Author Topic: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)  (Read 8454 times)

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
This is a highly experimental version of the game.  Your feedback is much appreciated on it, but be aware that parts of the game may be quite broken in this version.  I'll be out of town until Monday the 10th, so don't expect any fixes until then.  This is marked as a "postrelease" and has a version number of 1.013, same as the last official release, so you can switch between the two versions without incident (normally savegames from newever versions are not backwards-compatible, but that is not the case here).  This version will become the base of 1.014A when I return on Monday.

The Latest prerelease is now out: http://www.arcengames.com/share/AIWar1013Post1.zip

That version is an upgrade from version 1.012, so you have to already have 1.012 (or greater) installed. Just unzip it into your game folder (usually C:\Program Files\Arcen Games\AI War\ unless you specified something else). Please make sure that your unzip process keeps the folder structure from the zip file, rather than just unpacking all of the files into the base target directory.

What's new since 1.013:
(Cumulative release notes since 1.013 are attached at the bottom)

-------------------

-The default Auto AI Progress increase is now 1 per 5 minutes of gameplay, rather than 1 per 30 minutes.

-A whole bunch of more performance improvements have been made to the simulation logic, once again relating to how range checks are performed (this time it's regarding relative range checks), so the simulation performs ever-better with vast numbers of ships in combat on a single planet.

-Missiles no longer detonate until they are right on top of their target, rather than as soon as they are withing minimum striking range.  This makes lightning turrets way more effective against groups of starships or other ships.

-When players get too short on energy to finish building items in their queue, and then pause all constructors, the "short on energy" message now goes away.

-All ships now cost either metal or crystal, never both.  For all of the dock-based ships, missiles, etc, the cost is metal. For most other ships, including turrets, defenses, starships, and most economic ships, the cost is crystal.

-The way that build queues work has been completely revamped; the time required to build queue-based ships is now based on their metal/crystal cost.

-The build process for ships in queues is now a lot more streamlined.  There are no longer big pauses at the start and end of building each ship.

-Ships being built through queues now have their costs deducted over time, as each ship is gradually constructed.  The rate of construction is the same for all ships at a given constructor, so many times there will be a little bit of "excess build materials" left over at the end of building one ship. Those are automatically applied to the next ship to be built via the constructor.

-The displayed income rates for metal and crystal now are shown as a net value of the difference between the total income and the total per-second expense from queue-based construction.  Any ships that are directly placed to be built are not factored into this.

-Constructors are only now accelerated by engineers that are manually assigned to them.  Thus the net inflow/outflow of resources won't be messed up when a constructor is damaged and an engineer is assigned to it.

-Engineers will no longer try to accelerate constructors of their own volition (here again, to keep the net inflows/outflows stable and player-controlled).
« Last Edit: August 10, 2009, 09:53:33 pm by x4000 »
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #1 on: August 07, 2009, 03:18:14 pm »
Further note about the economic changes here.  See this thread:  http://arcengames.com/forums/index.php/topic,613.0.html

I think this new model works really well in concept, and in practice it also seems quite good so far.  Not to say that there might not be bugs, and the balance of how much various things cost might be way off in a few cases here.  I mostly based this on the old model, but threw in a few needed balance shifts along the way.

Such as:  cruisers are a lot more expensive now.  They already are a death squad from afar anyway, so the cost-to-benefit ratio is still quite good.  This just gives you more incentive not to let those cruisers get destroyed.

Also:  if something is taking crazy long to build, either build more constructors or assign engineers to assist (as always).  Even for the crazy-long build times, you can third those with engineers if you put three of them to work assisting the constructor.
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline freykin

  • Jr. Member Mark II
  • **
  • Posts: 69
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #2 on: August 07, 2009, 05:00:17 pm »
I'll give this a try tonight most likely, I actually love that type of economic model where it's showing your net income, ever since playing a ton of Total Annihilation.  Also, hooray on the engineers not assisting of their own volition!  It's been messing me up in my current game, with the Mk II engineers on my homeworld repeatedly going back and helping out when the whole reason I got them was to fix astro train damage.

Offline Kalzarius

  • Sr. Member Mark II
  • ****
  • Posts: 422
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #3 on: August 07, 2009, 06:32:01 pm »
A few issues I noticed in a brief continuation of my singleplayer game using this version.

1) Some of the construction times are incorrect.  For example, I saw Cruiser I takes 0:17 to build with my 3 engineers assisting, but Cruiser II claims 1:34, despite only taking 0:34 (in fact, when it goes to build, it initially appears 1:34 and then changes to 0:33).  The 0:34 is twice Cruiser I's speed, which is correct, since it's a doubling of the cost and therefore a doubling of the construction time.  I'm not sure where the hint is getting that extra minute from.

2) Engineers cannot be assigned to a constructor that isn't building anything, and the engineers leave that constructor when there is nothing left in the queue.

3) While I think the cruisers and bombers could both benefit from being built faster, it's the missiles I'm truly worried about.  Lightning missiles, for example take over twice as long to build as they have previously (8:40, I believe, compared to the previous 3:30 or so, with 3 engineers).  While I can appreciate balance, this is extreme for the multiplayer game I have going, where we are lucky to have an in-game second every 3 real-time seconds.  Every second counts at least 3 times for us.  :D

Otherwise, all seems OK.  I didn't catch any other bugs, but I think the numbers need more thorough testing and consensus from the community.

Offline eRe4s3r

  • Core Member Mark II
  • *****
  • Posts: 2,825
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #4 on: August 07, 2009, 08:00:29 pm »
Well its already been said - while the bomber displays 1:30 or so and actually only needs 33 seconds, the cruiser displays 1:30 and actually needs 2 minutes

Which is vastly slower than before

Generally, experimental indeed ;p
Proud member of the Initiative for Bigger Weapons EV. - Bringer of Additive Blended Doom - Vote for Lore, get free cookie

Offline MorphingJar

  • Newbie Mark III
  • *
  • Posts: 28
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #5 on: August 07, 2009, 10:58:54 pm »
Still trying out the new economic module.  Still not sure how well it sits with me.  But lets get down to the meat.

The Cutlass ship has been virtually crippled.  They Aren't attacking enemies anymore.  They still seem to do damage if you slam them into ships But they lost their ability to actually chase down ships and attack.  Also, when chasing down AI Cutlasses, simply being near any ships causes them to pause.  Almost like they are trying to turn around and hit the nearest guy, but just get stumped.

They seem like timid little mice rather than the voracious Tasmanian devils that they should be.

Edit.  Does not seem to be specific to the post release.  Am seeing this in the regular release as well.
Another Edit.  Once you get out of your home planet and branch out.  The cutlasses seem to start gaining some backbone.  At times they will be timid.  others they will charge right in.  Its hard to find some consistancy.  starting off with Cutlasses right away though seems to provide some imediate disapointments.
« Last Edit: August 08, 2009, 01:57:18 am by MorphingJar »

Offline darke

  • Hero Member
  • *****
  • Posts: 534
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #6 on: August 08, 2009, 10:46:37 am »
Haven't had a chance to play a 013 game yet due to busyness but I figured I'd take a shot at this. So far:

Start game is rather dull.

Now every building costs crystal to build (with the exception of the crystal generating structures... which cost metal; logical but very WTF-inducing when encountered when I got used to everything else costing crystal), it reduces a lot of the interesting decisions.

It used to be that you could juggle how many defenses you had, in comparison to how many ship building plants and engineers you had, and how much you should save back for your first energy generator because they all had a different crystal/metal cost (defenses tended towards crystal, offensive tended towards metal, energy was more crystal then metal, but needed a large chunk of both). Now the start game-play is "toss down a tractor turret at each wormhole, a couple of engineers, and as many space docks as possible to crank out ships since you've got nothing better to spend it on".

Ships are painfully slow to build, exacerbating the "toss down as many space docks as possible" problem.

Repairing with engineers is so very, very, very, very... *Zzzzzz...* Oh, yeah, slow. On the plus side it's gotten me to research and build a repair station. On my home planet. With the knowledge you start with. Yeah, slow. :) On the minus side? 9 minutes to build... *Zzzz...* Oh, yeah, on the really minus side? Repairing with repair station is so very, very, very, very... *Zzzz...*

Scout ships (and starships?) losing cloaking when you run out of power is a little... immersion breaking. Given said ships are by definition are ones that should be working on their own away from your supply lines. :)




Offline CautiousChaos

  • Full Member
  • ***
  • Posts: 100
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #7 on: August 08, 2009, 12:38:54 pm »
I'm kindof with everyone here as well.  Pretty big changes to the economics model that detract from the fun of the game. Appreciate the attempt, but it seems a case of fixing something that wasn't quite broken.

-cc

Offline eRe4s3r

  • Core Member Mark II
  • *****
  • Posts: 2,825
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #8 on: August 08, 2009, 06:12:10 pm »
Well actually it was not only quite broken but completely broken before

There is no way you could predict how many resources your shipyards use - so you are always micro managing shipyards which you shouldn't have to do.

If its properly balanced (and by the way, this is experimental so.. don't install it! ;p) then this will be far superior and better than the system before
Proud member of the Initiative for Bigger Weapons EV. - Bringer of Additive Blended Doom - Vote for Lore, get free cookie

Offline Revenantus

  • Arcen Games Staff
  • Hero Member Mark III
  • *****
  • Posts: 1,063
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #9 on: August 08, 2009, 06:18:15 pm »
Well actually it was not only quite broken but completely broken before

Heh, I think that 'completely broken' is something of an exaggeration. Perhaps inconvenient is a more appropriate word?

I haven't actually had time to properly test this post release myself yet, but I'm optimistic that the problems described here could be resolved by retuning the costs of structures and ships.

Offline eRe4s3r

  • Core Member Mark II
  • *****
  • Posts: 2,825
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #10 on: August 08, 2009, 08:00:57 pm »
Well, with that i mean that stuff before was fine, but if this change works out right, stuff will be.. better ;p

The previous system does not allow for "resource outflow" stats, which are kinda important - because we get resources per second - if resources were a one time thing, transported with harvesters etc. then the current system would be fine - but since the scale is so much larger the entire previous concept didn't fit.

Though i guess the experimental is a bad example of it - this change needs a lot of trials before it becomes that well implemented that its really an improvement, atm i look at missile build times and i see the problem.. missiles should build a LOT faster...
Proud member of the Initiative for Bigger Weapons EV. - Bringer of Additive Blended Doom - Vote for Lore, get free cookie

Offline darke

  • Hero Member
  • *****
  • Posts: 534
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #11 on: August 08, 2009, 10:16:08 pm »
At the moment it's turned the game from being pretty micromangement free in terms of economics (Note: I hate micromanaging economics), into a nightmare of screwing around with finding where I've got buildings turned on on planets, and generally pain-in-the-ass fiddling around with production queues. Take for example the following very, very, VERY common task in my current (sometime around the 9th or 10th try) at an AI10 game: I need lightning missiles.

Previously: I have lightning missiles in a queue. At some point it finishes, but I don't need one yet, and I'm running under the 15k a piece, so nothing happens. All is good since there's a 100% completed missile just waiting for me to need it. I decide I need it, I hit the "pause all" button, it stops all production of my ships all around the galaxy. (Yes, I do actually have ships being produced on planets other then the one or two main ones to pump out offensive ships.) At some point in time soon (a minute or so usually, it's pretty easy to estimate since I've got a little number underneath that says +X/second resources), I have greater then the 15k I need, so I hit "pause all" again to un-pause, and out pops my nicely completed Lightning Missile (or it could be a scout cruiser or whatever, I tend to produce all big ticket items like this).

Currently: I have lightning missiles in a queue. Except they may or may not be producing at any point in time, since I may or may not have positive resource balance, so I have to hunt down as to what's chewing up my resource. There is no easy way to do this, especially since it may be being caused by a random starship factory I took over in another system a while ago, and I set it to producing a couple of starships then completely forgot about it, or another of my missile manufactories building another missile. I don't know, and I have to find out because I can't just hit "pause all" and let it do it's thing. Usually this isn't a problem since it won't pop the ship out until I have a healthy chunk of resources, but since it's currently a continual drain, I have to track the damned thing down.

This is only going to get worse if it gets changed back from the current rather, well, boring method of having ships only cost one resource or another, since suddenly you'll have to micromanage every one of your docks, etc.

I like the concept of this, but it's really not working for the game since it's currently built around the previous method, and it doesn't have the dozens of buttons needed to "pause all but <thingie>" so you can actually play it without it being a economy-micro-lover's dream. :(

Also I never had a problem previously with estimating how much resources I was using. I used to glance at the numbers at the top of the screen enough that I could tell if the metal/crystal were trending up or down, and I could alter the number of crystal->metal converters I had enabled to compensate. Now I've got a "trending" number and a random flickering other number that tells me, well, pretty much nothing that I didn't already know. In addition it removes the actual useful value of knowing how much resources I have coming in per turn, so I can calculate how many turns it would take to get X amount of cash when I hit "pause all", which is important for things like dropping down buildings and the like.

Sure I can currently hit "pause all" and find that out, but instead of having two separate numbers that told me useful information, I now have two numbers, one of who's knowledge is completely redundant because I can get a good enough feel for it from the other.

IMHO, to use Revenantus' term, it's "inconvenient" now, and worked well enough before not to cause me problems. :)

Offline eRe4s3r

  • Core Member Mark II
  • *****
  • Posts: 2,825
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #12 on: August 08, 2009, 10:37:21 pm »
The fact that your missiles and ships build without resources paid is an exploit which is only there to prevent stalling of its production because of lack of resources. So in a sense, fixing that would make the problem worse.

I am going to assume that you have seen a IV factory stalled because your queued shipyards consume the required resources way faster. This new economic method is not there to complicate things but to finally fix the resource starving problem, which affects everything

And again, experimental means that only the concept is there - the resource prices and flow output of shipyards needs to be changed so that build times are the same as before and ships can still be produced at the same rate as before. Once that is done shipyards will no longer starve constructions unless your economy is starved. Which is the POINT ;p

Also - it might be worth mentioning that this is not the way the suggestion went too - the original suggestion just had 1 change - all construction costs are paid over the entirety of the construction with a fixed rate at the shipyards/spaceship factories/missile fabs (to predict ouflow)

If done right there would be nil micromanagement and no resource starving (unless its your fault ,p) but 1 varying resource type per ship makes matters more complicated than they should be.

Edit: By the way - the above suggestion would not have changed build times, so i am actually a bit confused why it was not implemented like that - While ship costs would need to be equalized (so that their resources fit with build time) the actual building time would remain the same. The thing is that ship costs go all over the place and they are not supposed to do that. But even then shipyards would just need to use 2 rates (for each resource) and scale the flow if the value is very small to fit with the larger value.

Also - the point of this change is partially to stop you from exploiting the 0 resource cost but build time still counting up bug of missile silos and spaceship constructors to get away with 0 economic management (and its not micro if it concerns resource rates) ;p
« Last Edit: August 08, 2009, 11:02:18 pm by eRe4s3r »
Proud member of the Initiative for Bigger Weapons EV. - Bringer of Additive Blended Doom - Vote for Lore, get free cookie

Offline darke

  • Hero Member
  • *****
  • Posts: 534
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #13 on: August 08, 2009, 11:48:26 pm »
The fact that your missiles and ships build without resources paid is an exploit which is only there to prevent stalling of its production because of lack of resources. So in a sense, fixing that would make the problem worse.

I disagree. I have paid the allocated time, I have paid the allocated resources, I have "exploited" nothing. I really don't think "playing the game exactly as it's designed" could possibly be "an exploit". I'm not even leaning on an obscure edge case here, this is core functionality! :)

I am going to assume that you have seen a IV factory stalled because your queued shipyards consume the required resources way faster. This new economic method is not there to complicate things but to finally fix the resource starving problem, which affects everything

Having not played anything other then AI10 for the last couple of months I haven't had an advanced factory under my command recently. Though as I recall the previous technique I used when I encountered this on a regular basis was I hit tab to go to the galaxy map, clicked on my front line base that was producing 90% of my ships, clicked on the space dock in the right hand side, then clicked pause. Suddenly I had more then enough resources to feed a Tech IV factory. The new resource management system has so far proven considerably more complex then the old one.

And again, experimental means that only the concept is there - the resource prices and flow output of shipyards needs to be changed so that build times are the same as before and ships can still be produced at the same rate as before. Once that is done shipyards will no longer starve constructions unless your economy is starved. Which is the POINT ;p

And "experimental" means I download, I test, and I whine in detail. Just like The Evil That X's wants us to. I mean really, it's not as if I responded "this change blows goats, turn it back to the old less-goat-blowing one". :)

Anyway, I need to get back to playing my AI10 game where I'm thoroughly exploiting an unintended intersection of abilities to make the game far easier then it should be, and I need to get the game won before The X-Man gets back, works out what I'm doing, and "fixes" it. :)


Offline eRe4s3r

  • Core Member Mark II
  • *****
  • Posts: 2,825
Re: Experimental Postrelease 1: 1.013 (Efficiency up, new economic model)
« Reply #14 on: August 08, 2009, 11:58:45 pm »
You haven't actually paid any resources when it counts up ,p

You can start the build counter to go up while its still "lacking resources" which is a exploit , you move the time to pay to when its finished instead of when its starting (which is the exact opposite of how the rest of the game works) ... while i agree that it helps minimize the micro management part, its clearly illogical to have for example - a nuke being build to 0 while you have not anywhere near the resource to pay for it or its construction. And then *pause* the fab to delay instant appearance on resource payment to a moment of your choosing, If you start a construction you should have to pay the price of it. Not when you want, but either at the start or entirely over the process of the construction itself.

The timer shouldn't even count up when you cue something and have not enough resources...

By the way - i didn't want to stop your from commenting/pointing out the errors - just that the implementation is not the in any way as it was suggested ;)

However - i am totally with you that the current experimental system is far inferior to the previous system ;)
Proud member of the Initiative for Bigger Weapons EV. - Bringer of Additive Blended Doom - Vote for Lore, get free cookie