Buffer (seconds) get stuck

[tribal wedge]

Often I build something using buffers and it works perfectly fine for like 20 minutes, for example, but then buffers just randomly stop activating. I even try to connect them directly to a switch or a button, but they do not work until I break them and put them again. I don’t know if it’s known thing.

https://youtu.be/4e6Sry4Dlgg
Here. It does not work until break it and put a new one.

https://youtu.be/qnPVNx407nU
This one is not exactly how I described it, but that's an example of it freezing until I put a new one. Weird things going on all the time with the buffer in my worlds.

[simple iron]

If you’re able to upload a save that easily reproduces the issue that would also help the devs

[tribal wedge]

Sure, I’ll upload one later today

[tribal wedge]

Where do I find a save file?

[tribal wedge]

https://youtu.be/qnPVNx407nU

https://youtu.be/4e6Sry4Dlgg

[mossy phoenix]

AppData/Local/Brickadia/Saved/Worlds

to find appdata, press Win+R, type %APPDATA% and press enter

[turbid plank]

this happens when the gate has to wait to send a signal and it receives a second signal before sending the first one.
after sending the first signal the gate will get blocked by the left of the original time from the second signal.

so example for config of 10 seconds:

0s - 1 pulse - gate waits 10s - no output
6s - 1 pulse - gate waits 10s - no output
10s - gate sends output from 0s
11s - 1 pulse - gates ignore
16s - no output from 6s
17s - 1 pulse - no output
21s - no output from 11s
27s - gate sends output from 17s

because of the pulse at 6s the gate locks and ignores everything until the 16th second

this should solve the problem before updates

[timber elk]

normally I just do a NOT from the buffer into an AND gate along with the pulse triggering the buffer

Seems to fix it

[mossy phoenix]

buffer -> NAND is a faster/uses-less-gates way to do this

[timber elk]

tru

[turbid plank]

that doesn't work...

the first signal arrives the and gate is still on to bug the gate until the gate clicks

and the extra click is only prevented for one tick, and the exact tick it resets '-'

[turbid plank]

this is the correct way of solving it

your solution solves
on immediately - off with delay
that doesn't bug as far as I tested

the bug I was referring to happens when:
wait off - sends signal with delay
that not gate stays on the whole time...

[turbid plank]

Yeah I just tested again, this is a clock, it doesn't bug at all and the videos refer to a different problem

I was also talking about a different bug so

[turbid plank]

clock needs to be before edge detector, put a tick clock configured as 0, 2

ok I tested again and that's not needed. What actually made it really work for me is this

only default minigame triggers the round start gate

[tribal wedge]

my mini-game is default as well, you can see on the vid

[tribal wedge]

how do you do that?

[turbid plank]

gate configured to 20 s

[turbid plank]

don't mind it in my map I actually had racing conditions

but that on the video is expected behaviour

[timber elk]

Nah, pretty sure I’m right

The buffer gate is broken in the video from what I can tell

It bugs out when it receives a signal when it is already sending signal right? So the NAND basically says- only send a signal if it’s not already sending a signal

Granted, this does introduce another problem, where the signal won’t get reset on minigame resent. But I haven’t been able to fix that without a buffer array

Also- why are you guys putting edge detectors on the exec gate? I thought those didn’t really matter

Not on exec at least, they are 1 tick anyways

[timber elk]

This isn’t a bug from what I can tell though. Would be nice if we had a way to kill the buffer gate early.

[turbid plank]

doesn't solve

[turbid plank]

this is the only "problem" and it is the expected behaviour

[turbid plank]

happens because of this

read what I sai in chat on the video

[timber elk]

It’s two separate problems, the first problem in OP is because buffer receives signal early and locks the gate up. The second is that when you attempt to solve this, it will instead occasionally skip an input.

Like I said though, you can fix this with a buffer array and a reasonable round reset timer. It’s annoying though, not sure if there’s another way to fix it

I guess you could make a manual buffer with a 1 second loop increment, that’s also obtuse lol

[turbid plank]

a buffer array works, but you need logic to choose what buffer to send the signal and block signal from previous buffers. the next problem is if you use all the buffers, and the next solution is add more hahaha

[turbid plank]

it's the other way around, the situation I placed locks the gate, which is the bug.
in every situation the OP said the signal is ignored, I didn't solve anything... I just said timers have the ignoring feature as default.

there actually can be another default for timers which is they replicate the original signal after time X; but that's not the version we have