PTE build 79270-pte now up

Com being weak is a problem - but making him able to spam walls can't be the solution. The commander is actually the one unit where I could see PERSONAL shields working, powered by energy until the energy producing infrastructure is destroyed. 5-10% rest damage afflicted while the shield is up. So either you take down energy first, or your snipe needs 10-20x the regular damage to take him down.

I think that was suggested before, more than just once. And I don't remember any real arguments against, except for the "it's not supported by the engine yet" one.

Damage prevented to energy consumed ratio would be the one, damage absorption ratio from 0-100% would be the other required value. No such thing as shield which need to build up or which can be shot down, but energy which is subtracted straight from the central energy storage for each shot. If energy isn't sufficient to pay for the entire cost, only proportional amount of damage is prevented. So a nuke on the commander can flush your energy storage completely AND kill your commander if unlucky.

I still don't see why the com being able to build walls is a problem. He is your primary fabricator after all.... It kinda fits for me. That said, I'm not against your suggestion.

Ta had the com cloak + unlimited fire rate on dgun if you had the energy for it. Meant late game you had to cripple your opponents infrastructure before going for com which is as it should be.

As it stands the only viable target for a major offensive is the commander. That is kinda a separate issue however. commander plus walls is an early game strategy anyhow. T2 just bypasses the walls and can take him down too quickly to risk it late game.

I like the splash damage changes on the grenadier. Glad to see some more defined unit roles. Not gonna lie though, every time I see hp decreases I get pretty sad. When I messed around with a mod that increased hp values across the board I really started to enjoy combat more. Oh well that's what mods are for I guess.

Full ack on that one.

But commander spamming walls ... it's the wall mechanic itself which doesn't feel right. It's a war of attrition as soon as comboxing starts, both players are hindered from doing anything else while under attack. The problem is, that the learning curve for doing anything else in such situation isn't gradually, it's an overhanging cliff as you need high APM before you can do ANYTHING else apart from building walls without killing yourself.

I still think that walls need a post-construction delay before they reach their full HP. Meaning that their metal/HP ratio isn't better than any other mobile unit unless they are left alone to "unfold" (until their energy wall raises).

Just read Jables' posts in this thread, he updates the opp and also posts the buildnotes in a comment.
But most changes are in balance and save/load fixes

That is actually a cool idea. Walls should need 10 seconds or say _after_ they are completed to enable their shield and before that provide no protection or have very limited hp or whatever is easiest to do engine wise.

Also as another idea I agree taking away the wall from the commander and buffing the commander (extra hp?) in return would cool. Direct commander fights are fun, but currently they only really work by using wall spamming in fight.
Walls should not be able to be created directly in fights. They should be something you setup in advance to fortify a location you already have secured so your opponent has to either bring arty or walk around it.

More info on this would be great. What exactly is being impacted? Is it failing to load, or just going slow?

Sure. I played the match yesterday and in build 79270-pte, but the replay list is not showing it yet so I can't give you a lobby ID. I'll keep periodically checking.

In general, what I saw was that after the AI commander built his initial base as I described above, it got stuck in a loop of assisting a vehicle factory for a while, wandering a short distance away, returning, and then assisting the factory again. It seems that this loop was perturbed when a large (other) AI group of units invaded the base, and had to be fended off by the commander. It was previously only using groups of tanks to fend off forces sent its way. After the commander intervened in the battle, finally started to build other structures, including (one) metal extractor, several more factories, and an orbital launcher.

The AI doesn't appear to react very will to orbital threats when it doesn't have defenses in place or avengers already; I've been able to place anchors above enemy commanders and slowly whittle away their HP without them particularly caring. Not sure if this is because they don't have sight of the threat, lacking any radars or umbrellas.

Aside from these few edge cases though, the AI generally gives me a run for my money if I'm not aggressive about expanding. In particular, I've been testing on the 4 planet random generated system, which has a large central planet, two smaller planets orbiting it, and a further out small planet orbiting the sun. This map tends to have a lot of water on the starting planet.

That one is because it didn't have sight on them. The AI is incapable of reading tracers. It would help the AI (and players!) a lot, if a unit became automatically "tracked" by radar whenever you can see on of its tracers.

Will we see a new random system generator in next stable patch?

Going slow, very slow.

Easiest solution for wall issues would be to separate them into two tiers. Reinstate the old wall, make it so you can't shoot through it, make it weaker, and make it the basic wall. Then have the current wall be an Advanced wall that requires energy.

to note: this is currently looking like our release candidate. Unless something pops up nowish, we'll be pushing it to stable.

I guess I'll have to keep copies of 78071 around until I can work out a minimal server-script hack for paste unit

How happy are you guys with the changes to the Deep Space Radar, in regards to balance considerations?

is copy paste unit locked in the engine? If not, maybe a mod could be made.

This isn't exactly what you are talking about, but it's ai related. I made anarchipelago style map that is huge. The AI seems content on spamming land units and never takes to the sea. Or if they do, it's very late game and puts them at a huge disadvantage. Is there anyway to program if the map contains more then 50% water the AI will focus navel? The AI seems pretty broken in maps like this. Having just played TA I noticed the AI focuses naval on heavier water maps.

I believe the issue is server-script, which isn't moddable in the usual sense. I think server shadows work, but the script files get loaded before server mods are sent.