PA Alpha Build: 52973 / 53072

Of course. I'll try to come at the problem with as many ideas as I can once we arrive at a more proper time to discuss it.

For now however, just the fact that you're specifically spelling out your intention that we shouldn't arrive at a situation where 'Advanced > Basic' is cause for jubilation on my part.

This build is great - I'm glad to see a focus on reducing pain points. I still find assisting and pathfinding flaky here and there:

a) If you queue a fab bot up to build a heap of mexes, then add a couple more to follow it, sometimes a few mexes later the assisting ones run off to a random spot and frolic in the field (run in circles next to each other). The original one continues. Improving their work ethic would be good, as you often don't see it's happened until much later on.

b) Sometimes you queue a heap of mexes up on a juice patch of metal spots - something happens to the fabber but there are now wireframes blocking building on those metal spots. Highligting all fabbers to cancel orders did not help.

c) The flow field thing seems to stop the comm and other units when they come up straight against a flat edge of something (power gen, etc). If they hit it on an angle they slowly get around it, but if it's perpendicular they don't 'randomly deviate' enough to find a way around one side of the structure. Perhaps the flow field costs around buildings should be a bit rounded in shape rather than square.

Great.

about metal points, is there going to be a placement fix in the works soon? the AI can't seem to advance if it can't access all nearby spots, and just sits there. Perhaps add a foot print to allow at least a basic extractor to be able to land on it for sure, so it can still be too close for use with an advanced... Also, in TA even without deposits, you could still build an extractor on the ground for a smaller amount of the metal, determined by metallic content in the ground, and on a metal world, there weren't any extraction points, since the world is made of metal...

I think the best would be to force metal extraction points to compensate for locations, even if it means they get nudged over and away from rocks, ledges, and water lines where they can't build extractors. (Or those points are auto removed due to inability to be built on for the sake of the AI.)

SO MUCH PERFORMANCE IMPROVEMENT.

Thank you so much for the optimization. Whatever you guys did, I can now happily play the game without rage quitting due to the amount of fps I usually get. Keep up the good work Uber!

I did notice my commander stuck within a metal extractor. I could not find him anymore and suddenly I saw a fat metal extractor, seems he merged with it. Giving it a new move order he moved out the metal extractor.

I agree with nano partially, as far as advanced... no, you will not get me to say it... t2 metal extractors, they could at least for alpha put out about half the metal they do, and still probably be desired upgrades considering there are only so many metal points and for you to get more you need to go to t2 mexes eventually, EVEN WITH the absurd macro from exploding and reclaiming t1 mexes.

But I do understand, even before you "officially changed the terminology", that you were aiming for putting non-copied units in t2, ones that don't do what t1 does at all, such as artillery as t2 not t1 and scouts and antiair being t1 not t2 so you would literally have to build multiple factories to produce both, yet their "order" is based on their necessity for one's own survival, needing scouting and anti air to even stay alive compared to artillery and levelers just to dig-in and snipe-at the enemy. Like two separate factories entirely, just one is pre-dependent on the other.

Such is a good model. Honestly, t2 mexes can in fact work if you were to balance it that claiming the usual 5 metal spots that spawn in a shotgun-spread-like pattern would actually build faster and produce more than a single t2 mex, you would have the choice to claim more mexes for more profit, or claim more efficient metal per spot. Remember, after you claim so much land you would have to go about condensing mexes into t2 anyway, so it wouldn't ever "uselessness" them, and it might even balance the out-of-control economy you can create with some tedious work.

I would say you should be rewarded for defending more land anyway, over defending one expensive mex which, may I add, also has the bonus of being very "hardy" of high health and thus might actually survive if bombers got through standard AA. The higher health is also a very good niche building a t2 over a t1 has, so it definitely doesn't need very ungodly metal income.

And optimization allowed me with 1g gpu and 4g ram to finish a game with an AI on a size 4 planet. Pretty smooth indeed.

I think that basic mex could have low cost, high health and small yield, so you would use it to spread to low defended areas, whereas advanced mex would be expensive, low health and high yield so that you are going to want to use them for their economic boost, but putting them into the frontlines, or low defended areas in general, would not be cost efficient.

As for generators, making the adv. pgen go nuclear when destroyed (in a smaller than comm scale) would definitely make you think before you build it in the middle of your production areas.

These are my ideas how to make the current resource structures feel less like upgrades and more like tactical choices. Oh and we definitely have to be able to build adv. mex on top of basic ones. The cycle of harvest is not fun(maybe make it automatic?). Also why not be able to do it both ways? (t1->t2 and t2->t1)

You would need to fire 45482 billion billion shots. I'm not sure how much artillery you build, but my estimate would be oodles of games.

Mine keeps hiding inside energy buildings. I can usually coax him out though. He seems to forget queued orders quite often too; I can see the outline of the queued buildings if I select him, but he just stands there. I usually move him somewhere and then re-queue the build orders.

Is it best to keep shift held down whilst selecting different buildings to construct, or release and re-hold shift?

I started playing TA again and I do notice that the explosion size of various buildings does affect how I build my bases.

The "uselessness" of the metal extractor points, is where the AI can not build on them cause there is no foot print space for the AI to build on them. You and I, are able to avoid those spots quite easily and able to keep moving, the AI can not. Try the polar region of a metal planet, or try one of the smaller lava planets with low lava. Both tend to have metal points in places the AI is trying to build metal extractors, The AI even tries to stay out of the water when it comes to placing shallow Metal extractors on a high water Lush planets. This is why I suggest removing the ones that can not be built on, or have them moved over so that the AI can build on them to keep the AI from not doing anything, would be easier to move over or remove non-accessible metal points, than to try to rewrite the AI to try to exclude those metal points.

Those aren't tactical (nor strategic) choices if you're required to have a certain economic output to build units efficiently.
They're still upgrades... flimsy and desperately costly to lose upgrades... but upgrades none the less.

Hmm.. I understand what you mean. I would prefer if there was only one tier of land/sea mex and gen and having better ecostructures means doing something.. harder. Like a solar panel on orbit or a huge reclaimer on a metal planet or something. However I don't think Uber is just going to scrap t2 ecostructures so what do you think should be done to t2 to really make it a stratetic choice? All I could think of that isn't a direct upgrade would be low cost - low health and high cost - high health structures with the same yield rate.

My hands are tied. Sorry.

As an aside though, this is what I'm using as a starting point.

I know that's a grossly oversimplified equation, but I find it helpful to study it in its most basic form.

Thank you for the reply. Other than the issue with the AI, I can't seem to find any bugs, just minor performance issues when I am well above 1000 units that are in constant motion, and that is amazing to say the least. In the mean time, I will keep looking for actual issues.

Nano, you can discuss all you want, but we won't be having serious discussions about it (Uber + Community) until we feel the game is in a state where such a conversation is fruitful.

We are not updating the build notes/post, but a small change was hotfixed live today, build 53072 - This should address the Intel HD 2000/3000 driver bug for people running Sandy Bridge chips.

Serious topics deserve serious discussions
I'll continue my efforts to be diplomatic.
no promises ...

Believe it or not, I'm actually looking forward to Multi-planet play, so I'm perfectly content for you to focus on that, rather than my ramblings for the time being.

Me as well, I don't mind if the last thing to work on are the AI interactions with metal spots as it wait's for me to toss "small" rocks at the planet it inhabits...