15:00 The problem with the long pipe (as stated in the paragraph above the video) is that the flow is not predictable. If you look at the green bars you can see that although it trends downward there are many spots were the fluid level randomly spikes up.
The chunk boundary problem is an issue I completely forgot about and it causes even more weird effects if there is a pipe junction on a chunk boundary.
37:51 - heheheheee, this reminds me of v=9hTosCUTmck :) [a Blue Man Group thing... it's about teh Intarnet, right?? Well...] 38:42 - I think, JD, that you're underestimating how likely it is for people to be utterly confused by that situation. Especially in more complex builds, but even in this one that's on screen. Myself, I don't think I'd be _confused_ by it, per se, but I'd likely find it off-putting. I agree, though, that if it's an animation that's perpendicular to the pipe or some other way indicating rate without appearing to flow along the pipe, that might be fine. I'm not sure what that'd look like, though, so now I'm even second-guessing that. Anyway, thanks as always.
This really seems over complicating something that gets used a lot. Add all inputs for a given section; cap it at max throughput. Now subtract all outputs. Doing per tick math on every single pipe on the map = no ups in endgame. Can do some basic -%10 of max every 20 pipes past some set distance from last input.
the example of the fuel etc about balancing out its so wrong... you need both fuel and oxidiser for the plataform to move.. wich gives you meteorites wich gives you ice wich gives you fuel and oxidiser... if you have more fuel than oxidiser you cant move the plataform.. you wont move... done so insted of having a fluid system that just work this guy wants an inserter and circuit to bypass something bad. -.- my god.. let the other guy have his own opinion..
Perhaps a current non issue (though while it's true that there are effective work arounds that does not mean it isn't a problem for everyone else without the knowledge for the work arounds) Who knows that you might need it on a pipe only planet or whatever not possible in the current system. New system seems quite simple Entire pipeline is effectively one tank, insert speed is unlimited to the trank, withdrawal speed is based on fill level, location and distances are irrelevant. Pumps are one way gates between tanks that drains one "tank" to fill the next.
Floating point rounding error. UPS. Ugly underground pipe to underground pipe chain. Tanks balancing using circuits at train station. Yeah, there are no problems.
Except it sounds like it became an issue: "However, during our playtesting of Space Age, it became clear that the existing system would no longer do the job. The production scaling afforded by the new machines and quality brought the flow algorithm to its knees." So... they did something with it.
15:00 The problem with the long pipe (as stated in the paragraph above the video) is that the flow is not predictable. If you look at the green bars you can see that although it trends downward there are many spots were the fluid level randomly spikes up.
The chunk boundary problem is an issue I completely forgot about and it causes even more weird effects if there is a pipe junction on a chunk boundary.
Yeah loving the improved fluid flow for megabaseing. And almost no sloshing. God that destroyed flow rate in BIG builds
a solid episode of FFF JD! Good job!
Hah. I see what you did there.
If you run pipes in parallel, don't you end up with a parallelepiped?
So… if I want higher throughput, I just need to maintain then segment full-ish?
37:51 - heheheheee, this reminds me of v=9hTosCUTmck :) [a Blue Man Group thing... it's about teh Intarnet, right?? Well...]
38:42 - I think, JD, that you're underestimating how likely it is for people to be utterly confused by that situation. Especially in more complex builds, but even in this one that's on screen. Myself, I don't think I'd be _confused_ by it, per se, but I'd likely find it off-putting. I agree, though, that if it's an animation that's perpendicular to the pipe or some other way indicating rate without appearing to flow along the pipe, that might be fine. I'm not sure what that'd look like, though, so now I'm even second-guessing that.
Anyway, thanks as always.
This really seems over complicating something that gets used a lot.
Add all inputs for a given section; cap it at max throughput.
Now subtract all outputs.
Doing per tick math on every single pipe on the map = no ups in endgame.
Can do some basic -%10 of max every 20 pipes past some set distance from last input.
Hope someone can make a mod to do this for the current version but i have my doubts
the example of the fuel etc about balancing out its so wrong... you need both fuel and oxidiser for the plataform to move.. wich gives you meteorites wich gives you ice wich gives you fuel and oxidiser... if you have more fuel than oxidiser you cant move the plataform.. you wont move... done
so insted of having a fluid system that just work this guy wants an inserter and circuit to bypass something bad. -.-
my god.. let the other guy have his own opinion..
this whole fluid nonsense is solving a non-issue
Perhaps a current non issue (though while it's true that there are effective work arounds that does not mean it isn't a problem for everyone else without the knowledge for the work arounds)
Who knows that you might need it on a pipe only planet or whatever not possible in the current system.
New system seems quite simple
Entire pipeline is effectively one tank, insert speed is unlimited to the trank, withdrawal speed is based on fill level, location and distances are irrelevant.
Pumps are one way gates between tanks that drains one "tank" to fill the next.
Floating point rounding error. UPS. Ugly underground pipe to underground pipe chain. Tanks balancing using circuits at train station. Yeah, there are no problems.
Except it sounds like it became an issue: "However, during our playtesting of Space Age, it became clear that the existing system would no longer do the job. The production scaling afforded by the new machines and quality brought the flow algorithm to its knees." So... they did something with it.
I justant to know when I can try it