FreeOrion

Forums for the FreeOrion project
It is currently Sat Dec 16, 2017 5:12 am

All times are UTC




Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 90 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
 Post subject:
PostPosted: Wed Jun 30, 2004 4:56 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
tzlaine wrote:
Quote:
Could you give reasons to support POOLING PP

There's no need to yell.
Didn't mean to yell, just very clearly emphasize, since the point was missed before.
Quote:
The reason the micromanagement argument applies to PP pooling is that without pooling, there is de facto no global build queue. Providing a consolidated interface to N local queues, while it is more convenient than going through N screens with individual queues, doesn't actually reduce the level of micromanagement you have to do. You still have to check on and manage items in each of the N queues.
What do you mean by "check and manage items in each of the N queues" ? If you order 400 ships, and 40 planets are collectively working on them, then you don't need to check anything other than the status of the whole order. There are no "local queues"... that's the point. Each time a particular planet finishes a ship, the global queue sets it working on the next ship in the order, until all are done. All you did was request a bunch of ships... once. Where's all the micro you're worried about?

If you're worried about worlds with several wonders enqueued after eachother... then this would be a) very rare and b) handled with a sitrep update when a wonder is finished (it is a wonder after all), so you don't need to check the queue at all, really.

The one case where there would be some micro is when dealing with local wonders... but you only need to check the build status of the wonders, not umpteen other queues / the whole empire / N of them. This wouldn't be any more complicated than if you were pooling PP either... the progress appears at the same place in the global queue, but has a not "at Bananastar IV" or somesuch, as opposed to nonlocalized orders which have "distributed" or somesuch...

When the location specific order is finished, you don't need to do any micro either. If a world has no specific build orders, then it's again open to orders without specific location, so it would start contributing to nonlocalized ship production, if there are any orders still being filled. (this would take into acount distance and such, automatically, so it wouldn't build the ship far from where you want it)
Quote:
Quote:
Quote:
Finally, I'd like to point out that every argument against the pooling of PP that has been mentioned in this thread can be applied to the pooling of RP. Planet-level research, anyone?
??? Since when? There's no strategic problem with having research available everywhere when complete. Research is not done only in a specific location... it's spread out for the whole empire (as far as I know), so it makes sense to pool RP. Building ships and wonders "should" be done in a specific location, which makes pooling PP problematic...
Alright, maybe not every argument against PP pooling applies to RP as well, but the major ones seem to. I was specifically referring to the "game balancing nightmare" that is supposed to accompany PP pooling, due to the radically different empire production rates possible. This specifically applies to the pooling of RP as well, but with RP there seem to be no objections.
That's one argument that partially applies to RP and PP. The nature of the imbalance for RP and PP is different... (see below)
Quote:
Why do you say that building ships and wonders at a certain location makes pooling PP problematic? I don't understand what you mean by that.
I refer to the dilemma concerning whether or not to limit local PP spending from the empire pool. If yes, how? If not this leads to imbalances, however this isn't the same kind of imabalance as with RP.
Geoff the Medio wrote:
drek wrote:
Just like research, you wouldn't be able to decrease the number of turns a project takes to complete via throwing more points at it. More industry=more projects, not faster completion time.
This, in of itself, does not eliminate the ability for any colony to produce with the full force of the entire empire. Ok, so shipyard in the middle of nowhere can't produce a Death Star in one turn... that doesn't mean it wouldn't be able to produce 400 small attack frigates in ~2 turns. This is just as imbalancing... assuming small / large ships are balanced.
(If I misquoted you, sorry... nested quote edited got complicated)


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 5:42 pm 
Offline
Designer Emeritus

Joined: Thu Jun 26, 2003 8:07 am
Posts: 935
Quote:
What do you mean by "check and manage items in each of the N queues" ? If you order 400 ships, and 40 planets are collectively working on them, then you don't need to check anything other than the status of the whole order. There are no "local queues"... that's the point. Each time a particular planet finishes a ship, the global queue sets it working on the next ship in the order, until all are done. All you did was request a bunch of ships... once. Where's all the micro you're worried about?


I fail to see how this is effectively different from pooling PP into a single pile--just more complex with the same effect.....unless you are saying that ships deploy at the same planet in which they are constructed then move to a central rally point, which is just not going to work.

1) What if the planet producing ship X from the global queue is attacked and conquered? What happens to the project? This is bad if the player is relying on your little queue robot, and it decides to build ships near dangerous areas.

2) What if the path between the producing world and the rally system is unsafe? What if the path becomes unsafe after the project has already started to build?

3) What if the global queue chooses a world for producing ships that, a turn later, I decide needs a shiny new building?

We know from experience that robot governors don't work well in a strategy game. A bot assigning projects to worlds is bound to make errors in judgement. The player has to constanstly police the queue, which defeats the purpose of having automation in the first place.

To put it another way, in a multiplayer game a player manually assigning projects locally is going to beat the snot out of a player using the automated global queue.

With a global pool of PP, and set rally/deployment points in the form of Shipyards:

1) There is only one queue to juggle.

2) The player is making a decision: which rally/deployment point do I want completed ship to appear at? If that decision should later on turn out to be unwise (due to enemy action, for example), the player has only his own lack of foresight to blaim, compared to crossing one's finger and hoping the robo-queue doesn't botch things.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 6:00 pm 
Offline
Juggernaut
User avatar

Joined: Fri Dec 26, 2003 12:42 pm
Posts: 872
Location: Germany, Berlin
drek wrote:
I fail to see how this is effectively different from pooling PP into a single pile--just more complex with the same effect.....unless you are saying that ships deploy at the same planet in which they are constructed then move to a central rally point, which is just not going to work.
I see the difference! The planet itself limits how fast something can be produced, not an artificial defined X PP per Y turns (which i realy dislike)

drek wrote:
1) What if the planet producing ship X from the global queue is attacked and conquered? What happens to the project? This is bad if the player is relying on your little queue robot, and it decides to build ships near dangerous areas.
The project is kill! What else should happen? It's the way i would like it, rather than simply lowering global build a little bit down, which doesn't hurt at all.
EDIT: I mean if there is pool production and you are building a big project, it won't hurt you if one of the supporting industry worlds was conquered. The project would simply continue. On the other hand, if you build it on a planet and this planet is conquered it would hurt you. (And that's the way it should be!)

drek wrote:
2) What if the path between the producing world and the rally system is unsafe? What if the path becomes unsafe after the project has already started to build?
Another route would be chosen. If there would be such a route, ship movement would stop at the closes system where no enemy is present. Btw. you should eep your empire together.
drek wrote:
3) What if the global queue chooses a world for producing ships that, a turn later, I decide needs a shiny new building?
Than say so. It up to us if we will save the already applied production or not. The robot wouldn't change build which are in the local queue.

drek wrote:
We know from experience that robot governors don't work well in a strategy game. A bot assigning projects to worlds is bound to make errors in judgement. The player has to constanstly police the queue, which defeats the purpose of having automation in the first place.
I don't know, but selecting places where ships should be build isn't that complicated. And even if a single decision is wrong, does it really matter?

drek wrote:
To put it another way, in a multiplayer game a player manually assigning projects locally is going to beat the snot out of a player using the automated global queue.

With a global pool of PP, and set rally/deployment points in the form of Shipyards:

1) There is only one queue to juggle.
Don't we have the same problem here? Couldn't i ask the same questions you did above? The only difference might be, that you will have much more planet than shipyards, but still a bunch of single queues.


Last edited by noelte on Thu Jul 01, 2004 7:54 am, edited 2 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 6:12 pm 
Offline
Designer Emeritus

Joined: Thu Jun 26, 2003 8:07 am
Posts: 935
Quote:
I don't know, but selecting places where ships should be build isn't that complicated. And even if a single decision is wrong, does it really matter?


If you want to play an optimal game in order to defeat another good player in multiplayer (or a cheating AI in single) then yes, mistakes do matter. In order to play an optimal game, you are saying I'll have to police the queue to ensure the queue bot isn't doing anything stuipid.

Sounds like Smac's viceroys, ie, useless. I'd do better just to turn off your queue bot and assign the projects manually to each local queue.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 6:43 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
There aren't really that many mistakes to make regarding an individual planet's build tasks. If a planet isn't building a specially located wonder then it can be a) building ships if it has a shipyard or b) building infrastructure.

In a build order, you could specify a shipyard not to use just as easily as one to use. You could also specify a group of shipyards to use for a particular order. You don't have to leave any decision up to the tool, except ones that are trivial, such as "I ordered 5 ships, and have four done/started building, and there are two shipyards that could do the last one, and one is 4 turns away and would take 5 turns to make the ship, and the other is 2 turns away but would take 10 turns to make the ship.. which should I use?" (the two options would be ones you specified that the build order could use). You could also modify a build order to include or exclude a particular shipyard after placing it.

SMAC viceroys built stuff without your input. The macro tool would only build what you order, subject to restrictions you can specify and change.

If you rally your ships somewhere that gets blocked or taken, you could be notified in the sitrep and asked to specify a new rally point. Altneratively, you could be smart and not put your rally point where it could be blocked or taken.

Edit: Even if you have pooled production, it seems to me that a tool like this would be pretty useful. And there's not really much less micro with pre-placement and pooled production, since you'd have to check all the queue for the status of all your stuff that's building.../Edit

drek wrote:
I fail to see how this is effectively different from pooling PP into a single pile--just more complex with the same effect...

Geoff the Medio wrote:
Geoff the Medio wrote:
drek wrote:
Just like research, you wouldn't be able to decrease the number of turns a project takes to complete via throwing more points at it. More industry=more projects, not faster completion time.
This, in of itself, does not eliminate the ability for any colony to produce with the full force of the entire empire. Ok, so shipyard in the middle of nowhere can't produce a Death Star in one turn... that doesn't mean it wouldn't be able to produce 400 small attack frigates in ~2 turns. This is just as imbalancing... assuming small / large ships are balanced.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 7:10 pm 
Offline
Programming Lead Emeritus
User avatar

Joined: Thu Jun 26, 2003 1:33 pm
Posts: 1092
Geoff the Medio wrote:
In a build order, you could specify a shipyard not to use just as easily as one to use. You could also specify a group of shipyards to use for a particular order. You don't have to leave any decision up to the tool, except ones that are trivial, such as "I ordered 5 ships, and have four done/started building, and there are two shipyards that could do the last one, and one is 4 turns away and would take 5 turns to make the ship, and the other is 2 turns away but would take 10 turns to make the ship.. which should I use?" (the two options would be ones you specified that the build order could use). You could also modify a build order to include or exclude a particular shipyard after placing it.


I think this makes Drek's point. Doing this for dozens of local shipyards is micromanagement at its worst.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 7:20 pm 
Offline
Designer Emeritus

Joined: Thu Jun 26, 2003 8:07 am
Posts: 935
*nod*

The UI for allowing the player to set these kinds of build orders would be a nightmare for the average user.

Quote:
This, in of itself, does not eliminate the ability for any colony to produce with the full force of the entire empire. Ok, so shipyard in the middle of nowhere can't produce a Death Star in one turn... that doesn't mean it wouldn't be able to produce 400 small attack frigates in ~2 turns. This is just as imbalancing... assuming small / large ships are balanced.


This isn't a bad point, certainly the weakest aspect of globally pooled production.

a) Maybe big ships are better than small ships.
b) The total maintaince on those 400 small attack frigates could be crushing compared to the total maintaince on 1 Death Star.

Still it's a good point. I'll have to ponder it some more.

edit: one thing: no ship should take only 2 turns to build. Low tech ships with low PP costs should be like gnats in late game...utterly unable to even scratch a high tech ship with a high PP cost.

Still the very idea of a player queueing up 200 low tech ships in place a few high tech ships in disturbing. Maybe as techs advance, the lower level, low cost ship parts are disabled--unbuildable. Civ3 disables building cavemen warriors once you reach a certain tech level, for example.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 7:45 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
tzlaine wrote:
I think this makes Drek's point. Doing this for dozens of local shipyards is micromanagement at its worst.
...?! how is this micromanagement? You only do it when you place an order or make a modification to an order. Micromanagement to me means dealing with lots of individual build queues and adding, removing and reordering things in them, and having to do this a lot.

(And who says there will be dozens of shipyards? They're supposed to be rare. (which can still be done without pooling))

What I'm suggesting is clearly macromanagement. You could add or remove a location for a build order with one click. You don't need to go through a bunch of separate screens or deal with separate queues to do this.

I don't see why this would be difficult for an "average" user to understand or deal with. It's no more complicated / confusing than control groups in an RTS.

Another issue for shared PP: You can build just as effectively at a shipyard that's very far from your core industrial worlds as you can in a core world shipyard. This means if you've got a single shipyard right next door to the enemy empire, you can suddenly produce a fleet of ships there without warning or delay due to transit. It might also become, in some cases, faster and easier to build a new fleet at this far removed shipyard than it would be to fly your fleet to the enemy... this seems imbalanced to me. Granted, it's not horrendiously bad, as you still have to build this shipyard and protect it, but it still seems wrong to me if you can build just as fast at a shipyard half way across the galaxy from your industrial worlds as you can from one near multiple industrial worlds. (yes the far removed shipyard would be an industrial world itself, and produce a lot itself without shared PP, but with shared PP, it would be just as effective as a core world and located far away, which is different)


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 7:52 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
drek wrote:
a) Maybe big ships are better than small ships.
b) The total maintaince on those 400 small attack frigates could be crushing compared to the total maintaince on 1 Death Star.
These seem like balance decisions that should be made for what make good ship combat, not so that our build system works better. Personally, I'd like to see small ships have a role even when you've got big ships around. Think marines in starcraft.
Quote:
Maybe as techs advance, the lower level, low cost ship parts are disabled--unbuildable. Civ3 disables building cavemen warriors once you reach a certain tech level, for example.
There were some interesting proposals that this might interefere with... Instead of replacing your laser scouts, you'd upgrade the lasers and they'd be still useful. See the brainstorming thread and the links therein:
http://www.freeorion.org/forum/viewtopi ... 1865#11865
(should avoid discussing this more... it's off topic)


Last edited by Geoff the Medio on Thu Jul 01, 2004 8:01 am, edited 1 time in total.

Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 8:37 pm 
Offline
Creative Contributor
User avatar

Joined: Sat Sep 13, 2003 6:52 pm
Posts: 1396
OK,

We all want your empire to be able to spend its PP on projects of various kinds

We all think these projects should have Some form of limitation of how much of your empire's production can go into them at once

The limitation can be of two forms
1. Only to do with the nature of the Project (ie a Death Star takes exactly 100 PP/tun)
2. Local conditions also influence it (This planet can only spend 20 PP/turn that one 140 PP/turn)



Part 2 can be broken down into Which Local conditions influence project production rates
a) The total locally produced production is the limitation on the total that can be spent on projects
b) others

essentially Local production is 2a
its main problem is that only Industrial worlds are effective And that means you need to have a small number of Focused Industrial worlds if you are going to keep ... Focused on a few Shipyards/planets w/ buildings.

I favor a 2b approach. other things like Total Infrastructure, Specific types of Infrastructure, Shipyard Rating, etc. could (and Should in some cases) be used.

Drek seems to favor a type 1 approach (with yes/no type limitations on placing.)



I think 2b would satisfy the goals of
1. Few important worlds/shipyards
2. No 'Industrial Starter' Colonies
3. Simplified management
4. Customizability (balancability, etc.)


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 9:59 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
Krikkitone wrote:
essentially Local production is 2a
its main problem is that only Industrial worlds are effective And that means you need to have a small number of Focused Industrial worlds if you are going to keep ... Focused on a few Shipyards/planets w/ buildings.
Your sentence stucture is a bit fragmented/confusing... but I think you're saying that it's bad to need to focus a world on industry in order to put a shipyard or wonder there. Why is this bad?

If there are wonders that work only at a farm / mine / research world, then these wonders could have appropriately reduced build costs (talking local PP only) so that they could be built at a non industrial-focused world.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 10:53 pm 
Offline
Lead Designer Emeritus
User avatar

Joined: Thu Jun 26, 2003 1:54 pm
Posts: 761
Location: Austin, TX
I think everyone who has posted in this thread has made their point. If you require more clarity, take it back to the DESIGN thread.

Let's keep this thread open for people who haven't posted yet.

-Aq

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 11:15 pm 
Offline
Programming Lead Emeritus
User avatar

Joined: Thu Jun 26, 2003 1:33 pm
Posts: 1092
Geoff the Medio wrote:
tzlaine wrote:
I think this makes Drek's point. Doing this for dozens of local shipyards is micromanagement at its worst.
...?! how is this micromanagement? You only do it when you place an order or make a modification to an order. Micromanagement to me means dealing with lots of individual build queues and adding, removing and reordering things in them, and having to do this a lot.

(And who says there will be dozens of shipyards? They're supposed to be rare. (which can still be done without pooling))


What more is there to queue management than placing an order and/or modifying it?

I thought you had conceded the point (as Aquitaine has) that local unpooled production queues will drastically increase the number of shipyards, when you mentioned this point as one of the "ok reasons" that "seems trival to you".

If this is not the case, I'll state my reasoning again: With pooling, you have a small number of shipyards, each of which can accept a relatively large amount of your empire's PP; let's say you have 5 of them. That's five queues to deal with. Without pooling, you can only build ships at the comparatively low rate of a single system's PP, so you will of course try to make as many shipyards as you can to put as much of your empire's PP capacity to work as possible building ships. Let's say you have 25 of them. That's 25 queues to deal with. I feel that fewer queues is an inherently better situation.

Quote:
Another issue for shared PP: You can build just as effectively at a shipyard that's very far from your core industrial worlds as you can in a core world shipyard. This means if you've got a single shipyard right next door to the enemy empire, you can suddenly produce a fleet of ships there without warning or delay due to transit. It might also become, in some cases, faster and easier to build a new fleet at this far removed shipyard than it would be to fly your fleet to the enemy... this seems imbalanced to me. Granted, it's not horrendiously bad, as you still have to build this shipyard and protect it, but it still seems wrong to me if you can build just as fast at a shipyard half way across the galaxy from your industrial worlds as you can from one near multiple industrial worlds. (yes the far removed shipyard would be an industrial world itself, and produce a lot itself without shared PP, but with shared PP, it would be just as effective as a core world and located far away, which is different)


This is true with or without pooling. Given enough time, you can build up a border world to be a ship-building powerhouse. With pooling, you still need to wait X turns spending Y/turn to get the shipyard, which should take a while. In particular, with something like a shipyard, the X should be fairly large, whether or not the Y is. With or without pooling, you should not be able to build a high-capacity shipyard at your border in a short period of time.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jun 30, 2004 11:38 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12042
Location: Munich
replied in design thread

http://www.freeorion.org/forum/viewtopi ... 2444#12444


Top
 Profile  
 
 Post subject: Begging your pardon
PostPosted: Thu Jul 01, 2004 2:13 am 
Offline
Creative Contributor

Joined: Tue Oct 07, 2003 1:00 am
Posts: 440
if I may make one last suggestion - not for or against anything, but perhaps something to reflect upon while considering the Production pooling issue.

=> Basically, how best does the proposed system we are contemplating serve a game that is very flexible and wide-ranging. At one extreme, we have a small galaxy world (40 stars) with very few planets, lots of computer opponents, and low tech. The amount of viable planets under the player's command could range from around 1 to 10.

At the other extreme, we have a huge galaxy (500 stars) with many planets, few computer opponents, and high tech. In this case the amount of viable planets under the player's command could range into the thousands.

Then, there is something in between. An empire growing from that 1 planet to the 1000 planet juggernaught. Does the system foster early game interest, while still being manageable in the late game? How well does it adapt as time goes on? Does it lend itself to being adaptable?

I like the discussions in this thread, because so far I've been swayed by one camp, then drawn back to the other... Perhaps the solution may be to have an option for both.. (Perhaps in early game everything is built at the planet level. Late game the player wants could toggle a 'switch' and have that planet transfer it's PP's into the global fund - with appropriate penalties and balances, of course??? I dunno)

- - - -

Will now be posting future arguments in design forum.

_________________
There are three kinds of people in this world - those who can count, and those who can't.


Top
 Profile  
 
Display posts from previous:  Sort by  
Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 90 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group