Post your optimal base/ship design tips and why!?!

[feelotraveller]

Yep.
 
Another small tip is if you create your designs with components in the right order you can get them powering up shields, firing weapons and/or extracting resources before the the base/ship is finished! 
 
Play around and see what you can achieve...  [:)]

[Kayoz]

ORIGINAL: feelotraveller

Realistically only one of each type of manufacturing plant is needed on a spaceport. Well in excess of 30 construction yards can be supported by one of each with no loss of speed.*

Have you actually tested your 1-plant-30-yard hypothesis? I'd love to see your results, because it's in complete contradiction to the results of everyone else.

[Canute0]

I didn't test the 1 for 30 hypothesis, but it sounds a but much for me too. 1 plant for 5 yard is more realistic for me.

The bottleneck arn't the plants or yards, most time it is one or more missing resources.

[Kayoz]

ORIGINAL: Canute

I didn't test the 1 for 30 hypothesis, but it sounds a but much for me too. 1 plant for 5 yard is more realistic for me.

The bottleneck arn't the plants or yards, most time it is one or more missing resources.

Here's how I theorized how Elliot's construction process works - it's consistent with what I've seen from testing.

ref post #5 -
http://www.matrixgames.com/forums/tm.asp?m=3096095

The result is:
1. If you want maximum ship construction speed from a port, you must keep a ratio of 1 manufacturing plant of each type per construction yard. Fewer plants can cause bottlenecks in your production.
2. More yards or plants on a construction ship gives no benefit. Only 1 of each (1 yard, 1 of each plant) is of any use.

[jpwrunyan]

ORIGINAL: feelotraveller

Everything you have written is bollocks. [:D] I just wanted to say that. (@jpwrunyan)

Yeah, I guess I am not surprised. It wouldn't be the first time, anyway. I had this nagging feeling that in the early days I got away with 0 cargo bays on my space ports. But I also remember at the time people were saying you should put mining components on your spaceports as well to get extra resources (which is definitely bollocks). So somewhere along the way I got fed up with doing so many things wrong and trying to sort the false from the true that I must have rationalized the default designs. My bad.

The whole space port/planet relationship is a bit esoteric, I think. Like the fact that planets have 3 invisible docking bays as well as their own construction yard.

Anyway, if you have cargo holds on a space port stuff does get put into them, right? Because I am pretty sure I have seen it happen... I believe stuff gets put into the cargo holds of defensive bases and resort bases as well, right? Was I halucinating? If so there must be some reason to it...

[feelotraveller]

ORIGINAL: Kayoz

ORIGINAL: feelotraveller

Realistically only one of each type of manufacturing plant is needed on a spaceport. Well in excess of 30 construction yards can be supported by one of each with no loss of speed.*

Have you actually tested your 1-plant-30-yard hypothesis? I'd love to see your results, because it's in complete contradiction to the results of everyone else.

Gee, your testing regime is non-existent. Because if it wasn't you might have learned something.

Read this thread http://www.matrixgames.com/forums/tm.asp?m=3180183&mpage=1&key=plants? for some enlightenment. If you do not understand the explanation I'll try to help. If you don't agree with the results do some testing and preesent your reults. I would be most interested since as far as I know no-one else has tested the working of plants in spaceports; certainly no-one has posted any results, let alone contested the conclusion.

[Kayoz]

ORIGINAL: feelotraveller

ORIGINAL: Kayoz

ORIGINAL: feelotraveller

Realistically only one of each type of manufacturing plant is needed on a spaceport. Well in excess of 30 construction yards can be supported by one of each with no loss of speed.*

Have you actually tested your 1-plant-30-yard hypothesis? I'd love to see your results, because it's in complete contradiction to the results of everyone else.

Gee, your testing regime is non-existent. Because if it wasn't you might have learned something.

Read this thread http://www.matrixgames.com/forums/tm.asp?m=3180183&mpage=1&key=plants? for some enlightenment. If you do not understand the explanation I'll try to help. If you don't agree with the results do some testing and preesent your reults. I would be most interested since as far as I know no-one else has tested the working of plants in spaceports; certainly no-one has posted any results, let alone contested the conclusion.

The question is of greatest potential speed - in any situation.

Once you have a bottleneck of a required resource and say 29 of 30 construction yards are paused waiting for the required resources - no construction can take place, because those yards have locked the production of the one available plant. If you maintain a 1-1 ratio, then other construction can continue as there are unoccupied plants which can produce ships (given no resource bottlenecks) while you wait for the resources.

Your hypothesis only works if the production capacity of the port is seen as a whole, in which case all production stops at the port regardless of the plant type required, in a situation of insufficient resources. However, I've seen ports continue to churn out ships when there's a lack of resources - so long as those ships do not require the lacking resource(s). Or so I seem to remember - I've gone off DW for the moment, since Shadows is looming on the (hopefully near) horizon.

Or so it seems. The editor doesn't allow us to edit cargo levels, so it's rather difficult to test this thoroughly. My experience, however, seems to suggest that a 1-30 ratio is a recipe for choking your production entirely.

But hey, you might be right - given the difficulty of testing at that granularity, I'll shrug and say "who cares?" until someone presents definitive information. 30-1 ratio seems daft no matter which way it works.

[feelotraveller]

ORIGINAL: Kayoz

ORIGINAL: feelotraveller

ORIGINAL: Kayoz




Have you actually tested your 1-plant-30-yard hypothesis? I'd love to see your results, because it's in complete contradiction to the results of everyone else.

Gee, your testing regime is non-existent. Because if it wasn't you might have learned something.

Read this thread http://www.matrixgames.com/forums/tm.asp?m=3180183&mpage=1&key=plants? for some enlightenment. If you do not understand the explanation I'll try to help. If you don't agree with the results do some testing and preesent your reults. I would be most interested since as far as I know no-one else has tested the working of plants in spaceports; certainly no-one has posted any results, let alone contested the conclusion.

The question is of greatest potential speed - in any situation.

Once you have a bottleneck of a required resource and say 29 of 30 construction yards are paused waiting for the required resources - no construction can take place, because those yards have locked the production of the one available plant. If you maintain a 1-1 ratio, then other construction can continue as there are unoccupied plants which can produce ships (given no resource bottlenecks) while you wait for the resources.

Your hypothesis only works if the production capacity of the port is seen as a whole, in which case all production stops at the port regardless of the plant type required, in a situation of insufficient resources. However, I've seen ports continue to churn out ships when there's a lack of resources - so long as those ships do not require the lacking resource(s). Or so I seem to remember - I've gone off DW for the moment, since Shadows is looming on the (hopefully near) horizon.

Or so it seems. The editor doesn't allow us to edit cargo levels, so it's rather difficult to test this thoroughly. My experience, however, seems to suggest that a 1-30 ratio is a recipe for choking your production entirely.

But hey, you might be right - given the difficulty of testing at that granularity, I'll shrug and say "who cares?" until someone presents definitive information. 30-1 ratio seems daft no matter which way it works.

Oh, come on, what a lack of intellectual rigour. It is easy to test. (And to shrug 'who cares?' after claiming that everyone else has results which show otherwise and asking for my testing results smacks of bad faith at the very least. If you don't care and really don't know why did you bother to issue such a confrontational challenge?)


First generate a galaxy (new game). I like to use a minimal galaxy both size and star numbers, for computational speed purposes, but any will do. I suggest turning off pirates and monsters so they don't interfere while you are testing. One opponent set at maximum distance. Give yourself an excellent homeworld, and the opponent a harsh one (not necessary but helps to keep them from interfering if your test runs long; excellent world means more resources for experimentation without requiring actually playing to gather them ). Starting technology (keeps throughput of resources to a minimum), oh, and keep it at a single colony for convenience.

Pause your start game. Delete (via editor or scrapping) all your ships (partly convenience partly so that they don't interfere). Save this (save1) so we can make other tests later. Retrofit your spaceport so that it has 30 yards and one of each plant (since this is what we are testing). Don't bother with anything else until this retrofit is complete. Now save your game (optional save2) in case you want to reuse it/test other similar things.

a) Order 30 ships of any given single type (for convenience since we want them to all finish at once) at your spaceport (the only limitation is making sure that you have the resources present to build them all). Pause and unpause until you find the point at which (give or take a day) the first components of each ship are built (can be slightly messy but the more precise the better). Write down the date. Now unpause until you get the message that the ships are finished. Write down the completion date and then figure out how many days elapsed between when the first components were added and when the ships were completed and divide this number by 6 (the build tic is six days) rounding slightly to an integer to eliminate messiness. This gives you the number of build tics between when the first components were added (definitive start date) and when the ships were completed. Write this number down and underline it.

b) Load save1. Retrofit your spaceport to have 30 yards and 30 manufacturing plants of each type.* (Optional make save3 in case you want to retest.) Order 30 ships which are of the same type as in a). Repeat the process as above until you have the number of tics between the definitive start and the completion date. Now if the Kayoz hypothesis is correct then tics a) will be much greater than tics b). If the feelotraveller hypothesis is correct then tics a) will equal tics b). What did you find?

*[Actually having at least 2 plants of each type will suffice for the experiment. ]


Not difficult and not particularly granular. Nor was it difficult to come up with.

Similar tests can be conducted to confirm (strictly speaking provide evidence to support) my hypotheses on construction yard speed (easy, just vary tech levels) and manufacturing plant speed (slightly more time consuming since you need to count the total size of components in each category produced each tic including those added to ships - note that the remainder is carried over as components started but not finished).


Resource shortages are a red herring and I am not sure why you introduce them. (But to be explicit for curious minds.) If there are shortages of resources for producing the components needed the plants will continue to producing other components which there are resources available for until only the components requiring the missing resource(s) are unproduced. It does not matter the number of plants you have. If there are shortages of components for ships in the construction yards then the yards will continue to build those ships with other components until only those not available are left to be built.

[Kayoz]

ORIGINAL: feelotraveller

Resource shortages are a red herring and I am not sure why you introduce them. (But to be explicit for curious minds.) If there are shortages of resources for producing the components needed the plants will continue to producing other components which there are resources available for until only the components requiring the missing resource(s) are unproduced. It does not matter the number of plants you have. If there are shortages of components for ships in the construction yards then the yards will continue to build those ships with other components until only those not available are left to be built.

It's not a red herring - that's the whole point I'm laying a question mark over. Do the plants cease production (locked waiting for resources) or do they continue to produce other components that aren't lacking in resources? I'm unsure as to which is the correct answer. If they continue to produce, then you should see your builds that are stalled for resources complete for all components BUT those which are lacking. However, I seem to recall seeing ships that were 1/4 built and nothing more done till the offending resource arrived; at which point construction continued. On the other hand, if you're correct, then everything else should have been built except for the few components. Nothing in your test addresses this question.

Perhaps you missed the whole query of "is plant locked and waiting due to resource shortage". Need I spell out the difference for you in the case of resource shortage? Do I really have to do that? Your test is not testing the mechanics of plant production, but that of construction speed as dictated by the construction tech level. You seem to have failed to notice that construction yards only build at a certain rate - regardless of how fast plants are manufacturing. To make it simple for you - you're exercising the wrong game mechanic.

My main disinterest in the whole maximum ratio is due to the fact that a planet will almost never have sufficient resource to build 30 ships at once; and that I've never been in the position where - even if I had the resources at a planet, where I needed so many ships at once. If I am on a building spree like that, I'm generally gearing up for a major war - or it's too late and the delay of waiting on the default port ratios makes no difference when considering that the ships will almost assuredly come out with zero fuel and absolutely assuredly with zero shields (making throwing them immediately into combat rather pointless). Besides which, Erik's posts seem to hint that major changes have been made and the time required to test this sufficiently will be of no value and completely wasted once Shadows is released.

Once again, I shrug - "who cares?" Default ratio works fine for almost all circumstances.

[Darkspire]

Darkspire

[justmax]

Regardless of the validity or lack thereof of any of this, some of us also take into account the 'gaminess' of certain tactics. For example, not putting cargo bays on a space port is ludicrous; and one plant of each kind might be just fine for 20 construction yards? If so, these things aren't WAD based on the default ship/base designs but bugs that people are taking advantage of. That might be good and dandy for you, but all that tells me is that you're incapable of actually beating the AI on it's own terms, so you look for cheats like this to give you an edge because somehow these bugs feel less like cheating than just opening up the editor and erasing that incoming enemy fleet about to pound your colony to mincemeat.

Of course a spaceport needs cargo bays. It also needs a proper ratio of plants to construction yards. That's the way the AI builds; ignoring it is just another way of cheating past an unintended deficiency in AI designs.

[jpwrunyan]

ORIGINAL: feelotraveller
Read this thread http://www.matrixgames.com/forums/tm.asp?m=3180183&mpage=1&key=plants? for some enlightenment. If you do not understand the explanation I'll try to help.

I wonder how I missed that thread. That was fascinating. I almost feel like maybe Codeforce accidentally made the construction capacity of plants 10x greater than they intended. I think it is obvious from the default designs that they intended 1 construction yard to be able to keep up with the speed of 1 plant. Since you need 1 of each plant to build any ship, this led to 3 construction yards per set of plants. So small space port is 3/1/1/1, medium is 6/2/2/2, and large is what again? 12/4/4/4??

Anyway, I just want to confirm, first of all, the way construction works is that each plant will manufacture spaceship part associated with its type. So energy plants will build engines, shields, etc. Weapons plants build armor and weapons. High tech plants build colony modules, command centers, etc.

Once the parts are constructed, the construction yards attach them to spaceship under construction (basically matching the part to the blueprint). Construction yards have a one-and-only-one relationship with something being built. If you order a ship, you need one construction yard to build the ship and only one construction yard can be used to build that ship. In this case naval ship construction has always been a helpful analogy for me. You build the parts in the factories and them cart them over to the ship chassis sitting in the bay (at a dock).

So to establish a common lexicon:
Manufacture means the assembly of components from raw materials (which are then put into storage... you can see them in storage on a planet)
Construction means attaching a component in storage to other components onsite to build a ship (or base in the case of Constructors and planets)

Aside: the reason Constructor ships only have 1 construction yard is because they can only build one thing at a time. So they are automatically always bottlenecked in that regard.

Finally, manufacturing plants never pre-fabricate anything. When an order comes in at a site to build 10 spaceships, the manufacturing plants just see that they need 10 command centers, 50 armor plates, 20 reactors, etc. and start building them. So there is always a lag between construction yards actively constructing and the first available component being manufactured.

I think we all understand this and no one is disputing it, right? This is the in-game process.

The thing is, manufacturing plants can produce components at such blindingly fast speeds that it would require dozens of ships to be constructed simultaneously at the same location for any bottleneck to occur on the manufacturing side. I had always assumed that the time it took (for example) for a Hightech Plant to produce 1 command center component was the same as the time it would take for a construction yard to attach 1 command center to 1 ship (thus 3 construction yards per plant set being optimal in most cases). But if I understood the other thread correctly, it would be more like 10 command center components could be constructed in the time it took for a construction yard to attach it to one ship. Yes, I am assuming the manufacture of a component doesn't happen simultaneously with the construction of that same component, I am just trying to keep it simple here.

So what we are essentially arguing about is the concrete values for time it takes to manufacture and time it takes to construct, right? Because it is these speed values that I find myself not understanding.

edit: math ratio error and fatal typo

[justmax]

You're right; it's a bug, a mistake in the game. The manufacturing plants produce items far more quickly than they can be used by the construction yards. If you look at the default base designs (as you clearly did) then you'll see right off the bat there's an uncorrected error in the code. Things should work exactly as you've outlined but they don't, leading to potential cheats.

[jpwrunyan]

Wow, that would be a bug worthy of Initech. Can someone from Matrix maybe comment on this? Or do I need to necro this thread every 7 days with a post about Dwarf Fortress?

I'll do it.

[Tanaka]

ORIGINAL: jpwrunyan

Wow, that would be a bug worthy of Initech. Can someone from Matrix maybe comment on this? Or do I need to necro this thread every 7 days with a post about Dwarf Fortress?

I'll do it.

[:D]

[Darkspire]

ORIGINAL: jpwrunyan

Wow, that would be a bug worthy of Initech. Can someone from Matrix maybe comment on this? Or do I need to necro this thread every 7 days with a post about Dwarf Fortress?

I'll do it.

Blow the construction yards, they work, its the the design files in conjunction with the policy files that need looking at [:D]

Darkspire

[feelotraveller]

First up, this thread is about 'optimal base/ship designs'.  Whether you want to build your spaceports with zero cargo bays or include only one plant of type each is a separate issue.  (As an aside: I play to have fun and have finished my games many ways.  Cheating -which this is not, strictly speaking- is a concept which I don't understand in a single player game since is not a competition but rather an entertainment.  I personally find it much more 'gamey' to run around conquering cpu player homeworlds, although I did have fun game that way once. [:D]  But each to their own.)

Yep jpwrunyan, that's a good summary.  Speed for construction yards is the figure listed on the tech screen divided by 100 (so starting tech construction yards build 2 components per tic) which gives the number of components fitted to a ship in a yard per build tic.  Speed for manufacturing plants is the figure listed on the tech screen divided by 100 which gives the total size of components that the manufacturing plant can produce in a tic (so components totalling 300 size for starting tech).  Each component is manufactured by the plant corresponding to its tech tree category.  A couple of fine grained points to note: components can be fitted in the tic they are produced and components can be started in one tic (using part of the required capacity for their production) and finished in another.

It was my initial conclusion/reaction that this was a result of a mistake in the game code and that the plants were producing 10 times more quickly than they should.  However I have mentioned this fact many times now on the forums and even opened a tech support ticket stating this view months ago and there has never been a reply or comment from the developers/publishers nor any adjustment of the code (which on the face of it looks like a quick and simple fix).  So now I am not so sure.  Perhaps they are not listening, but perhaps it is as intended or very hard to fix.  Their continuing silence leaves this issue unresolved (in my mind at least [;)]).

Edit: it turns out it was over a year ago. Go here if you want to add your concerns/bring it to the attention of the powers that be.
http://www.matrixgames.com/forums/tm.asp?m=3018477&mpage=1&key=plant&#3018477

[justmax]

I think the issue is pretty straight-forward, regardless of a response in the tech forums: a) the standard designs include cargo bays in the space ports and a certain set number of plants per construction yards; it doesn't take a rocket scientist to figure out what that means, and b) the AI uses these designs. If exploiting a bug to get yet another advantage over the AI is oky-doky with you, then by all means - go ahead. Nobody in the world is going to care what you do in your single player game, on your computer, except for you.

Me, I see this as a cheat no different than opening up the editor and doing unto the enemy with the 'delete item' cursor. Cheats spoil the experience for me, so I avoid them.

[jpwrunyan]

I don't think anyone is calling anyone else's gameplay "cheating", "exploitation", or any other thing that merits a defense. For example, if I reload a game I consider that a personal defeat. But anyone here who reloads after a mistake doesn't have to feel I am disparaging them. They just aren't as totally awesome as me. But that was predetermined already. We all have opinions as to what is exploitative and how it sits with us. The fact that this debate can even take place is because the game is woefully unchallenging at the moment.

And no one is saying "cheats make the game fun, YAR!". It's all just opinions. You can't have a discussion about optimal design (such is the OP) without also discussing exploits. That's just due diligence. I think we can discuss what the exploits are without endorsing them. And we can say "you should have 0 cargo bays and 30 construction yards on your spaceport" to make a point about the game mechanics and not actually be encouraging the behavior--not that it matters. That is how I have understood the conversation so far.

Anyway, I apologize if I am misreading your reply. I just personally don't want the discussion to become about how the game should be played because I find that boring and emotional. I would rather understand the mechanics and shed some daylight on what's going on.

Finally, I also bumped the tech support ticket. At least 3 people here (including me) seem to think this is a bug.

[Bingeling]

ORIGINAL: justmax

I think the issue is pretty straight-forward, regardless of a response in the tech forums: a) the standard designs include cargo bays in the space ports and a certain set number of plants per construction yards; it doesn't take a rocket scientist to figure out what that means, and b) the AI uses these designs.

Cargo bays on spaceports makes total sense. Any real spaceport would have them. There are also all kinds of good game reasons why blowing up an enemy spaceport should not destroy all (or much) of the cargo present at that colony. Which makes having cargo shared with the colony a very good idea. I don't think "useless cargo bays on spaceports" make the top 100 list of AI issues with designs... It hardly matters.

There could be some of the same behind the plants. Something going quite belly up in construction if capacity is limited. Or it could be a simple typo. As it is now, construction works, and stops once there are resource shortages. I can imagine quite a few changes to plants (like blocking all construction if resources are missing for the "current" job) that could screw things quite badly. Resources and construction blocks are hard enough to understand even if blocked plants are not added to the mix.