This is purely the system. Actual map will use different units, a different power source behaviour and power link. Also the town hall will have a buff with a death effect to temporarily disable the grid.
Also notice how I turned the beam upside down to make it look like swinging cables.
Otherwise it is ready unless there is something you can think of to improve it?
Maybe altering how the cables are destroyed might make it look better.
Also there is an issue if there are multiple power sources it will be first come first served.
I will upload an improved version
Edit: made a more or less finished version with any unused stuff being removed. Just need to import it and hook it up to new units both in the validators and units data types. Even made the power link orthogonal to the one used by the protoss.
Stuff that needs changing are the power nodes and power source units. I believe the source is going to be the town hall. Also need to change the splat actors used by the Power actor. Also maybe changing the normal and specular textures of the beam.
I decided to see if killing the unit actually removes the cable. Most of them worked, but there were a random 3 cables theat hung from the pylon even though all the zerglings were dead. I killed the pylon, but the cables just hung there.
With pylons have them tall with a circular apical attachment point so you cannot go wrong with them. With the power source/user structures just have a crown of attachment points and use an AM that goes by distance choosing the closest impact site to the source and the closest launch site to the target.
Also look for models with the attacher bones because those can be used as beams.
Oh and we can adjust sized of the various radii and model scales as needed.
Done some testing of your system and my opinion is to go back to the drawing board visual wise. It does not support multiple sources, has major issues if several links are activated at once, is not very stable as far as visuals go and has moderate lag with just 100 links in close proximity. Here are some screenshots showing some problems I found.
Also try to avoid having a Search Area effect linked to a Create Persistent effect
The Extend By Unit Radius flag on the search area effects explains your radius problems.
I would recommend avoiding beam models that flow in a specific direction because your system often causes them to flow in opposite ones.
Also with my power system they are meant to be dirty great big power cables and not tentacles but that is just the model used and can easily be swaped.
I agree. Beams seem to be the protoss thing (I'm not saying it is, but just saying) Anyways, I'm not sure if this could be possible, but there is in the doodads, a cable that's connected to 2 poles. I'm not sure if you could use that. But if you can, then it would be good becuase it's already looking like a cable.
Just look at the bones the model has in the previewer. If it has attacher it could be used as a beam. If you mean to it as an alternative to spanning the gaps there would be the problem of estimating the scaling.
Or just use it's model to replace that of the archon beam actor and see if it works.
In the top of the lower right pannel you should find a small button with a tiny icon of a bone on it. A folder is then shown called bones.
You still are not able to link multiple sources. Also there is still the issue of all the persistents. For cutting the power just do like I did and have a perpetuating death response that momentarily disables the network. Still has issues forming a proper network.
Maybe consider removing the orphan destruction pathway for your beams and replacing it with something else? Also maybe put in a target sorts in your search?
I tried using channeled affects to reduce lag potential, more like Lilianna's idea I guess. I still have to work on it. But the idea is that powered pylons wont create that huge amount of searches until their power source is killed. I haven't checked the systems uploaded here recently, but it looks like we don't have a definite system yet.
In short you did not check the perfected system I posted in #85. My system only searches if there are targets that have not yet been connected to the visuals. It forms perfect networks despite hardy destructive testing. It also seems to only have caused a stir with the choice of model used for the beam.