Logic Memory Units Have A Broken Functionality

[Fridge Lord]

I am trying to use logic writers to set the value of a logic memory unit. The purpose of the circuit in the test case is to get a wall cooler to come on at -10C, cool to -15C, then turn off until temps reach -10C again.

To do this we need a memory chip to "remember" the 0 value once temps reach -15C, and stop the wall cooler. So, trigger a 1 on -10C, remember that 1, trigger a 0 on -15C, and hold that 0 again.

You need 2 logic writers to write the 1 and the 0 to the malfunctioning memory chip in a stable fashion.

Once you use one, you cannot use the other one. It's as if there is an override or the signal is hanging in the network, even if you turn the first logic writer off.

Eventually the memory chip will just break and neither writer will write a setting to it anymore.

Please look into this.

[Fridge Lord]

So, upon further tests, if you have 2 logic writers writing to the same memory chip, they will eventually break the memory. Powering them off and on will not cause them to change the memory's state.

However, if you use the screwdriver and click one of the writer's bottom right knobs for send "Setting" to memory, it will send the data and change the state of the memory chip. So something is going on there.

[VegasGoat]

Yeah as far as I can tell right now, the logic writers only send the value when it changes. Turning off and on again does not rewrite the value. I don't know if this is intended or it's a bug. I came across this when trying to have a set of buttons that would each write a different value to a pressure regulator when pressed.

[BrazenBran]

I've seen this as well with this current patch. I don't remember having the same issue in previous patches but it's hard to recall exacty as I have been swapping between saves and re-building etc.

If it is actually the intention that you can't have to writers on the same memory chip at the same time I think unfortunately its another example of the current circuitry system requiring way too many chips to acomplish simple tasks.

[Myst Leissa]

Originally posted by BrazenBran:

I've seen this as well with this current patch. I don't remember having the same issue in previous patches but it's hard to recall exacty as I have been swapping between saves and re-building etc.

If it is actually the intention that you can't have to writers on the same memory chip at the same time I think unfortunately its another example of the current circuitry system requiring way too many chips to acomplish simple tasks.

I /think/ Memory Chips are only supposed to be changeable from the screwdriver - the problem with this setup assumes a logic writer can write to a memory chip - when in reality that is a read-only connection (despite having 2 data ports)

As a Workaround try using the static memory chips in combination with Processor Units: Use the memory values for constants and compare them against the reader values.

[Fridge Lord]

You can get around it if you use a Logic Select to trigger the respective logic writers writing to the same memory chip. There may be other solutions as well.

It has something to do with the logic writer changing its state. If it stays static, it will not write properly. For example clicking the output knob of the writer to "setting" again makes it send the value.

[BrazenBran]

Not being able to write to memory chips would be a glaring issue as you could only have constants and no variables.

[fugger_matte_der_Held]

Hi all...
I was looking a bit arround for known bugs and found this thread - looks like my problem (unclear if bug or featur) fits in here
My setup:
- 1st LogicMemory with value 0 set (supposed to be static)
- 2nd Logic Memory (supposed to be variable) in this case I set it to random value |=0 i.e. value=10
- Logic writer reading from 1st Mem writing to 2nd mem

Target:
the setup as above was a simplified one to hunt the problem down
my final Target was: reseting 2nd mem to the value stored in 1st mem

Problem i face:
it looks, like the LogicWriter expects any change/update on its input-side to "activate" the output a second time

Testings I did:
1.
- setup as explained above with logicWriter turned off
- switching on the logic writer will perfectly read value=0 from 1st Mem and write this value=0 to 2nd mem
- I use Screwdriver to manually change the value=0 of 2nd mem to 1
- value=1 is now stored in 2nd Mem and does not change
- fine so far (even if it was expected by me that the Logic writer will write the value=0 to 2nd Mem again
2.
- setup as explained above with logicWriter turned off
- switching on the logic writer will perfectly read value=0 from 1st Mem and write this value=0 to 2nd mem
- switching off the LogicWriter
- I use Screwdriver to manually change the value=0 of 2nd mem to 1
- Switching on the LogicWriter
- nothing happens (unexpected - i thought powering off and on the LogicWriter would reset it and reread the value=0 from 2nd Mem and rewrite value=0 to first mem)
3.4.5. and so on...
I tested with different input devices as logic switches, an autolathe ArcFurnace and so on
as long as the value of inbut is not changing no further writing happens


i figured out that there are 2 things which can be done to make the Value=0 is written to 1st Mem a second time:
1. using screwdriver to "turn" the output-var-screw on LogicWriter (there is just one option which is "Setting") - this will make the LogicWriter is getting reseted and write the value=0 to 1st Mem which was read from 2ndMem
2. I use the Screwdriver to manualy set 2nd Mem to anything (except 0) and then set it back with screwdriver to value=0

so my guessing: the LogicWriter will write the value from input to the output one time
to make it writing a second time it expects any update/change on the input-side
in general. this behavior is not too bad, as it enables me to use multiple LogicWriters writing to the same LogicMemory - I'd call it a feature :-) (also this is most likely good for performance issues?!?!)

What I think is a bug, is the fact that due pwercycling the LogigWriter (turning off and on again) it is not getting reseted... Also interesting is and most likely is connected to the described issue:
lets say I have an input value=21 on the LogicWriter - pointing with mouse on the LogicWriter will show me this 21. If I switch off the LogicWriter it will still continue to show the "21"...

oh and sure with batchWriter it is not possible to write to a single mem, but the behavior in general seems to be exactly the same :-(

hell what a bug report :-) hope this is helpfull to you guys

Edit: the issue/case described happend on NON-dedicated server