This is my small blueprint library for effortlessly creating laser defense walls. It would be nice if you have comments or suggestions for improvement.
I put special emphasis on it being modular and flexible and still being simple. It needs only 3 different basic layouts for the whole wall including corners. You can create straight walls, rounded corners (inner and outer), and diagonal walls. All blueprints can (and need to) be rotated and mirrored to get all variations of horizontal and vertical combinations in connection with corners and diagonal walls. No rework required to patch wall parts together.
It's based on the roboport grid size of 50x50.
Blueprint book:
If the components are provided via the logistic network, the wall will build itself, so to speak.
Wall components:
straight wall - for building horizontal and vertical walls
corner (short) - the corner that will connect a horizontal with a vertical wall to form an inside corner
corner (long) - the corner that will connect a horizontal with a vertical wall to form an outside corner
diagonal - the combination of one inner and one outer corner, which magically forms a diagonal wall. You achieve the same by alternating inner and outer corners. This diagonal wall is a just pair of both corners with a nicer laser turret spacing. It's the only blueprint that cannot be mirrored without breaking the layout, so don't mirror it.
Screenshots:
Additional components:
Artillery grid - for placing artillery turrets with optimal spacing. Use the roboport as anchor for correct placement. The artillery turret ammo is supplied by the logistic network via requester chest.
Radar grid - for placing radars with optimal spacing. Use the roboport as anchor for correct placement.
Corner with roboport connector - if you create inner corners, it might happen the robots will try to fly directly across enemy territory and running out of power. By separating the logistic networks, you can avoid this. This blueprint implements a connector to request the required items from one network and provide it to the other. A requester chest requests items from the source network, and an inserter moves it to a storage chest in the destination network. A constant combinator is used for configuring which items are transferred. As default, the combinator is disabled and instead the red wire is used for configuring the items on a global scope. The belt is used to fill robots into the robotport, so the system is able to bootstrap itself independently. It will also keep the robot count, so if some are destroyed, they are refilled. The desired number of robots can be configured in a separate constant combinator.
Notice that after placement there is a 1 tile space between this roboport and the neighbored roboport to the left that is provided by the blueprint you place left of this blueprint. The inserter between the chests is exactly on that space.
Roboport connector - the connector without wall parts
Wall supply connector - example how you can connect the wall to the logistic network of your base.
Note:
Logistic network: a buffer chest is placed next to every roboport to distribute material across the wall, so repairs and replacements can be done on a timely basis. The buffer chest is configured via green wire on a global scope.
This wall can be used as soon as you unlock construction robots. Logistic robots, the requester chests (for logistic network handover) and the buffer chests are not required to be unlocked - it's nice to have later, but not crucial. The same with artillery. Artillery will prevent enemy nest creep, but this is mainly for the late game with huge bases where the walls need to be fully autonomous.
The wall strength was tested with default enemy difficulty. It is able to cope with behemoth mobs who start to appear with evolution > 0.9. However, automatic repair is required after an extended amount of operation due to the ranged spitter attack, so make sure you provide repair kits on the network for fully autonomous operation. Even if one part of the wall is overrun by (for example) 30 behemoth biters, it is kept alive and firing by the construction robots providing supplies, so it is actually invincible as long as there is electric power.
Example with all tied together (no manual editing - just put the corresponding blueprints down):
Simple modular laser defense wall
Re: Simple modular laser defense wall
Are these open segments really slowing the biters down, or is it just for show?
Never tried those myself.
Never tried those myself.
Pony/Furfag avatar? Opinion discarded.
Re: Simple modular laser defense wall
If you look closer, you see one stone behind every opening. You can see that from a larger mob of enemies, many of them will crowd in front of such openings and struggle to get through. They don't bite, instead they turn and run back and forth in front of such openings and don't get through. Not all, but 2/3 perhaps. Only the rest are able to directly run through every obstacle and directly attack the walls in front of the turrets.
I assume (guess) it has to do with collision boxes of enemies and the wall. The path finding algorithm tells the enemies which tile is free and can be walked through, but due to the collision boxes, the location on a tile from where an enemy is actually able to walk through an opening seems limited. So the walking algorithm makes an enemy walk back and forth until it reaches a location from where it is actually able to pass and not hold back by the collision boxes. It looks similar to a pea you throw into a funnel. The pea will circle inside the funnel around the hole until it finally reaches it.
If you observe enemy mobs attacking, you see this layout in general makes enemies slower and waste their time with trying to find the right spot to get through, thus more time for shooting them and less time for them biting the wall. Other designs rely on pure fire power (for example with flame thrower turrets) to prevent enemies directly run to and biting through their inner (any only) wall.
I tried to find a balance between wall item requirement and slowdown efficiency. There are other designs here and on reddit who waste (in my opinion) wall items. If you think it's still a waste, just remove the outer wall parts and only keep the inner wall. The blueprints are more about the modular design that makes assembling a larger wall easy, not about the outer wall layout.
I assume (guess) it has to do with collision boxes of enemies and the wall. The path finding algorithm tells the enemies which tile is free and can be walked through, but due to the collision boxes, the location on a tile from where an enemy is actually able to walk through an opening seems limited. So the walking algorithm makes an enemy walk back and forth until it reaches a location from where it is actually able to pass and not hold back by the collision boxes. It looks similar to a pea you throw into a funnel. The pea will circle inside the funnel around the hole until it finally reaches it.
If you observe enemy mobs attacking, you see this layout in general makes enemies slower and waste their time with trying to find the right spot to get through, thus more time for shooting them and less time for them biting the wall. Other designs rely on pure fire power (for example with flame thrower turrets) to prevent enemies directly run to and biting through their inner (any only) wall.
I tried to find a balance between wall item requirement and slowdown efficiency. There are other designs here and on reddit who waste (in my opinion) wall items. If you think it's still a waste, just remove the outer wall parts and only keep the inner wall. The blueprints are more about the modular design that makes assembling a larger wall easy, not about the outer wall layout.
Re: Simple modular laser defense wall
No-no, I don't consider this a waste at all. I was genuinely curious, I never invested time into defence simulations, and wanted to profit from your experience, since you obviously did, considering you made a whole bp book.
Thank you for the insight, I appreciate it.
Thank you for the insight, I appreciate it.
Pony/Furfag avatar? Opinion discarded.
Re: Simple modular laser defense wall
Look on the video in that post of mine (1st picture is a video) until you see the huge mob on the left. This is an earlier version of that wall, but you can see what I mean with the enemies crowding and struggling to get through one of the outer openings:
viewtopic.php?p=558808#p558808
viewtopic.php?p=558808#p558808
Re: Simple modular laser defense wall
Hello Tertius,
I was impressed by your soft wall already, so I give it and try your blue print book based on the idea of "gap_and_spike".
Whole book is nicely done, looks clean. The isolation feed is nice trick (thanks for the inspiration ), however ... .
1. All the blueprints are saved as Snap to grid + absolute. It makes is somehow not fitting a smaller spaces. I mean, you have land grab larger than you wanted. Perhaps Snap to grid + relative would be better.
2. Also I would recommend save the wall with landfill, at least the lower part with towers(the upper part, the actual soft wall dont need landfill), then it would better fits into edge of body of water.
3. I would recommend to add constant combinator to each section sending signal "S", then number of bots count like S * 3 = con_bot , S*1 = logi_bot. This will prevent a 2 sections of wall infested by 100 bots.
4. I noticed you supply a door to the request chest, but there are no doors used.
3. I see you effort to make a cheap wall, capable defend wandering groups - Settlers. It works, no doubt. However there should be a big warning, "Do not use arty !"
Because, if you do, a column of bugs come and eat your cheap wall for a snack.
Yummy
BTW: as you could see on the map, your wall is in the front of mine. The same battle testing for my wall result in zero damage or minor scratches. e.g. no single entity destroyed. Research level of the map is 10 for every space tech which is combat related , any developing base will be eaten horribly.
I was impressed by your soft wall already, so I give it and try your blue print book based on the idea of "gap_and_spike".
Whole book is nicely done, looks clean. The isolation feed is nice trick (thanks for the inspiration ), however ... .
1. All the blueprints are saved as Snap to grid + absolute. It makes is somehow not fitting a smaller spaces. I mean, you have land grab larger than you wanted. Perhaps Snap to grid + relative would be better.
2. Also I would recommend save the wall with landfill, at least the lower part with towers(the upper part, the actual soft wall dont need landfill), then it would better fits into edge of body of water.
3. I would recommend to add constant combinator to each section sending signal "S", then number of bots count like S * 3 = con_bot , S*1 = logi_bot. This will prevent a 2 sections of wall infested by 100 bots.
4. I noticed you supply a door to the request chest, but there are no doors used.
3. I see you effort to make a cheap wall, capable defend wandering groups - Settlers. It works, no doubt. However there should be a big warning, "Do not use arty !"
Because, if you do, a column of bugs come and eat your cheap wall for a snack.
Yummy
BTW: as you could see on the map, your wall is in the front of mine. The same battle testing for my wall result in zero damage or minor scratches. e.g. no single entity destroyed. Research level of the map is 10 for every space tech which is combat related , any developing base will be eaten horribly.
Re: Simple modular laser defense wall
I see you did the test in map editor mode. Did you manually set the enemies? How? I'm a bit worried, because I tried to place nests in a way the game will produce them, placement and density, then shoot at them with artillery. Was all good. I even created a zoo in my sandbox and let the enemies replicate themselves to exactly recreate how the game will distribute enemies. For me, there were never such big mobs generated by artillery this way. The wall was always able to cope with them. I tested with all standard tech researched but no infinity techs. Evolution is a 0.93 or something like that.
It will definitely be overrun if you select behemoth biters and spitters in map editor and directly place them by holding down the mouse button and hover with the mouse over an extended area. If you shoot into such a mob, the wall is overrun. But such mobs are not realistic. The game will not create them, at least not with the default enemy settings.
This is my zoo in my sandbox I used for testing: viewtopic.php?p=558474#p558474 (the save doesn't contain the wall I present here. It's just some wall).
Many thanks for the comments. Very helpful. I try to improve according to this. Especially the comment with grid mode - this is definitely an issue. I usually resolve by placing a blueprint somewhere, immediately cut and then place without grid constraint... a dirty kludge. I didn't use relative grid mode, because I didn't understand it in the past. Seems time to look again. I also tried to use half resolution (25x25) at development, but there was an issue with rotating/mirroring as far as I remember. I will look again, but to have a perfectly rotatable blueprint you need an even number of tiles, as far as I remember.
About the door: well, the construction was kind of rushed and the doors were planned but forgotten to actually build. It seems a waste to include a door into every wall segment. I planned to include a variant of the straight wall with a door. Don't ask what I'm currently doing to get to the other side of the wall.
Landfill is an idea. Currently, I just let end the wall at the water. The game behaves exactly as I want: no ghosts on the water. If there is a need to landfill, only small patches are required to fill small water puddles in front of the big water, if at all. At least this is how I do it. How do you want it?
The logistic network handover is also not fully matured. I developed it and let the wall deploy itself in my sandbox, but just started using it in my current real game. What you can globally configure is the number of spare bots (10). The number of bots actually deployed on a wall segment isn't a global setting - it needs a local setting on the second constant combinator located on the connector blueprint. Default is 50 for both. I didn't make this a global setting, because every wall segment has a different size, so it needs its own amount of bots.
It will definitely be overrun if you select behemoth biters and spitters in map editor and directly place them by holding down the mouse button and hover with the mouse over an extended area. If you shoot into such a mob, the wall is overrun. But such mobs are not realistic. The game will not create them, at least not with the default enemy settings.
This is my zoo in my sandbox I used for testing: viewtopic.php?p=558474#p558474 (the save doesn't contain the wall I present here. It's just some wall).
Many thanks for the comments. Very helpful. I try to improve according to this. Especially the comment with grid mode - this is definitely an issue. I usually resolve by placing a blueprint somewhere, immediately cut and then place without grid constraint... a dirty kludge. I didn't use relative grid mode, because I didn't understand it in the past. Seems time to look again. I also tried to use half resolution (25x25) at development, but there was an issue with rotating/mirroring as far as I remember. I will look again, but to have a perfectly rotatable blueprint you need an even number of tiles, as far as I remember.
About the door: well, the construction was kind of rushed and the doors were planned but forgotten to actually build. It seems a waste to include a door into every wall segment. I planned to include a variant of the straight wall with a door. Don't ask what I'm currently doing to get to the other side of the wall.
Landfill is an idea. Currently, I just let end the wall at the water. The game behaves exactly as I want: no ghosts on the water. If there is a need to landfill, only small patches are required to fill small water puddles in front of the big water, if at all. At least this is how I do it. How do you want it?
The logistic network handover is also not fully matured. I developed it and let the wall deploy itself in my sandbox, but just started using it in my current real game. What you can globally configure is the number of spare bots (10). The number of bots actually deployed on a wall segment isn't a global setting - it needs a local setting on the second constant combinator located on the connector blueprint. Default is 50 for both. I didn't make this a global setting, because every wall segment has a different size, so it needs its own amount of bots.
Re: Simple modular laser defense wall
It is nests spawned by editor on clear land. All the bugs are born by itself. Smog/Polution is not present.
I know there is some precise ratio of biters vs spitters nests, I dont care. I just swing mouse back and forth then add other kind. My eyball aim is about 70% bitter rest spitter. Then the nest is hit by slow firing arty. There is a reason for slow fire, you dont hit too many bugs with next shell. so more of them survive to come. development level is close to 1 I guess. It is an old map with about 500 hours. Pure vanila, only mod is calculator. Try "Wave defence" build-in mod, you get columns of bugs even bigger. Or you might try research added range on arty, you will get overhelming response.
Nah, doors are pretty.
When your wall meet water in the middle of segment so roboport is not there ... . Maby it is becouse all my walls are somehow automated so I need all segment to make it work. Anyway it is better/nicer to make sure always stamp complete line of turrets with backline infrastructure.(I think) If you carve turrets into water, you don need the soft wall in front, obviously.Tertius wrote: ↑Mon Jan 10, 2022 5:32 pm Landfill is an idea. Currently, I just let end the wall at the water. The game behaves exactly as I want: no ghosts on the water. If there is a need to landfill, only small patches are required to fill small water puddles in front of the big water, if at all.
... the sensor wires all way long on the belt are painful. Memory cell where input inserter adds then output inserter subtract would be more elegant. Nevertheless, i Like the idea of an logistic isle.
Re: Simple modular laser defense wall
That's what I say about the bugs in my software projects as well...
Well, it seems I didn't test actual late game behavior enough. My current real game deployment is on a new save where I just recently researched the last laser/energy weapon upgrade tech. And where I just finished artillery research. The wall suffered damage and occasional destroyed turrets without the latest tech, but after research it was all good (however, no behemoths yet). Behemoths were tested only in my sandbox.
I also just got artillery in that save and started placing it to shoot all the nest creep and to free land I intend to expand my base to. That was great - no issues.
I didn't observe it before (I played without enemies), but the nest groups seem to grow over time. What started at map generation as location with about 4 nests, are now locations with 8 or more nests. I don't speak of newly spawned locations, I speak of existing locations.
I assume I need to observe all this and see how the wall is able to cope with all that. Real game is different to a sandbox. On the other hand, I don't want a wall around my whole base that's consuming half of all my resources and is half of the time just sitting there.
Makes sense. I thought about an extra blueprint just for the required landfill. Some kind of stamp I put down first, then the actual wall. I want to avoid creating landfill at the coast where I want the wall just end. But that could also be achieved probably by deactivating the personal roboport and destroying the ghosts on the water before the robots actually build it, so landfill could be included in the real blueprint as well.gGeorg wrote: ↑Tue Jan 11, 2022 7:01 am When your wall meet water in the middle of segment so roboport is not there ... . Maby it is becouse all my walls are somehow automated so I need all segment to make it work. Anyway it is better/nicer to make sure always stamp complete line of turrets with backline infrastructure.(I think) If you carve turrets into water, you don need the soft wall in front, obviously.
Definitely, the wire is painful. Especially if one needs to patch wall segments together that are separated by water or cliffs. In this case, a "local" configuration seems more appropriate - it's there, you need to switch on the deactivated constant combinator and disconnect the wire into the big network. Or you just destroy all cliffs and use landfill instead of using the geography.
The combinator computes the difference between the requirement and the available stock. Memory cells with counting? I'll look into that, but I wanted to keep the amount of combinators low. It's just a wall after all, not a computer.
However, the buffer chests don't have a local default. That's on the todo.
Re: Simple modular laser defense wall
I think idea "dont need a castle wall around all the base" makes sense. Current wall can hold small group od settlers. In default game they are 15-20. Get notiiced you would get about 7 behemots at once per group. It is ok. Eventualy, you can add mines
If you want to keep idea of "lean wall", then make an Upgrade. e.g. another blueprint to the book which can be stamped over this, and make it more solid. The same way as I did my CloverLeaf Power plant. Then, you build lean wall all around the base, only in case you need arty to claim territory, you upgrade the wall and add arty.
If you want to keep idea of "lean wall", then make an Upgrade. e.g. another blueprint to the book which can be stamped over this, and make it more solid. The same way as I did my CloverLeaf Power plant. Then, you build lean wall all around the base, only in case you need arty to claim territory, you upgrade the wall and add arty.