2-bit and 4-bit bit splitter and bit maker

[agile iron]

Pretty self explanatory. There is a lot of situations (including the main campaign) when you need to split a wire with much less than 8 bits.

[chrome edge]

Can't you use indexers in campaign?

[agile iron]

In levels like "arithmetic logic unit" or "instruction decoder" you cannot. Beside, this isn't really what indexes are for. They are to give you access to arbitrary bits, not to get all bits in order. I mean, you can use them to simulate splitter (not merger) but is't too much setup and it makes the circuit harder to read. I would rather have too long splitter.

Since we already have byte versions of splitters in 3 sizes, if would be fitting to also have 3 bit versions. (Or they should be literally the same components but that's another issue.)

[charred hollow]

I've wanted to make a 4-bit computer, I think it would make it easier without having to using the 8-bit wires

But I haven't tried the alpha branch so idk if the recent updates solve my problem

[worthy onyx]

I really hope we can see 4-bit components, since 4-bit operations are pretty common.

eg. the Z80 may be 8-bit, but the Z80 ALU is 4-bit

[wraith igloo]

Those components exist in the alpha (if it's about 4 x 1b <-> 4b and 2 x 1b <-> 2b, those were added recently. The green ones with configurable bit width already existed for a longer time.)
Logic and math gates with configurable bit width also exist in the alpha.

[worthy onyx]

I just noticed that, although, I've noticed a lack of bidirectional pins, those going to make a return?

Disregard: Just read the output pin is the new bi-directional pin, makes sense

[elder wadi]

This is a thing in the [save_breaker] alpha already.

In fact, this was done, because the decoder components were changed
to have a singular N bit inputs instead of having multiple 1 bit inputs
(for the selection / address input, the disable pin is still there).

[agile iron]

I don't think they were there when I was making the request. Since they are implemented already, let's shift the goalpost. We want a 3-bit version now.

[violet lance]

you get powers of 2, gyou don't have to use all the pins on it.

[frigid verge]

An AND will always be false if you leave a pin unconnected; a NAND will always be true.

[violet lance]

Those are irrelevant to the FR, why bring them up?

[frigid verge]

Misremembered what thread I was looking at, sorry

[mighty hornet]

in alpha there are concatenators that will create any bit width you like by just hooking up the appropriate wires (the concatenators themselves still have 2, 4 or 8 inputs but those inputs are "magic" and can produce any output wire width if you use them right)... on the splitter end you've got indexers which can pull out any bit (or run of bits)... just stack three of those on top of each other... the fixed-width splitters and makers are effectively only useful for cosmetic purposes at this point (eg: if you want exactly 8 bits, using an 8-bit maker might be better than a concatenator just because its more explicitly obvious what's happening when you come back to the schematic a month later)

[agile iron]

Yeah, you don't need to use all pins but it's hard to fit 2 3-bit splitters and one 2-bit splitter next to byte maker if you have to substiture 2-bit with 4-bit splitters.
3-bit signals shows very often in instruction decoders.