off from an Arduino sketch, it's just that the loops are BASIC style rather than the C/C++ style used in Arduino.
And like any other skill, the main reason why coders are fast at this is because they've practiced a lot and they recognize common patterns.
Hope Splitsie makes a ToDo list before he takes to long a break. So he doesn't forget what he's up to.
@Splitsie has upgraded the "free passive vent via broken pipe" to a "free high-volume passive vent via kerploded door".
There is a nice simple grow light script...nothing fancy, it just turns them on for 10 mins then off 10. Works well. Soybeans seem to kak it if you do the slightest thing wrong.
Technically, you only need them off for 5 minutes, not 10. Genes can shift that a bit, but if you sorta split the difference, turning them on for ~12.5 minutes (1500 ticks) and off for ~7.5 minutes (900 ticks), you get better growth rate (light above the minimum required raises the Light Efficiency and thus the overall Growth Efficiency of the plant) without hurting the crops. I _believe_ genes can only shift that by 25%, though, so 6.25 minutes of darkness might be enough to handle even plants that manage to accidentally mutate to maximum darkness required.
Judging from the wiki, the pressure regulators move 10 times as much gas as the volume pump (100
Easier yes, but seldom better.
When Volume pumps are controlled by IC, they are usually switched off most of the time or set to very low volumes. Since they use energy proportional to their volume setting, they can be very energy efficient. The pumps on my atmo system consume less the 1W.
Also the volume of the regulator is dependent on how your pipe network is configured. The volume would be good in some places, but bad in others. Sure, you can manipulate that with one-way valves and such, but it's counterintuitive and easy to mess up.
2 Options missed. 1. Mod in an autolathe to the other printers.(it's really the same process as SE XML modding.) 2. Quit and start over.
Also, you don't need a volume pump before the advanced furnace. It has a built in pump.
I'm trying to remember why I put that there and I honestly don't recall if it was there for the basic furnace and then not replaced, or if I'd had some other idea :P
@3:42:00 Going to recommend again that you create a Configuration cartridge for your tablet. You can point it at any device that has a logic port and see every single logic variable it accepts or exposes, and their current values. It's _incredible_ for writing any sort of IC10 code, because you can very easily see what values are available, which can be written to, and what they do when you do things. For example, you could see what the names of the vars are for the input and two output pressures on the Filtration unit, and what their current values are, which also tells you what unit their in (kPa, in this case) by comparison to your Atmo cartridge (same for temperature, actually. Very obvious it's in Kelvin from that cartridge).
Almost completely agree with this advice, but it will not show you slot variable, that's why I will make a slot reader while writing code for an unfamiliar piece of equipment, it's an easy way to help write code when your trying to see what is available, since slot information is not very well documented.
Ya, the fact that it doesn't show slot vars is a serious annoyance mine. That said, Splitsie isn't likely to interact with slots vars until he tries automating Harvies, and the config card is still ridiculously handy regardless.
@@KaedysKor he will need it to read the filter levels also
@@timothyloveridge4051 ah, ya, he did talk about using a script to read those. Though I think that was partly driven by his belief, prior to chat correcting him, that filter units with no or exhausted filters would pass everything to the filtered output, rather than everything to waste like they actually do. So reading the filter slots is really just a convenience thing to indicate exhausted filters, rather than a necessary thing. And given that heavy filters last for-freakin-ever, I personally don't see it as worth the effort.
That said, an onboard script that auto-detects the gas being filtered by the unit based on the filter cans inserted can be very handy.
The RBLF feels sorry for all the Patients the Splitsie probably exploded when they were hooked up to Oxygen. Poor patients… 😂😂😂😂😂
Not only should you stash a backup Autolathe somewhere, anywhere, you should build a second one to sacrifice to the gods of hacked save files. Seems only fair.
Splits your base is a plumbers nightmare ! 😂
Hey Splitsie, With running your base air through the storage and filter system, couldn't you have a gas sensor on the pipe to detect nitrous and pollutant and the moment it does detect it, it turns off the normal base system and turns on a pump to direct it to the pollutant and nitrous filtration so it runs through that and once the pollutant and nitrous is no longer detected it automatically switches back to bypass base storage
Make another autolathe kit and put it somewhere safe, like on the shelves next to your seeds. You got caught by surprise this time, but now you know the problem exists, you have time, and you have the resources to avoid having to cheat in a solution a second time.
Not just the autolathe, also the parts you need to build the autolathe, so 2 iron sheets, 4 minimum cable coil, 2 plastic sheets, since those items also need to be created by the autolathe.
you will need a wastewatertank and treatmant at some point in that room.
I got to say you could use the labeler to set your settings instead of clicking multiple times
at about 30 ish minutes I realised that the heavy cable that was disconnected in the middle of the fabrication room was the grow lights for the greenhouse and I'm 50 mins in thinking that Splitsie is never going realise and the plants are doomed.
wooo someone realised and told Splitsie about 10 minutes after I posted
Yup, was lucky I didn't miss that message 😂
Natural Selection in action, Darwin is smiling down upon you.
Should the cold filters not have pumps on their outputs? for both filtered and returned mixed gas?
This will improve efficiencies all around.
Basically what you did for your "in series" filters.
Just because the others are in parallel and already having an efficiency hit, shouldn't mean they need to take a 2nd punch to the gut.
That recovery was easier than expected. I guess you didn't burn that much oxygen after all? Next time, make bigger explosions! 😉
still need to make the grass room too
Gonna put a large vent in your gas handling room and airlock it?
That CO2 jetpack is like a portable grill atm.
Oh, good point about the hot gas, I should really replace it with cold CO2 as soon as I can. I wasn't planning on locking and vacuuming out that room, but equally, I shouldn't need to go in there too often so I guess we'll see how I feel about it when I get to that point :)
@@Flipsie one big issue is your main cooling loop for heat exchange is only pressured at 320Kpa so you could add a lot of capacity to help it cool.
If it helps with name batching from lbn or sbn commands, you need two hash IDs. One for the device itself and one for the exact name you gave the device.
something like a volume pump you named "NOSPump":
define VolumePump 12345678 #I dont know what a volume pump hash is and cant look it up but it is a simple copy from stationpedia and pasting.
define NitrousPump HASH("NOSPump")
sbn VolumePump NitrousPump ON 1
would reference every volume pump named "NOSPump" on the network and turn them on. Hopefully it helps for later so that you wont need as many pins.
With a proper system, would next be water making? and personally I dont see a problem with an h2 combustor if you regulate the output and not have it attached to a blammo canister. But understandable if you want it in a separate room... just curious if you will use the filtration for cooling and getting water. I will say though that when you make water you will get some very hot water and gases that could double as furnace hot gases and could potentially replace your heater setup.
what about making a spare tank in a separate area, that stores and extra supply of oxygen?
Am guessing different plants need different light timings. Probably why my poor soybeans die. Oh well.
36:00 I have not said "No it's not, no it's not" so many times!
I know I should pay more attention to the programming bits for when I eventually play this but I think I’d rather just be chill
lol I'm not sure I'm all that helpful on that front, maybe on what not to do :D
Did you ever shut off the grow-lights again after you overrode their automatic control?
The automated shut off should still function normally when it gets to the next shut off time 🙂
my pentium I5 4 core struggles with Stationeers and firefox open at the same time. calculations for pressure changes are noticible in airlocks taking much longer to do and videos on YT freeze but audio continues until i go into menus or pause the game. tabbing takes a while too. runs allright without a browser open. my CPU maxes out before my 16GB or RAM
You did better at keeping calm during your explosive moment.
I would have been a lot more annoyed, and it would be full of swear words.
I agree lol
While I disagree other lathes should be able to build an autolathe (there needs to be some kind of lose condition in a sandbox game), I was surprised to learn when I check just now that it isn't available from any of the appliance vendors either. Cheating in the autolathe kit seems legit in this case. If you want to "balance" it maybe run the credit card through the furnace to simulate an emergency premium humanitarian delivery or some such thing.
Edit: Follow up thought, since this game is based on SS13 and you build specific lathes from a base machine frame in SS13 by adding a specific motherboard to them, maybe there's a place for the electronics lathe to be the replacement with a manually buildable machine frame.
Well, as long as you have the electronics printer, you have one machine frame in it.
@@Pystro That is a solid point
@@chriswoodend2036 But now the question might become: If you only have the autolathe, how do you get the motherboard to replace an electronics printer?
Would not the "lose condition" be if _all_ of your crafting machines got blown up? It just seems super silly to me that losing your autolathe is a effectively a death sentence. Only Appliance Trader sell the kits to make new ones, and if you don't have a trading setup already built and running, you're just straight dead, period. I don't see any reason a sandbox game needs to have a complete "well I guess I just die" condition just from losing _one_ device. Like, if your entire crafting setup explodes, that's one thing, but right now, if you lose your autolathe for any reason, you're completely beholden to finding an Appliance Trader to be able to restart at all, and if you don't have trading set up yet, your save is dead, full stop, no recovery possible. It's utterly silly that the autolathe is _that_ make-or-break of a building.
Adding a recipe for it to the other crafting buildings (probably not the oven, but certainly the pipe bender, electronics, rocket, tools, and security) seems just fine to me. You still have a distinct hard-lock lose condition, it just requires _all_ of your crafting facilities to explode, not just one specific one. Actually, the first thing I did after seeing the start of this stream was go mod in a copy of the autolathe recipe to the other crafting facilities for my local game.
@@KaedysKor The appliance trader will buy autolathes but they aren't on the trader sell tables. You literally cannot get a replacement autolathe without an autolathe.
This game gives me strong VR vibes. I sit VR only? or just feels that way watching it?
I don't think it has VR at all
@@Flipsie Interesting. I think it just looks a bit that way. When i was watching TFE, Nev, You, and the destructo 9000, Before that series went away, i remember seeing something that reminded me of watching someone in VR playing a FPS, or something similar. I don't Think i am describing this the way that makes sense. so forgive me.
was looking foreard to stationeers next week
Who read Marsian? Who is reminded about storyplots?😅
In theory could you cool a base using a sterling engine?
No, as you'd need the cool substrate to begin with so the cooling has to be done somewhere
@@Flipsie I was more so thinking on a colder planet like Mars or Europa you could use planetary Atmo for the cool substrate.
Okay, try not to cry, try not to cry, try not to cry...
Okay, I failed...
YOU GOT THIS SPLITSIE!
I'm sorry Splitsie, could you please tell me what was the name of that new game? Our Kin? Arc In? Archeon?
@Splitsie I still think the right choice for recovery on this one is to roll back to a save before you placed the combustor and pretend it was all a nightmare.
That said, you did learn to make a spare auto lathe and store it away from your active equipment.
Something I found about this game is that it has quite a few situations where recovery is impossible:
- loosing the autolathe
- loosing your suit
- running out of oxygen/water/food
- loosing one of your tools (only early game before you make the tool crafter)
I find it a pity that you couldn't rename your Channel 'Crapsie' just for this Episode...as general 'Aww crap' Moments go, this is Monumentally bad...Just as an Aside, I only play Permadeath Survival Where Possible...This happening to me would mean a Complete Restart...Sorry Buddy, But thems the Dice...:)...
The difference between "Mode" and "On" on filters only matters if using the internal IC slot. If you turn yourself off via "On", you can't turn yourself on again. But with "Mode" you can toggle between Processing/"Active" and "Idle" though.
Ah, that makes sense 🙂