New armor system?

Nobody is suggesting that bots will replace tanks. That is a fallacious straw-man. What people are suggesting is that robots can compliment more traditional vehicles. They excel in some areas where tracked vehicles will fail, and vice versa. This is both an interesting concept in reality and in Planetary Annihilation.

Which segues onto my next point. I'm not quite sure you understand the role of designs based upon reality in games. Nobody is suggesting that games should be realistic for the sake of realism. What people are suggesting is that realism has a place in informing some designs. I explained what those roles are in a post some time ago, but it is as relevant now as it was then:

These are good guidelines for when reality is worth paying attention too. I would suggest that the idea of unit behaviour being dependant upon the basics of physical projectile motion and unit bulk embodies aspects of all three of these guidelines. This is why using ideas such as fragile planes and tough tanks are interesting concepts and ones which make sense. Stop seeing it as a dichotomy between fun and realism, and start looking at how we can use inspiration, possibly from reality, as something to enhance the fun.

Is it too late to suggest the 'everything can shoot everything' solution?

I'd suggest that if you are getting a lot of edge cases maybe the methods are flawed. Just like with any system really, if you aren't getting the results you are looking for maybe a different approach should be considered.

As for the "edge case" of the factory thing, that is simply a result of how that system was set up.
You simulate projectiles so that they can theoretically hit anything in their path. You have a straight forward damage system that does equal damage to all things they hit. You asked for this.

It is better in such a system to anticipate that all things can happen, and consider the cases where they do. Which is what TA did, tho with varying levels of success. For example, how much health does a factory have to be able to be killed by missiles designed to take down bombers? If you want bombers to be countered easily by anti air perhaps they should have less health rather than making the missiles more powerful.

But if you must impose arbitrary limitations I think you should start with looking at the targeting system. You don't want anti air weapons to target ground units, but this is a case where it happens anyway.
If it's anything like supcom, the targeting does not distinguish between layers, only between unit movetypes. The aircraft could be underwater and it would still be targeted, along with any nearby subs, because of this. What you really want is for that system to only target things that inhabit the air layer, and not aircraft sitting on the ground, or in a factory. Or things that are transitioning. So basically you want to talk to Mavor about the layer system because it has a number of flaws like this.

HP aren't necessarily good just because they are simple. It is a very basic system for simulating damage, and frankly I wish games would start to move away from it. It is a relic from a time when you needed to do these calculations in your head with pen, paper and dicerolls.
I cast my anti air missile for +214 damage against air! Roll for initiative!

I am not against armor systems as such, they do add some depth to it. The problem with armor systems is that they can be highly unintuitive if used incorrectly. So they must be clearly defined and visible. It is of course unreasonable to expect an aircraft to be as heavily armoured as a tank, or a military structure. So it makes sense that they would take more damage from light weapons like AA missiles, machinegun fire or flak bursts.

Speaking of zero hour missiles, one of my favourite things about those was that they would run out of fuel and fall down if they didn't hit anything. Wonderful little touch. Missiles in PA are really derpy right now flying in circles all the time.

@madsci I think that whole point can be summed up neatly in one line:
It doesn't have to make sense, it just has to look like it does.

The reason t2 air invalidates t1 air is because they got more DPS for cost and more health for cost which is a shame in my opinion.

If we are gonna base our arguments on realism then we should remove Health Points altogether. Unless we make up some techno-mumbo jumbo where future armor is modular where a unit can keep functioning at full efficiency because it basically replaces damaged parts until it reaches a critical threshold and explodes.

I could support directional armor, various armor thickness and variable weapon penetration. However, PA have not set out to use those mechanics so even if I like those mechanics it might be inappropriate for me to argue of the inclusion of such mechanics.

Consistency is important in my opinion. I should be able to get a feel of how a unit performs even if I haven't tried it against all other units. Inconsistent damage modifiers with few to no visual cues breaks that.

You are wrong. Of course an anti-bot and anti-light unit can be more expensive than ordinary tanks. Make a unit that have Area of Effect and kills bots in 1 shot. Voilà! It is now very effective against clumped bots. DPS isn't the only balance lever you have. It could even have shorter range than tanks too. Special damages promotes lazy design.

Take a look the unit list in Zero-K. http://zero-k.info/Static/UnitGuide
How does the lack of armor classes restrict the simulation?
On the contrary, by lacking armor classes, Zero-K have managed to design a counter-system based mostly on physical properties such as Area of Effect, projectile velocity, unit speed, unit acceleration and unit size.
Also explain how Armor is WYSIWYG.

Damage modifiers is the easy way and if you use it too much in an inconsistent manner you end with an illogical mess that doesn't make sense to the player.

Where did I ever say anyone was saying that?

Since you're not capable of parsing my argument let me boil it down:

The bots in the game would not exist in a world with tanks. Ever. They would most certainly not be faster than tanks.

Arguments based on realism are stupid and should be laughed out the forum.

And what are you basing that assumption on?

I'd say that having trillions of different units is what truly adds complexity, more than armor effects ever could.

I mean, just by looking all those different kind of units makes me feel like "wow, why would I bother to learn all this?".

More units also add depth and variety. Without armor classes, it is also much easier to abstract and compare units mentally and statistically.

Well I think it is a matter of taste. Some people prefer more unit choices. I think most TA players does.

My point were that you can have unit variety and Rock-Paper-Scissor unit balance without armor classes.

Yes, you don't necessarily need armor classes, but they can be helpful sometimes. The devs just saw that the game was better with these changes. Was that a right decision? Dunno. At least it fixed some issues.

To me that looks very much like units being assigned roles and tinkered around with until the outcome is desirable. Instead of creating roles first and then assigning units to them.
But on to addressing the issue at hand: Plus one internet dollar to bmb's suggestion of making targeting category dependent on actual layer position. That would have solved the problem immediately without edge cases nor outcries. It would have been simple and intuitive, too.
Like having your cake yet having your cake.

Not Metal Gear Solid. It says that walking tanks are expensive, and they have their terrain flaws, but they are more nimble in each direction and are better at climbing terrain. So they would both have their ideal terrain that the other couldn't use.

But again, we are talking realism and nonrealism. Bots don't matter, it is detrimental to remove bots from this game because hurr durp.

I like madsci's post about how chasis should be used to determine a unit's movement and qualities. Don't get me wrong, this game already does that, vanguard is fat and slow, doxes are small and nimble, even planes look their part. More could be done too.

I want to make a mod where chasis has a base movement type for all units, and modulars arm the chasis. So the weapons all function the same on every chasis and the chasis functions the same despite weapons. People reference battle earth 2145 or something. I try to stay away from directionalism and in fact never played that game, but sort of like that.

Earth 21xx did that quite well. Had a vastly smaller scope, though, which kept that whole design part manageable most of the time.
On topic: That whole armor debate comes down to technical design competing with gameplay design. And while Uber are very bold when it comes to the technical parts of the engine, the direction that gameplay is taking is, well, less bold. Which is a bit disheartening, honestly. There is only so much that can be modded.

Ok, i'll bite...
Is anyone seriously having a problem with a system like this that improves gameplay and balance without having ANY noticable negative side effect? If you see fighters shooting at an air factory while flying over a base does that ruin the game for you? Or is the game ruined when the factory doesnt explode afterwards?
Yes, it would probably be possible to prevent such cases from happening at all and there are certainly many suggestions in this thread that might work, but is it realy worth even considering to invest time money and resources into something that is already fixed and wouldnt even be noticed if it hadnt been announced?
Why the hell would anyone even care about this, if the edge cases are fixed that is GOOD even if the fix isnt in the minds of some players as elegant as it should be.

That fix obfuscates the underlying issue that is what is bad about it. Instead of fixing landed air-craft being classified as airborne and/or reducing fighter damage to not-insane levels which both would have solved the core problem.
That, and it sets a precedent for additional 'fixes'.

There is no underlying issue. Games are simulations, not exact replicas of physics. There are trade offs to be made for many reasons. Sometimes you have to do things like this so you don't rabbit hole on issues for months and end up right back where you were.

Yes, there is: Anti-air weaponry attacking not-air units. Which landed aircraft and factories clearly are. And the cause for that is the static layer categorization in combination with units able to change layers.
When solving those problems of course tradeoffs have to be made. But claiming there is no issue when the simulation does behave in a consistent yet unwanted way is just wrong.

Exactly this. I have a question for all of you. Besides Raevn, you shaddup for right now, but if Raevn woulda kept quiet, would ANY of you knew EXACTLY what Uber did if they just announced they:

-Fixed fighters destroying air factories

...If they were that vague in their patch notes, and Raevn/MaxPowerz/Colin/Trialq/Other Modders all didn't spill the secret, would any of you knew or cared how it was fixed?

Random Forum Posts: Oh thank god t2 air got nerfed, AND they fixed fighters killing air factories. GO UBER!

Generally, you are against and arguing against units doing different damage to different types of units. Generally, this is not happening here. This is being used currently as an easy way to COMPLETELY REMOVE UNINTENIONAL DAMAGE. There is a difference between completely removing unintentional damage, and rebalancing all weapons ingame to do different damage multipliers on different units.

Fighter missiles still hurt ground units if they miss/try to track an already dead target. I learned that the hard way when one stray AA missile took out my botfabs.

The problem here is the following: The simulation model yielded consistent yet unwanted results. Instead of changing the model or the expectation for the model's outputs the results were tinkered with. What will happen when aircraft are allowed to land in the midst of a ground army? Will ground units in general be added an armor type? What with potential landing pads, carrier units?
Every time a unit with a 'layer'-tag enters a different layer you have a new edge case to deal with. Just take the Astraeus - that thing is switching layers like no one's business - is it a ground, air, orbital unit? All of them? None?
So why the decision to apply an intransparent and unflexible fix instead of changing the static nature of 'layer'-tags?
Yeah for increasing modding options (which is an assumption, who says the armor 'system' will ever be exposed?) but for gameplay that is supposedly treading all manner of new ground this armor fixing business looks deceptively like a step backwards.