Factorio Forums

Interesting, though I have to wonder if it's better than simply joining belts, with only one side filled, with a splitter?

the idea from what I gathered is it splits the two lines and then merges the four lines into two mixed, then combines the two mixed. Could file under a buffer style merge

Skellitor301 wrote:

pyrolytic_tungsten wrote:Shouldn't this thread be stickied or something?

That is up to the mods

Well, I still mean, the right place for this is in the wiki and not forum, but I see also, that this is useful, so I made it sticky.

Well, that's *how* it does it, but the inferred end goal was the combination of two items on two belts heading in opposite directions, into one belt that heads perpendicularly to the original two, with a throughput as high as possible.
*shrug* still useful in any case.

ssilk wrote:

Skellitor301 wrote:

pyrolytic_tungsten wrote:Shouldn't this thread be stickied or something?

That is up to the mods

Well, I still mean, the right place for this is in the wiki and not forum, but I see also, that this is useful, so I made it sticky.

Thanks

and I do agree that being on a wiki would be helpful, I just cant work on that and this one, which is why if anyone wants to start a wiki version they are more than welcome to

n9103 wrote:Well, that's *how* it does it, but the inferred end goal was the combination of two items on two belts heading in opposite directions, into one belt that heads perpendicularly to the original two, with a throughput as high as possible.
*shrug* still useful in any case.

No doubt, though I gotta say it would make game play a bit more interesting if you do this instead of just merging. if not for efficiency, then just for a complex appeal

Updated:

• Fixed a few errors I found in the list
• Added the Open Wall Defense Layout

The "Open Wall Defense" seems like a bigger PitA to get back in through, than it is to just go through a wall or two, (since there's a lot more that you have to take down for re-entry.)

Skellitor301 wrote:if anyone wants to start a wiki version they are more than welcome to

https://forums.factorio.com/wiki/inde ... %26_Tricks

I'll answer any questions in the wiki board.

n9103 wrote:The "Open Wall Defense" seems like a bigger PitA to get back in through, than it is to just go through a wall or two, (since there's a lot more that you have to take down for re-entry.)

I would just use normal belts honestly, express belts would be difficult. But hey, looks cool

Skellitor301 wrote:

n9103 wrote:Well, that's *how* it does it, but the inferred end goal was the combination of two items on two belts heading in opposite directions, into one belt that heads perpendicularly to the original two, with a throughput as high as possible.
*shrug* still useful in any case.

No doubt, though I gotta say it would make game play a bit more interesting if you do this instead of just merging. if not for efficiency, then just for a complex appeal

I have 44 assembler 3s making advanced circuits on 1 line. The last time I added assemblers I was having issues with electronic circuits and plastic not getting to the end of the line with the simple merging I had done previously. I suppose if the the belts feeding the merge were perfectly packed on the bottom side (see screenshot) it would have been fine but that isn't practical. Assuming the merged belt moves up, items on the bottom side of the incoming belt will merge onto the belt blocking the items on the top side from getting on. Small gaps on the bottom side of the incoming belt won't be big enough to let items from the top side of the belt to merge on resulting in small gaps on the merged belt. With slower belts you can just have the belt at the merge be 1 speed faster but that isn't an option here. I suppose the other alternatives would have been to route new belts along the outside of the line and use long inserters or set up a whole new line but this seemed to be the best option. The animation I made was of a more idealized setup since the real thing is a bit more messy with unneeded optimizations, odd belt routing, and abandoned belts.

Personally, I always move to a 2nd belt if I find that I can't get enough through on a fully compacted fast belt.
But since you're not doing direct insertion of copper wires (tsk tsk) I guess that's not an option without rebuilding the whole setup elsewhere.
That's something to consider though, I almost always have logistics going pretty well by the time I'm using so many advanced circuits that a full red won't supply them anymore. Then I'm in the process of a modular logistics setup for the circuits at least, if not plastics as well. Don't suppose that's an option?

n9103 wrote:Personally, I always move to a 2nd belt if I find that I can't get enough through on a fully compacted fast belt.
But since you're not doing direct insertion of copper wires (tsk tsk) I guess that's not an option without rebuilding the whole setup elsewhere.
That's something to consider though, I almost always have logistics going pretty well by the time I'm using so many advanced circuits that a full red won't supply them anymore. Then I'm in the process of a modular logistics setup for the circuits at least, if not plastics as well. Don't suppose that's an option?

My current advanced circuits line meets my needs and could probably take a few more assemblers on the end before the belts saturate. The line actually predates getting robots researched. I don't really see a problem with copper wire on a belt for advanced circuits given that each copper wire assembler currently feeds 11 red circuit assemblers. I'm not sure how well that would work if I wasn't using Dytech's very fast inserters. That said wire on a belt is a bad idea for green circuits since it takes 1.5 wire assemblers to feed a green circuit assembler and my green circuit line is packed with productivity modules and speed beacons. If you look in the upper left corner of the screenshot in my previous post you can see some of my green circuit line.

I use mostly belts with robots only handling finished goods, processing modules, and a small amount of other stuff. I use production lines branching off the central buses which expand well until you hit belt saturation. My main copper and iron lines are pairs of fast belts except in a few areas but a 3rd fast copper belt +furnaces would be a good move since it sometimes isn't enough. Sure it isn't as modular as what can be done with logistic bots but that's fine by me. Admittedly I'm at the point in the game where there is nothing left to research and I have enough modules for 10 rocket defense.

Direct insertion for CU Cable to Green circuits is advantageous over belt insertion (you need about the full capacity of 3 CU Cable Assemblers for each pair of Green Circuit Assemblers), and CU Cable to belt to Red Circuits is advantageous since each CU Cable Assembler can feed about 8 Advanced Circuit Assemblers (very, very slow). If you need to poke, you could tsk Tungsten for having Superfast inserters directly inserting into the Circuit assemblers since, with a stacking bonus, standard inserters almost keep up (at least until Assembler Level 7s replace the level 3s).

DerivePi wrote: If you need to poke, you could tsk Tungsten for having Superfast inserters directly inserting into the Circuit assemblers since, with a stacking bonus, standard inserters almost keep up (at least until Assembler Level 7s replace the level 3s).

That doesn't actually work for my setup even though I have all techs researched. The circuit assemblers are running at +40% or +90% speed with +40% productivity and I did my test with a pair of the +40% speed assemblers. Regular inserters between the wire and circuit assemblers make them pause frequently, fast inserters cause an occasional very brief pause, and very fast inserters cause no pausing at all. I don't currently have any mods installed that add additional assemblers or stack size bonuses. My only dytech mod is dytech inserters.

ssilk wrote:

Skellitor301 wrote:if anyone wants to start a wiki version they are more than welcome to

https://forums.factorio.com/wiki/inde ... %26_Tricks

I'll answer any questions in the wiki board.

I am go to put it up.

Edit: done that you think?

Could use some cleaner images. Several of the examples have excess machinery that isn't involved in the example.
A wiki should present a moderate to high degree of professionalism, while forums are mostly conversational and casual.
But in the end, that's just my opinion.

<ssilk applauds loud to Lee_newsom>


@n9103: Yeah, well, wiki needs better pics. But currently I don't care much about that. The current task is
- create content
- order content
- make content nicer readable

anywhere after that there are things like pics, typography, layout, etc.

Skellitor301 wrote: It's a good idea to put rail signals on the parts of a track that you will traffic across often, this will give you a heads up if a train is coming and spare you the annoyance of becoming railkill

I do not think this is a good tip at all. Signals introduce more blocks and complications. The only time you can place one when you feel like it is on track you know will only have one way traffic, else it will mess up.

pyrolytic_tungsten wrote:How to join items onto 2 sides of an express belt faster than a T joining intersection.

I have a request.

Could someone do this, but for items coming from one side?

(So turning two parallel belts with items A and B into one belt running perpendicularly to them with item A on one side and item B on the other, with both parallel belts continuing.)

I have a couple designs but thus far none of them work well for express belts that aren't absurdly large.

Guess it depends on your definition of absurdly large, and how dense you're expecting the output.
An express splitter outputting the upstream side onto a fast belt leading into an express belt being fed by the downstream side will have better density than a simple T, but there's room for improvement. This wouldn't be exceptionally large, as you would really just need to do a couple splitters and tunnels to get the feeds to oppose each other.