Public Review: Buildings

Past public reviews and discussions.
Message
Author
User avatar
Krikkitone
Creative Contributor
Posts: 1559
Joined: Sat Sep 13, 2003 6:52 pm

#61 Post by Krikkitone »

Prokonsul Piotrus wrote:
3. Imperial Caps..works because it automatically does so (there are only a limited number to deal with because the game says so)
What do you mean by imperial caps? I hope it wouldn't be something like in Civ3 where at some arbitrary level of cities build, happines of civ would plummet?
No by 'Imperial Caps' I mean a empire may only have so many of a certain building. (like Civ 3 Small Wonders.. although there the Cap was always 1.)

emrys
Creative Contributor
Posts: 226
Joined: Fri Oct 24, 2003 3:44 pm

#62 Post by emrys »

@guiguibaah:

So now you're proposing a system where you first select everything you want to build, then wait a few turns, then decide where to place it, then wait a variable number of turns depending on the location.

So everything you want to build requires you to revisit the decision at least once, after a few turns gap, and takes two sets of time to build. I can only see that leading to "Why did I want that warp interdictor, erm yes hang on, oh that was where I was going to put it... I think..." and "why do I have to wait twice for the same damn thing!"

Why not just combine the two parts of the decision (what and where) and make them at the same time, thus saving thought, and having only one wait for the item, i.e. pre-placement.

After all it seems we would have to put so many restrictions on the post-placement system to prevent it being a licence to 'cheat' or just simply annoying to opponents that we'd end up with a system that was pre-placement in all but name, and clunkier to boot.

LithiumMongoose
Audio Lead Emeritus
Posts: 188
Joined: Fri Jul 25, 2003 1:52 pm
Location: Cincinnati OH, USA

Re: Selecting item construction

#63 Post by LithiumMongoose »

guiguibaah wrote:Either
1 at 100%
2 at 50% / 50%
2 at 75% / 25%
3 at 43% / 33% / 23%

So you're not stuck waiting for 50 turns to construct a death star, while you have a zillion frigates waiting behind it.
Artificial constructs. Whenever it can possibly be avoided you're better off not imposing things like this on the user. Besides if you only had one build queue, one wonders why you'd put a Death Star in front of a zillion frigates in the first place. A far better solution would be to allow re-arranging things in the queue, maintaining any work that's already been done on them. (Nevermind that you're better off with serial production anyway, if it's available at all which is what I was talking about in my previous post.)

@emrys:

Nail on head, imho.

guiguibaah
Creative Contributor
Posts: 441
Joined: Tue Oct 07, 2003 1:00 am

#64 Post by guiguibaah »

So now you're proposing a system where you first select everything you want to build, then wait a few turns, then decide where to place it, then wait a variable number of turns depending on the location.

G - Both actually. You may select items beforehand from the build list and place them on the planet before their parts are constructed and available for placement. Once they are completed they are automatically placed.

Image
It could look a little like this. By dragging items from the build list onto the planet they get placed in the imperial build queue (or, you can do the reverse, by ordering it in the build queue and placing it later). It offers both the ability for both pre-placement and post-placement.

The extra 0-4 turn wait on underdeveloped systems is to prevent what happens often in classic pre-post systems like the CIVs or Moo2. You crank your taxes to full, select the city and keep "buying" those tanks every turn until you've wittled down the oncoming invading army.

- -

I think the mathematical equation would look like this.
Turns for item completion (on the planet) = [ Bcos - Epro ] + [ 4 - Pval ]

Epro is total empire industrial production
Bcos is building part cost.
4 is arbitrary number I just through up of.
Pval is the "GDP" of the planet described below.

4 - 100% of planet has cities
4 - 75% of planet has cities, planet's focus is industrial
3 - 75% of planet has cities
3 - 50% has cities, planet's focus is industrial
2 - 50% of planet has cities
2 - 25% of planet has cities, planet's focus is industrial
1 - 25% of planet has cities
0 - New colony

You wouldn't have to wait if you built them in your core industrial worlds (Pval of 3 or 4). They should pop up immediately.

= = = =

Why not just combine the two parts of the decision (what and where) and make them at the same time, thus saving thought, and having only one wait for the item, i.e. pre-placement.

G - If you chose to pre-place by dragging the item from the build list, you wouldn't have to worry about waiting twice. A system that offers both gives you the ability to order items from the empire build queue and let it stay in storage, giving you a flexibilty for drastic times. EX: A border colony was hit hard by an ion storm and lost it's hospitals. Since you so wisely chose to create a 'safety surlpus' of 1 hospital, your feighters rush to the planet with pre-built hospital parts and machinery to complete the hospital in 2 turns.

= = = = =

After all it seems we would have to put so many restrictions on the post-placement system to prevent it being a licence to 'cheat' or just simply annoying to opponents that we'd end up with a system that was pre-placement in all but name, and clunkier to boot.

G - I agree, the waiting time is there to prevent the rampant cheating that occurs in pre-placement games like CIV, MOO2, etc that allowed you to instantly 'buy' production. It gets frustrating when you are attacking an enemy border colony and every turn a death star pops out, built from a huge empire 13 jump nodes away.

G- The advantage of having a both pre and post placement systems is that not everything may need to be placed at a 'planet' for production. A system using both pre and post-placement allows for the option of placing things on systems, stars, ships, black holes, dancing rabbits signing about toilet paper, and moons. Or what the heck, you could even order the parts in your build queue and give them to your allies. Would be interesting if the game allowed for unique buildings.

G - I guess I'm saying to pre and post, "why not both?"

= = = =




= = = =

RE: Option to split imperial build Queue
1 at 100%
2 at 50% / 50%
2 at 75% / 25%
3 at 43% / 33% / 23%

So you're not stuck waiting for 50 turns to construct a death star, while you have a zillion frigates waiting behind it.

= = = =

Artificial constructs. Whenever it can possibly be avoided you're better off not imposing things like this on the user.

G - Why not? If you don't want to split your imperial build queue, just leave it at 100%.

Besides if you only had one build queue, one wonders why you'd put a Death Star in front of a zillion frigates in the first place. A far better solution would be to allow re-arranging things in the queue, maintaining any work that's already been done on them. (Nevermind that you're better off with serial production anyway, if it's available at all which is what I was talking about in my previous post.)

G - True, rearranging items and distributing production is a good idea (I like it). Splitting the queue offers players the option of "Ok, I'll split my queue in half, 75% and 25%... and just build all my military stuff in the 75% and all my civil stuff in the 25%", cutting down on some micromanagement.[/b]
There are three kinds of people in this world - those who can count, and those who can't.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#65 Post by drek »

....


Splitting the queue up into percentages (75%/25%, or whatever) is silly. Of course the player would want to build a single project as fast as possible, in order to maximize ROI--it would be bad strategy to split the queue and develop a bunch of projects slowly rather than one project fast.

Plus, the split implies a slider. No sliders, please.

As I've said before, the build queue ought to resemble the research queue. Each project takes a set number of turns to complete, and a set number of point per turn must be donated to the project. In that way, Industry (like Research) becomes a capacity--the more Industry an empire has, the more ongoing projects an empire can work on.

Emrys thoughts on post-placement mirror my own. Nothing to add.

LithiumMongoose
Audio Lead Emeritus
Posts: 188
Joined: Fri Jul 25, 2003 1:52 pm
Location: Cincinnati OH, USA

#66 Post by LithiumMongoose »

drek wrote:As I've said before, the build queue ought to resemble the research queue. Each project takes a set number of turns to complete, and a set number of point per turn must be donated to the project. In that way, Industry (like Research) becomes a capacity--the more Industry an empire has, the more ongoing projects an empire can work on.
I understand the argument for that, however I must respectfully disagree. Research is a fundamentally different animal. It can seem like a bottomless pit for spent effort until you finally have a breakthrough, because you're searching for something small (the right formula, theory, material, or whatever) in a vast sea of possibilities. A set amount of effort (points) over a set amount of time makes sense here because, past a certain point (ie the point cost), you usually can't hurry this sort of thing. Sure there are exceptions, but I doubt any other research model on this order of simplicity would do much better.

However with construction, it's a question of doing, not finding. And doing can always be sped up by throwing more money/workers/factories/whatever at it. (Don't take that as me advocating rush buying of production with money, b/c I'm not, but it fits for the theoretical argument I'm trying to make.) The returns will probably not be as good as linear past a certain point, and certainly it's possible to make things worse rather than better if your system isn't robust and efficient enough to handle all the input, but it *can* be done.

On a more practical note, it's sounding like the only things that will be available to build are wonder-esque buildings (either limited by caps or built-in drawbacks), and ships. Just taking the latter as an example, under your system every ship would have a fixed build time, in turns, probably based on hull size and other factors that went into its design. I can see a minimum # of turns being imposed, if ppl so desire, to account for the diminishing returns effect of adding more and more input, but a fixed value here feels artificial. Consider a small empire that can just barely meet the input requirement for a 20-turn Death Star, and a large empire than can meet it 4 times over. After 20 turns the former would have one such ship and the latter would have 5, but before then the large empire has no advantage.

In other words, you're trying to disallow the full-empire serial-build by imposing artificial values on how long something takes to build. Emrys' approach also forces parallel building, but uses the existing game world to do it: any planet within a certain distance of the build site can contribute. It is then up to the player to set up their production hubs as efficiently as possible, limiting overlap and still trying to keep things optimal from a military/tactical standpoint. Because of this challenge, and the fact that it gives map geography more importance, this is the approach I favor.

emrys
Creative Contributor
Posts: 226
Joined: Fri Oct 24, 2003 3:44 pm

#67 Post by emrys »

guiguibaah wrote:If you chose to pre-place by dragging the item from the build list, you wouldn't have to worry about waiting twice. A system that offers both gives you the ability to order items from the empire build queue and let it stay in storage, giving you a flexibilty for drastic times. EX: A border colony was hit hard by an ion storm and lost it's hospitals. Since you so wisely chose to create a 'safety surlpus' of 1 hospital, your feighters rush to the planet with pre-built hospital parts and machinery to complete the hospital in 2 turns.

...

G - I agree, the waiting time is there to prevent the rampant cheating that occurs in pre-placement games like CIV, MOO2, etc that allowed you to instantly 'buy' production. It gets frustrating when you are attacking an enemy border colony and every turn a death star pops out, built from a huge empire 13 jump nodes away.
To me these two points seem to be in contradiction to each other. If your arbitrary delay on placement allows one, it should allow the other, if it prevents one, it should prevent the other. Unless the hospital costs a heck of a lot less to build than the death star, i.e. a hospital could actually be built on this world in about the same time as the placement delay, in which case why not just build the thing.
I think the mathematical equation would look like this.
Turns for item completion (on the planet) = [ Bcos - Epro ] + [ 4 - Pval ]
I presume this is not what you intended to write, since if the total empire production level were more than 4 units above the cost of the item, you'd be able to place it anywhere in negative time. I find it likely that any workable delay formula will end up effetively giving you pre-placement. Because of course that's the point. The reason we hate the Civ style "buy a new army each turn in a border world", is because it seems totally unreasonable, because an 'undeveloped' world should be able to turn out a lot less than a 'developed' world, in fact, less by roughly the proportion of their development. Otherwise what does 'undeveloped' mean? The whole reason you're wasting your time picking off this worthless world is because it's worthless. I.e. it shouldn't be able to effectively defend itself without plenty of planning by your opponent (i.e. unless it's really a trap).
G- The advantage of having a both pre and post placement systems is that not everything may need to be placed at a 'planet' for production. A system using both pre and post-placement allows for the option of placing things on systems, stars, ships, black holes, dancing rabbits signing about toilet paper, and moons. Or what the heck, you could even order the parts in your build queue and give them to your allies. Would be interesting if the game allowed for unique buildings.
Pre-placement would allow for all these things as well, we are not intrinsically restricted to allowing building only at planets. In fact, I personally would like to see the number of projects that get built at planets be pretty small, e.g I'd like to see all ships get built at shipyards which exist at system level.

Under your placement method, you can't work out the build delay until you know where the thing is to be placed, because the location affects the delay. In the pre-placement model you can't work out the build time until you know where the thing is to be placed, because the location affects the maximum build rate, nothing in that says that an asteroid in the middle of nowhere can't have a maximum rate at which industry from you planets can be applied to it, or even that the enemies homeworld couldn't have a maximum build rate at which you can build something you've offered to construct for them under a treaty.
lithium mongoose wrote:However with construction, it's a question of doing, not finding. And doing can always be sped up by throwing more money/workers/factories/whatever at it. (Don't take that as me advocating rush buying of production with money, b/c I'm not, but it fits for the theoretical argument I'm trying to make.) The returns will probably not be as good as linear past a certain point, and certainly it's possible to make things worse rather than better if your system isn't robust and efficient enough to handle all the input, but it *can* be done.
This is simply a realism argument, and quite debateable at that (I'd think most real world large scale projects probably experience very rapid diminishing returns above their optimal rate). From the game point however it is irrelevant. If we say you can only build a given rate (i.e. a project takes this much per turn for this many turns, no more, no less), that's that. Equally if we say projects have a fixed industry cost but you do that in a hundred turns or one if you want (a la Civ ad absurdum), that's that. Or if we pick some arbitrary overdriving rule like moo3 or the expense of paying for production in Civ that'll be how it works. Since I'd say any variation is about as intuitive as any other, the key issue is what effect our choice has on the game dynamics, not how realistic, or aesthetically pleasing it is.
Consider a small empire that can just barely meet the input requirement for a 20-turn Death Star, and a large empire than can meet it 4 times over. After 20 turns the former would have one such ship and the latter would have 5, but before then the large empire has no advantage.
Well, the larger empire would have 4, so that's hardly a bad situation. More importantly, under the pre-placement multi-location parallel build system advocated by many, the larger empire could build the deathstar, and be doing something else with it's spare 3* capacity, say sequentially building a fleet of 5 turn support ships?, so they'd end up with the support ships rolling off the line as they were completed, i.e. in four waves, and the death star ( or two) finishing up at the same time as the fourth wave of support ships.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#68 Post by drek »

However with construction, it's a question of doing, not finding. And doing can always be sped up by throwing more money/workers/factories/whatever at it.
Debatable, but moot as it is a realism arguement.
Emrys' approach also forces parallel building, but uses the existing game world to do it: any planet within a certain distance of the build site can contribute.
If planet X is the middle of planet A and B, which planet recieves the benefit of X's industry? I think it would be difficult to represent and control the flow of resources between systems on the UI.

LithiumMongoose
Audio Lead Emeritus
Posts: 188
Joined: Fri Jul 25, 2003 1:52 pm
Location: Cincinnati OH, USA

#69 Post by LithiumMongoose »

drek wrote:Debatable, but moot as it is a realism arguement.
True. My apologies. It just felt like something that should really matter in this case. You're of course correct, it doesn't...
drek wrote:If planet X is the middle of planet A and B, which planet recieves the benefit of X's industry? I think it would be difficult to represent and control the flow of resources between systems on the UI.
Yeah, this occured to me earlier as a potential problem as well... Well, depending on how the galaxy map is put together, I doubt X would ever be identically equadistant from A and B, so whichever one it's closest to is the one its industry would go to. But you'd need some sort of visual overlay for the user to see what's in range of what, what the effect of overlap in a certain situation is, etc... Doable, but a bit of a mess. I guess the idea of a system or small area of systems with a lot of planets well-suited to heavy industry being especially special, was a bit much to hope for. Oh well.
emrys wrote:I'd think most real world large scale projects probably experience very rapid diminishing returns above their optimal rate.
I guess what I was describing earlier was exactly MOO3 overdriving, which I'd forgotten about, somehow... But you're right. Just because a large project has an optimal rate that's also very large doesn't mean the optimal rate doesn't exist. Sorry I fell into that trap.
emrys wrote:Equally if we say projects have a fixed industry cost but you do that in a hundred turns or one if you want (a la Civ ad absurdum), that's that. Or if we pick some arbitrary overdriving rule like moo3 or the expense of paying for production in Civ that'll be how it works.
I'm in favor of pre-placement and parallel building. That much I'm sure of. The problem is full-empire serial building is preferrable if it's an option. So how to disallow it. Nothing in the above quote does. The fixed-cost, fixed-time approach does because of the fixed-cost part which sets a maximum industry spent per item, which leaves leftover capacity, which goes to waste if you don't spend it elsewhere, so parallel.

How about the basic fixed-cost, fixed-time mehod with an overdriving option then? The idea being you could shave some turns off the time to build by letting the item use significantly more industrial capacity than it normally does, but at an obviously-inefficient ratio... I could live with that. No more than 25-50% off the normal build time, not talking about a 1-turn solution here, just *some* sort of way to accelerate things (at the cost of other things) if the need arises... i.e. your empire is being invaded, those support ships don't do any good without the Death Star in front of them, the big beast might be enough to hold a key choke point by itself but only if it gets there sooner than normal build time will allow for... Unlike rushing it with money, which then becomes stockpiled production, you're taking it out of current active capacity; in the example you're costing yourself the support ships which would've otherwise built in parallel...

guiguibaah
Creative Contributor
Posts: 441
Joined: Tue Oct 07, 2003 1:00 am

Hospitals and other stuff like that.

#70 Post by guiguibaah »

To me these two points seem to be in contradiction to each other. If your arbitrary delay on placement allows one, it should allow the other, if it prevents one, it should prevent the other. Unless the hospital costs a heck of a lot less to build than the death star, i.e. a hospital could actually be built on this world in about the same time as the placement delay, in which case why not just build the thing.


I think I see your point - Under your model, you would select "alpha centauri A", and flip into the build list say "Build Hospital here" - in which your entire empire would shunt production into the item, allowing it to be built in about 15 turns (since that's how big a slot building project (or baby wonder) would take).

I like that idea too. But how would you slow down production on un-industrial worlds, or border worlds?

You do raise an important point about building things that could take the empire only 1 turn to complete that I didn't think about. I'll have to get back to you on a solution... As for now, I can suggest that all 'slot wonders' or 'slot buildings' or 'baby wonders' like hospitals take at least 10-15 turns to build.

But I do prefer having them in an empire inventory, and taking longer on underdeveloped worlds to install. The lock-and-key effect, as said by Henri the 8th. I could even trade a 15 turn hospital to the trilarians for a 10 turn Oceanography studio that I can't build since I'm the Sakkra. Or, just queue up 6 4-turn frigates, and when they are ready, drop them in a system that has a shipyard.

My vote is option 2a with build option 3.
There are three kinds of people in this world - those who can count, and those who can't.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#71 Post by drek »

I like that idea too. But how would you slow down production on un-industrial worlds, or border worlds?
We don't.

As it stands now, Farm worlds and Science worlds export their product to the entire empire, without penalty. Why should industrial worlds be special in this regard?

(the except would be blockades cutting off a system from the rest of the empire, but that's a v.4 issue.)

For me, it's an ease of use issue. If industry and minerals work as food and reasearch does, the UI is cleaner.

Tyreth
FreeOrion Lead Emeritus
Posts: 885
Joined: Thu Jun 26, 2003 6:23 am
Location: Australia

#72 Post by Tyreth »

drek wrote:....
Splitting the queue up into percentages (75%/25%, or whatever) is silly. Of course the player would want to build a single project as fast as possible, in order to maximize ROI--it would be bad strategy to split the queue and develop a bunch of projects slowly rather than one project fast.

Plus, the split implies a slider. No sliders, please.
This sounds very familiar - such an argument also occurred on the old forum's thread about technology. Some wished to use sliders for technology to distribute RP's. I pointed out that it is far better to maximise each specific area for a project to conclude, then switch that slider to 100% in another area to complete that...meaning sliders introduce a micromanagement nightmare.
As I've said before, the build queue ought to resemble the research queue. Each project takes a set number of turns to complete, and a set number of point per turn must be donated to the project. In that way, Industry (like Research) becomes a capacity--the more Industry an empire has, the more ongoing projects an empire can work on.
This is what I think too. It keeps the game using a similar system rather than numerous ones, and reduces the learning curve.

Regarding slots - I think it is good to place an arbitrary limit in order to increase strategy. It reduces your ability to generalise your system/planet's capabilities, and forces you to specialise. When you specialise you need to cater more for your individual strengths and weaknesses - weaknesses that your enemy can take advantage of.
Also, without slots we will inevitably introduce other limits - eg, population or maintenance costs. The difference here is that with slots it is more immediately obvious what the planet's capacity is, whereas with maintenance, population, etc, you need to do some more in-depth research to see.
Slots could be available based on current planet infrastructure and/or population and/or size.

As for empire wide queue, I'm beginning to see why it might not be advantageous. If most buildings are more specialised in their purpose (which is our intention), then your decision is far more likely going to be related to a particular planet's need. Meaning your decision order will be:
1. What is my need at this time
2. Where can I address that need
3. What do I build to address it
Whereas with more generic buildings like the ones in MOO2, SMAC, etc, you think:
1. What is my need at this time
2. What do I build to address it
3. Where can I address that need
The latter favours an empire wide building queue. There is something about an empire wide queue that is appealing, but I can't place my finger on it. I don't like it if it means that I am going to be placing my buildings as soon as I decide to construct them every time.

noelte
Juggernaut
Posts: 872
Joined: Fri Dec 26, 2003 12:42 pm
Location: Germany, Berlin

#73 Post by noelte »

@contructing buildings, planet queue vs. emipre queue

First, the main reason why i don't like a empire build queue (at least when it comes to buildings) is the well known post placing and i don't think an argument as "it don't matters whether things pop up out of nowhere or not. It's a video game" is valid, at least not for me. As already said, this way you don't have to plan but simply fullfil your need at the moment when placing buildings. I think post-placement isn't an option.

Limiting slots might be a good way to increase strategic decisions. But i also think the UI shouldn't match a specific amount of them because i'm pretty sure that we will have mods with more or less slots. Also the number of slot should be dependend of planet size, moons, technology...

If we decide constructing buildings per planet queue (very natural to me) then wonders should be handled the same way!? All what's left is constructing ships which is done at shipyards.

Right now, i think local queues are the way to go. (maybe with some kind of dev plans) I don't see that an empire queue would reduce micro.

At the moment, is the building/placing of starbases solved? Maybe i'm to closely bound to moo2 but i think it could be handle the same way, meaning: you can build defence systems via planet queue. (maybe smaller ships too). Which leads to the question "Do we really need shipyards in a small amount?". At first i liked them but don't know if i still do. They have imo the same problem as the empire queue has, placing.

drek
Designer Emeritus
Posts: 935
Joined: Thu Jun 26, 2003 8:07 am

#74 Post by drek »

I like the empire wide queue because it solves the problem of sharing resources between planets and systems, without additional game systems like the supply units in civ.

You'd preplace objects like like civ/smac/moo2, but instead of appearing in a local queue, it shows up in the global.

noelte
Juggernaut
Posts: 872
Joined: Fri Dec 26, 2003 12:42 pm
Location: Germany, Berlin

#75 Post by noelte »

drek wrote: I like the empire wide queue because it solves the problem of sharing resources between planets and systems, without additional game systems like the supply units in civ.
I don't see why that's special to empire wide queue. If a planet doesn't produce anything it's PP's is transfered to a gloabel stock (note, i don't mean stockpiling). Those PP's can be routed to planets which produce something.
drek wrote: You'd preplace objects like like civ/smac/moo2, but instead of appearing in a local queue, it shows up in the global.
I remember this point, but you will always be able to use the local queue when adding/changing/removing items. That way there is no gain from empire queue.

As i think we don't do things like "Construct 10 buildings, i don't care where to place them", the main marco argument for the empire queue is imo gone!?

So, what is the empire wide queue good for?

Locked