I suppose you could call this "dynamic water logic" because it relies on flow instead of pressure. I was expecting hydraulic relay valves ("static water logic"), which are widespread in industrial control applications, and perhaps even more commonly, the automotive automatic transmission:
Almost all the material I could find about computing with water seems to focus on dynamic gates, essentially completely ignoring the century-old[1] knowledge of hydraulics that's been present all along. Perhaps this shows just how isolated the different disciplines can be.
[1] I have a book from the early 1900s about water-operated elevators, complete with descriptions of all the valving required to control the hydraulic ram. They were logic gates, before they were even called logic gates!
I want to build a megalithic computer using a river flowing down a mountain. The AND gate is a enormous door lifted by an enormous arm attached to an enormous block hollowed out from the bottom so that it can float.
Input is done by almost lifted doors that can be moved by applying enough man power.
i don't think this would 'explain' a logic gate to anyone who doesn't already understand a logic gate. this model doesn't convey any information that would help someone reason about a logic gate. If you didn't know anything about logic gates and you saw this, how would it improve your intuition? If I imagine these water streams in my head, does it help me understand how, why or what logic gates do?
i learned logic gates playing mods for games I was playing... I could have a motion sensor and see when I NOT'd it it made its signal the opposite. I could AND two of my sensors and see that I now had a signal combining both. I think this kind of thing, a model that simulates how logic gates are used gives a much better intuitive understanding of what they're for and how to use them.
I have zero physics background, and this is something I've never understand. I knew what a logic gate did, but the core idea of "how does a gate just know whether it's this AND this" always seemed like magic.
I had a serious lightbulb moment with this video. Although, I had to download it and slowly scrub through several times.
I still don't actually understand how this would work with electricity, but I can see how it's broadly possible.
This doesn't help you see how it works with electricity, because it's not really how it's done with electricity.
Put two valves after each other on a pipe. The water can only flow if the first valve AND the second valve are open.
Fork off a pipe into two pipes with a tee, put a valve on each side, and connect the other ends of the valves together with another tee. Water flows through the whole thing if the first valve OR the second valve is open (or both).
That's how it works with electricity too. The valves are themselves controlled with water/electricity too.
Armed with this information, you should try to understand voltage, resistance, relays, and transistors, in roughly that order. That may help to bridge the gap from the water gates here, to transistors which are solid state; you can't see them moving, but they largely accomplish the same basic function and can be used to create the same kinds of gates using just electricity.
Disagree. I once had a conversation with a sibling who was trying to understand logic gates, and I ended up using a verbal explanation of a very similar setup (in my example the AND gate was a pipe with a hole in the bottom just big enough to drain one but not both of the potential incoming streams, and another hole partway up leading into an output pipe). I think pipes can provide a useful example that seems more real and tangible than semiconductors. As an aside, I've heard/read more than one source use water as an analogy for electricity -- I think it can help people reason about, for example, friction impeding rate of flow before they're comfortable enough thinking about electrons that this would make sense intuitively.
“Logic gate” can (and does) refer to the idealized concept, and is already disctinctly different from the implementation, e.g., an electronic logic circuit. Gates can be implemented with relays, physical switches or water.
> You can build logic gates out of just about anything.
Does that fact (which I never denied) in any way provide the original "pipes" picture any explanatory power, helping somebody to more understand the how the logic gates actually used in our devices function?
I still don't see that any answer here up to now disproved my original claim: "what is happening in the actual logic gates [meaning the gates that are actually used in our devices] has no similarity with that [the pipes thing] at all -- so any "intuition" one might hope to gain from these two pipes will be wrong."
I still claim that.
The "pipes" don't "explain" anything but demonstrate their own functioning.
TTL or CMOS aspects are also not relevant for this discussion, contrary to what "neuralRiot" writes.
Friend, I'm not saying you were wrong or trying to argue with you. Just pointing out that different people learn differently, and it's good to have different ways to illustrate and explain things.
You never know when one kind of explanation may give someone an "aha" moment that they didn't get somewhere else. And sometimes it's good to do things like this just for fun.
For example, my very favorite visualizations of sorting algorithms are the Hungarian dancers:
My intuitions about the world (and I guess most everyone's) roughly divide it into two kinds of things: agents and inanimate objects. It's most obvious and natural then to see a computer as a kind of agent, built out of smaller dumber sub-agents. (There have been psych studies showing young kids often include computers in the 'living things' bin, so I don't think this is just me here.) In the end with this intuition you get logic gates as particularly simpleminded agents.
Maybe this is just me, and it feels stupid to admit it, but I felt a definite shift of perspective on really seeing logic gates as inanimate. I had a related shift about F=ma around the same time, which is hard to articulate without sounding even stupider. You want to think of it like "you push, and the object accelerates" when from this nonanimate point of view there is no you causing things, there's only a network of constraints.
CMOS logic is actually really simple and easy to understand (from a functional model of ideal CMOS transistors). Indeed a water analogue shouldn't be necessary if one wants to understand modern gates themselves.
Which is not how actual gates work, if you’re gonna get nitpicky since most of of them are composed of CMOS transistors. You can see a “real” AND gate internal circuit here:http://www.ti.com/lit/ds/schs057c/schs057c.pdf
http://www.cogpro.com/chapters/E-CruiseOMatic/images/E-Cruis...
Almost all the material I could find about computing with water seems to focus on dynamic gates, essentially completely ignoring the century-old[1] knowledge of hydraulics that's been present all along. Perhaps this shows just how isolated the different disciplines can be.
[1] I have a book from the early 1900s about water-operated elevators, complete with descriptions of all the valving required to control the hydraulic ram. They were logic gates, before they were even called logic gates!