Anyone good with AE subnetworks?

[malicious_bloke]

I don't think that's really an exploit. It just seems like sensible cable management to me.

I have one "trunk" cable that I can use to service remote locations, then calve P2P tunnels off it, it lets me have up to 1024 channels wherever I feed the cable off to.

It's an absolute godsend when I'm dealing with hundreds of applications in a dozen or more dimensions.

To those who have the strength of will to use them, Colin grants such gifts...

 

[KingTriaxx]

Two-fold benefit to the carrier network: 1) I'm not spending any channels on the main network for the tunnels, which means more channels for functional blocks. I'm aware that channels are basically unlimited, but space on the sides of the controller isn't. 2) It also means I can keep the main network with much lower power than the Carrier. Cables use power, and if you have to run a long way, it's cheaper on power and materials to run a single cable. Eventually you reach the point where the bridge is even more efficient, but that's a lot of power, and a lot of cables. And even at that point because you can run multiple colors of cable through the bridge, I leave that on a 'Carrier' network.

That way even if the carrier network loses power, I usually still have the main network with all it's storage and auto-crafting with power enough to get the others running again.

 

[kris449]

Two-fold benefit to the carrier network: 1) I'm not spending any channels on the main network for the tunnels, which means more channels for functional blocks. I'm aware that channels are basically unlimited, but space on the sides of the controller isn't. 2) It also means I can keep the main network with much lower power than the Carrier. Cables use power, and if you have to run a long way, it's cheaper on power and materials to run a single cable. Eventually you reach the point where the bridge is even more efficient, but that's a lot of power, and a lot of cables. And even at that point because you can run multiple colors of cable through the bridge, I leave that on a 'Carrier' network.

That way even if the carrier network loses power, I usually still have the main network with all it's storage and auto-crafting with power enough to get the others running again.

To be honest I thought you were suggesting something like this:

But yeah that clearly doesn't work

I built my original network off a tutorial so I had a very specific idea of what a "subnet" is; everyone seems to have a slightly different way to set them up though.

 

[KingTriaxx]

Yes and no. For organizational purposes, that works perfectly fine. You can actually put those right beside each other, and then the color coding keeps them separate, so you know which one is what kind of p2p at a glance.

Basically you don't need the colored cables. All three types come in through the green cable. It terminates in three different p2p tunnels. (You cannot pass an me p2p through an me p2p. It's strange, but there you go.

 

[GreenZombie]

This is the P2P channel exploit:

Non of this is a subnet. It is one big network. Purpose of this is to avoid running several different Dense Cables around your base and just use one(white) as carrier. AFAIK this was not an intended feature, but it quickly gained popularity among people who did not like the added Channel mechanic.

I don't get that this is an exploit. Building a carrier network without this exploit would look exactly the same, but with 1 more controller dedicated to the isolating the carrier.

But, if we were actually worried about controllers, the network shown only needs 1 anyway

 

[rhn]

I don't get that this is an exploit. Building a carrier network without this exploit would look exactly the same, but with 1 more controller dedicated to the isolating the carrier.

But, if we were actually worried about controllers, the network shown only needs 1 anyway

I personally see it as an exploit since it wasn't originally intended, and it completely circumvents the channel mechanic that was supposed to make the mod less of a faceroll. And make the mod interesting and challenging to use, instead of just instant win.
From what I have heard(terrible source I know) the author of the mod disabled the ME P2P tunnels by default until a fix could be made. But the users wanted the exploit and turned it back on. So I suppose by popular demand it has just become a feature now.

Me P2P tunnels also doesn't really make sense to me with the channel mechanic(in AE2). You need these huge cables to handle the bandwith to your devices? But a P2P tunnel just magically compresses 32 channels down to one? If this is possible, why don't all the wires use this technology by default then?

 

[GreenZombie]

I personally see it as an exploit since it wasn't originally intended, and it completely circumvents the channel mechanic that was supposed to make the mod less of a faceroll. And make the mod interesting and challenging to use, instead of just instant win.
From what I have heard(terrible source I know) the author of the mod disabled the ME P2P tunnels by default until a fix could be made. But the users wanted the exploit and turned it back on. So I suppose by popular demand it has just become a feature now.

Me P2P tunnels also doesn't really make sense to me with the channel mechanic(in AE2). You need these huge cables to handle the bandwith to your devices? But a P2P tunnel just magically compresses 32 channels down to one? If this is possible, why don't all the wires use this technology by default then?

But its not circumventing the channel mechanic: You need the same number of P2P interfaces - wired the same way - to carry the data that a non exploit version of the network requires.

I think we can agree that the channel mechanic was poorly considered, and mostly rewards the wrong kind of player behavior (i.e. encourages the players to create cpu intensive builds rather than rewarding efficiency). And I agree totally on the channel compression "magic" - it really does make no sense.

 

[rhn]

But its not circumventing the channel mechanic: You need the same number of P2P interfaces - wired the same way - to carry the data that a non exploit version of the network requires.

You don't need to think about wiring logistics, space for wires, cost of long distance wiring etc. It totally removes the benefit versus cost/work challenge of the channels, not encouraging you to use many of the cool mod combos and compatibilities to save channels.

 

[GreenZombie]

You don't need to think about wiring logistics, space for wires, cost of long distance wiring etc. It totally removes the benefit versus cost/work challenge of the channels, not encouraging you to use many of the cool mod combos and compatibilities to save channels.

I'm curious then if this would be considered exploitative use of p2p tunnels. Because this - to me - is your 'p2p channel exploit network - but done they way that channels are fully intended to be used.

And it does not add any additional considerations wrt logistics, space, cost, etc.

 

[malicious_bloke]

That is the basic essence of how all my networks end up.

Except I run the trunk route off the same ME Controller as the rest, I just make mine a huge cube and put it in the center of The Shrine of Colin as was intended.

 

[rhn]

I'm curious then if this would be considered exploitative use of p2p tunnels. Because this - to me - is your 'p2p channel exploit network - but done they way that channels are fully intended to be used.

And it does not add any additional considerations wrt logistics, space, cost, etc.

It is all just a personal opinion. And I feel that if I can run wires from the controller to where I need them, then I do so.
Only place I have used the P2P ME tunnels, is cross Quantum network bridges(32 channel bandwith limit). To avoid having 5-10 Quantum bridges between the same two points, or running 5-10 dense cables for hundreds/thousands of blocks(and chunkloading that whole stretch). And that still feels wrong to me.

I would personally prefer if ME P2P tunnels used the amount of channels that it tunnelled, or some similar downside to give more consideration about using them instead of just a way to simplify wiring. That would bring the P2P tunnels back to more what they were in AE1.

 

[KingTriaxx]

Perhaps boosting the cost depending on the type of cable run into them?

 

[kris449]

Yes and no. For organizational purposes, that works perfectly fine. You can actually put those right beside each other, and then the color coding keeps them separate, so you know which one is what kind of p2p at a glance.

Basically you don't need the colored cables. All three types come in through the green cable. It terminates in three different p2p tunnels. (You cannot pass an me p2p through an me p2p. It's strange, but there you go.

I know how it all works, just liked the color coding. The point I was trying to make is that I thought you were saying you could wire the p2ps without using any channels from the "main" network. I tried it with a quartz fiber also, but either way I wasn't expecting it to work. Sorry, bad with words.

As to whether it's an exploit or not.. I'ma stay out of that one. Really I thought it was all just to save resource cost of cabling.

 

[KingTriaxx]

Ah. Well, yes, you can, because p2p doesn't cost a channel to have something run through it. Only the two channels for the devices on the carrier network. So you run a Dense cable with 32 free channels in, what comes out on the otherside has 32 free channels available.

Only the line connecting the two p2p connections uses channels.

 

[kris449]

Sooo off topic a little but since you friendly folks have been helping me with AE stuff so much, has anyone ever tried setting up this ASIMA, or anything like it? I'm really fascinated by that.

 

[KingTriaxx]

I'd never seen it before, but it takes a ludicrous amount of power, so not me.

 

[Inaeo]

Sooo off topic a little but since you friendly folks have been helping me with AE stuff so much, has anyone ever tried setting up this ASIMA, or anything like it? I'm really fascinated by that.

I'm always tempted to build a large scale automatic mining machine, but I usually come to the conclusion that my PC won't handle it. Not to mention the influx of resources we mostly just end up sitting in storage, so I may as well not have spent the effort to collect them. I do love the engineering challenge that it represents, as well as having production lines to process it all without bottlenecks, however, so I'm always tempted.

Over my Evolution as a player, I've moved from a mindset of "Must collect all the things!" to "Must use all the things!". If I have 4k wheat sitting around, it's actually worse than if I had never collected it in the first place. If I can turn it into a usable commodity (food, fuel for several mods power gen, breeding animals for other needed resources) I'm OK with having a buffer stock, so long as it isn't getting out of hand.

Going back to the heart of the thread, I try to keep my ME Storage to a minimum (usually a Drawer wall on a Storage Bus) and save all the actual networking for processing.