projRed redAlloyWire & insulatedWire output problem with CC commanded toggleLatch

[moTechPlz]

I have been tinkering with the computerCraft interface using openComputers to change the output state of a toggleLatch as you suggested when i asked for wrenchable dip switches. I've made this test chip: When i invert the output through the CC interface with the setGateProperty command. The outputs of the IC does change but a projRed.redAlloyWire or insulatedWire connected does not change with it. This problem can be circumvented by either;

• connect to the output with projRed framedWire, bundeledWire, logic gate,nullCell etc
• use redstoneDust or redLogic.alloyWire etc (can't test bluePower)
• have a inverter, buffer, andCell etc between the toggleLach and the IOBit

[Victorious3]

My sane side's telling me that I should just remove that feature as it is like playing Russian Roulette.

Changing the internal flags of a gate can cause all sorts of different weirdness as the only thing I can do is calling a loose neighbor update, that can have all kinds of bad consequences. You might have to accept your workaround, however, that the other types of wires interact properly is weird.

I'll have a look at this.

[evg-zhabotinsky]

Right now, it should be easy to allow CAD-like manipulation of ICs by CC. The real question is, how much sense would it make? It is not FPGA after all! Most changes you can make to gate meta are no less intrusive than changing gate IDs (like rotation, mode, and maybe even delay). Others are just internal bits, like cached inputs or outputs or "state of that input 1 tick ago" (used for edge detection). Only few properties don't fall under above 2 categories. I think they are latch states, maybe randomizer output. Resetting timers or tripping pulse formers might also make sense, but it will have to be implemented separately.

Also, IMO being able to query IC internals using nearby computer is unrealistic. If it is not FPGA, that is. And I guess it is not an FPGA.