Rocket launch shouldn't be able to destroy science

[melichor]

There are several cases where something in factorio can be destroyed:
1) Weapons (duh, that's kinda their purpose)
2) Biters, well, that's a game mechanic
3) Collisions, yeah, same thing basically
4) Mining production facility before it's done with current recipe, or if it has active fuel in it. This is okay, kinda, I understand the technical problems with it, and it's even kinda intuitive, because if you mine before it's done, you don't get the ingredients and you don't even get the text, that you got them, so it's something that player can find out
5) Mining anything fluid related, this is kinda obvious too
6) Letting Rocket launch infinitely. If you launch a rocket, while it has over 1000 science inside it, it will cap at 2000 science inside and truncate rest of it. It basically destroyed the resources that were put inside, for no reason. This is so unintuitive, and I personally destroyed like 25k space science before I figured it out. I know how to solve this in the game, but I didn't know I have to solve this. I had no idea that this is a problem. I assumed the rocket wouldn't launch, if there's no more space for the science, pretty much like it is with every other production facility.

My suggestions?
1) Put there an option, pretty much like there is with "auto launch with cargo"
2) Disallow launches, while it has >= 1000 space science inside

[urza99814]

Wow...is that really how it works?? I've played over 3,000 hours in this game and I still didn't know that. Wonder how much I've lost because of it...

It'd be nice if there was some kind of indicator of such things, since it's pretty unintuitive for random structures to just work differently -- Nuclear reactors have the same issue, although I knew about that one from the FFF posts, and the new heatpipe visualizations will help show that as well. But I don't see how anyone would know that rocket launches behave that way too...

If that's actually how rockets behave (haven't had time yet to confirm it myself), then I'd say that's the biggest UI problem that I've seen in this game...

[conn11]

Space science stacks 2000 packs, with one stack per silo. A rocket launch gives you 1000 so only a third launch without drainage is deleting anything. A steel chest holds 48 stacks of anything, therefore only your 99th unobtained rocket launch is deleting science packs, this is of course if you are not using active providers or some other mean of transport.
Should you really require such a specific feature, the circuit network is your friend...

[kbk]

Space science stacks 2000 packs, with one stack per silo. A rocket launch gives you 1000 so only a third launch without drainage is deleting anything. A steel chest holds 48 stacks of anything, therefore only your 99th unobtained rocket launch is deleting science packs, this is of course if you are not using active provider.
Should you really require such a specific feature, the circuit network is your friend...

Unfortunately, this "build more storage" approach doesn't really work
On most of my attempts I utilise a dumb "Science secondary bus" approach to science pack output, which involves a bus of sciencepack-filled belts parallel to main bus but driven backwards to labs, which are behind the start of main bus. Time after time I forget to fill the tech queue and end up with full belts. Even if I'd bother doing some buffer storage, a ~8k SPM-capable megabase fills up a chest every 12 minutes or so, and 10 chests only allow me a two hours of happy-go-lucky immersion in some other stuff around such a busy base before the moment I yell "Oh poop I forgot to turn science on again" at "Press T ..." button. Two hours is not that big of a time span in Factorio terms so yes, circuits are the only way.

Because the subject is in fact a design decision, one can only hope that endgame transition will become richer some day, and building that produces space packs will get separated from silo

[conn11]

Unfortunately, this "build more storage" approach doesn't really work
On most of my attempts I utilise a dumb "Science secondary bus" approach to science pack output, which involves a bus of sciencepack-filled belts parallel to main bus but driven backwards to labs, which are behind the start of main bus. Time after time I forget to fill the tech queue and end up with full belts. Even if I'd bother doing some buffer storage, a ~8k SPM-capable megabase fills up a chest every 12 minutes or so, and 10 chests only allow me a two hours of happy-go-lucky immersion in some other stuff around such a busy base before the moment I yell "Oh poop I forgot to turn science on again" at "Press T ..." button. Two hours is not that big of a time span in Factorio terms so yes, circuits are the only way.

Because the subject is in fact a design decision, one can only hope that endgame transition will become richer some day, and building that produces space packs will get separated from silo

Sure, but with a 8 SPM Megabase is it really that much to ask to use one decider combinator per silo outputting for example to the satellite input inserter to stop unwanted launches? The op gives a working solution of a possible circuit condition.
I don’t see the value for a broader playerbase, outside the megabase territory storage is trivial and you can always toggle the autolaunch feature. Don‘t leave your most advanced (Vanilla) production line completely unattended or without working setup, seems like a reasonable choice of game philosophy.
IMO with space science production the silo has actual worth above a fancy looking win the game button.

[Mylon]

Use an active provider chest. If space science ends up clogging your storage, then you have the least of all possible problems to worry about.

[Serenity]

Use an active provider chest. If space science ends up clogging your storage, then you have the least of all possible problems to worry about.

Or as said, buffer it in regular chests and only insert the satellite when there are enough free slots. That works very well with the science bus approach.

[kbk]

Sure, but with a 8 SPM Megabase is it really that much to ask to use one decider combinator per silo outputting for example to the satellite input inserter to stop unwanted launches? The op gives a working solution of a possible circuit condition.
I don’t see the value for a broader playerbase, outside the megabase territory storage is trivial and you can always toggle the autolaunch feature. Don‘t leave your most advanced (Vanilla) production line completely unattended or without working setup, seems like a reasonable choice of game philosophy.
IMO with space science production the silo has actual worth above a fancy looking win the game button.

In my opinion, the reason this topic exists is that silo introduces lots of new stuff gameplay-wise while being internally a miscommunicated jenga tower of whatever is labeled a design decision. What op suggests is hardcode what is otherwise doable with simple circuit logic into a yet another hack to fix what just really have been miscommunicated. This has already been given a "Won't fix" resolution in the bug section, so that's one of the reasons why I thought of moving the problematic part of silo elsewhere instead. Doing so should allow for better communication of silo mechanics, and that does not necessarily mean reducing the silo to a one-armed bandit.

For example, that entire science pack output might be moved to a some kind of landing pad for a new reentry vehicle which is launched at silo as a payload to collect some scientific data (or any other applicable internal resource, similar to how rocket parts never leave the silo). To address communication issues, the vehicle should return with only as much data as the pad can absorb any given moment. Data is then transformed to science packs. And, to make things harder, before you can launch such a vehicle there might be a need to maintain an orbital fleet of satellites, which have a small chance to break, to deorbit and even to fall on ground with each passing gametick. And then there might also be a probability for unsuccessful launch which depends on the satellite fleet size and quality etc etc

[Koub]

What bothers me with the silo is that it's the only building to actually destroy stuff when output is full. And there is no warning that I know (tbh I didn't even try to achieve science pack loss, the first time I produced space science, I had already seen posts about that, and took my precautions).

[conn11]

In my opinion, the reason this topic exists is that silo introduces lots of new stuff gameplay-wise while being internally a miscommunicated jenga tower of whatever is labeled a design decision. What op suggests is hardcode what is otherwise doable with simple circuit logic into a yet another hack to fix what just really have been miscommunicated. This has already been given a "Won't fix" resolution in the bug section, so that's one of the reasons why I thought of moving the problematic part of silo elsewhere instead. Doing so should allow for better communication of silo mechanics, and that does not necessarily mean reducing the silo to a one-armed bandit.

For example, that entire science pack output might be moved to a some kind of landing pad for a new reentry vehicle which is launched at silo as a payload to collect some scientific data (or any other applicable internal resource, similar to how rocket parts never leave the silo). To address communication issues, the vehicle should return with only as much data as the pad can absorb any given moment. Data is then transformed to science packs. And, to make things harder, before you can launch such a vehicle there might be a need to maintain an orbital fleet of satellites, which have a small chance to break, to deorbit and even to fall on ground with each passing gametick. And then there might also be a probability for unsuccessful launch which depends on the satellite fleet size and quality etc etc

Should a post 1.0 release feature offer some additional space exploration content, a special "receiver building" might be justified, currently it would only separate space science production from the silo without fixing the product destruction issue.
I agree with Koub, that it's a UI related topic. A popping up info box warning about possible science pack overflow on the auto-launch checkbox would communicate it properly. Furthermore the use of some absolute basic circuit conditions is a good solution to allow for some hyper specific conditions without overcrowding the base Vanilla interface. Encouraging players to use them early on, helps keeping more complex, but fun processes (like AOP) in game in the long term.
On the other hand the current dev choice of having this one special building punishing players for not paying attention is also a perfectly valid option, space science stacks most generously after all.

[kbk]

Should a post 1.0 release feature offer some additional space exploration content, a special "receiver building" might be justified, currently it would only separate space science production from the silo without fixing the product destruction issue.

Unfortunately, yes. That's because I'm convinced that this product destruction issue is one specific part of the novelties around the silo I've mentioned earlier, and it is not going to get fixed:

Hm, this seems very, very unfortunate, that the game just quietly destroys my rather expensive resources, without even making it obvious that this is happening. Out of interest, what is the argument or reason that this isn't considered a bug but a feature?

Because bugs are things that aren't working as we intended them to. In this case I explicitly programmed it to work that way.

I do hope, however, that any kind of big enough addition to space science might help change this as well.

[Optera]

What bothers me with the silo is that it's the only building to actually destroy stuff when output is full. And there is no warning that I know (tbh I didn't even try to achieve science pack loss, the first time I produced space science, I had already seen posts about that, and took my precautions).

It only bothers me for the same reason.
Everything until space science conveys to players that it stops producing if one output is full.

For basic oil processing that was an issue large enough to dumb down the recipe and break many builds, but rocket silo seems like no one wants to touch it since it has so many special scripts running.

My solution is simple, feed from silo to an output chest with stack inserter, wire the chest to the inserter placing satellites and have that only operate when there's less than x in the chest.

[ssilk]

I don't see the need:
After start output space science to a (provider) chest with a stack inserter. Now automate inserter: input the satelite to rocket only, if there are less than x space science in storage (logistic network or circuits depends on your taste). Automation completed.

[bobucles]

I don't think anyone's saying that an overflowing rocket silo can't be solved. It's just weird because it doesn't fit with the design principle of "safe failure" seen in the rest of the game. When a machine overflows, it stalls. When an arm can't deposit, it waits. When a train can't pass, it stops. When nuclear power overheats, it doesn't explode. Nothing else in the game has a direct punishment for failure, yet for some reason the rocket silo destroys your science if it isn't exported in time.

Destroying rocket science doesn't seem like a feature either. It doesn't seem to help with anything(does it?), there's no direct way to turn it off, and rocket science is a key ingredient for the post-game infinite research so you can't really have too much.

[Lubricus]

I don't see the need:
After start output space science to a (provider) chest with a stack inserter. Now automate inserter: input the satelite to rocket only, if there are less than x space science in storage (logistic network or circuits depends on your taste). Automation completed.

It's not a problem if you know about it. And it seems like there is lot of players don't know it.

[ssilk]

You need just to watch it. Like everything else in Factorio...

And even me learns new things for Factorio every week.

[Optera]

I don't see the need:
After start output space science to a (provider) chest with a stack inserter. Now automate inserter: input the satelite to rocket only, if there are less than x space science in storage (logistic network or circuits depends on your taste). Automation completed.

0.17.60 made a massive oil change with the reasoning of blocked outputs being an insurmountable hurdle for new players.
In light of that having blocked outputs destroy products is a much bigger and worse learning experience.

Wubes official stance has to be consistent.
If BOP is too complex for dumb newb players then Rocket silo mechanics are as well. I still think new players are smart enough to figure out blocked outputs in time for bop and dont have to be held by the hand until aop.

[planetmaker]

I believe the OP has quite a valid point: the game behaves counter-intuitively in this case (even when it's reasonably edge-case). One would expect the rocket to simply not launch if the science pack storage inside the launch facility is full.
Practical solution: allow the rocket to start and produce all the science it usually does when the science storage is <2k - but don't cap the science storage at 2k (effectively allowing it to store up to 2999 space science packs). This solution would not stop anyone winning the game and it would make the game consistent without changing much of how things work.

All other workarounds with circuit logic or provider chests etc of course do their job - but they are just that: a workaround around an counter-intuitive part of behaviour. This is also true when this counter-intuitive behaviour happens only in late game and doesn't affect many people.

(In essence I wonder how much science I might have lost myself due to the exact same reasons as the OP. Nice find! )

[conn11]

I believe the OP has quite a valid point: the game behaves counter-intuitively in this case (even when it's reasonably edge-case). One would expect the rocket to simply not launch if the science pack storage inside the launch facility is full.
Practical solution: allow the rocket to start and produce all the science it usually does when the science storage is <2k - but don't cap the science storage at 2k (effectively allowing it to store up to 2999 space science packs). This solution would not stop anyone winning the game and it would make the game consistent without changing much of how things work.

All other workarounds with circuit logic or provider chests etc of course do their job - but they are just that: a workaround around an counter-intuitive part of behaviour. This is also true when this counter-intuitive behaviour happens only in late game and doesn't affect many people.

(In essence I wonder how much science I might have lost myself due to the exact same reasons as the OP. Nice find! )

I disagree, utilising the circuit network isn't a woarkaround, but integral to a possible playstyle. In most cases even one active provider chest (or alternativly atrain wagon) will likley prevent any science destruction, considerung the huge stack size of space science. Should a player continue after the rocket launch, he is usually approaching megabasing territory after a while. In those scenarios the circuit network is a very helpful and powerful tool, not to say mandatory. Encouraging the usgae "early" on in a factory just capable of porducing space science isn't that bad of a design. Although it's debatable if players should receive a warning before they turn auto-launch on. Enhancing the UI seems to be a completely adequate solution.

[planetmaker]

I disagree, utilising the circuit network isn't a woarkaround, but integral to a possible playstyle. In most cases even one active provider chest (or alternativly atrain wagon) will likley prevent any science destruction, considerung the huge stack size of space science. Should a player continue after the rocket launch, he is usually approaching megabasing territory after a while. In those scenarios the circuit network is a very helpful and powerful tool, not to say mandatory. Encouraging the usgae "early" on in a factory just capable of porducing space science isn't that bad of a design. Although it's debatable if players should receive a warning before they turn auto-launch on. Enhancing the UI seems to be a completely adequate solution.

I'm not arguing against usage of circuit networks in the least. Yet - as argued in the OP - the behaviour is inconsistent and against expectations (thus I consider the suggested things workarounds). And it is not good means to "encourage" usage circuit networks this way at this place in a totally unobvious way IMHO. Power usage is much more suited to that task.