A Solution To Flying Spaghetti Monsters

[Zoolimar]

Hmm, this is a very good point. I'll have to think on how to fix this. Anyone have any suggestions for how to fix using systems as conduits to avoid distance debuffs?

This?
System blobbification

Or you could do something simpler with bounding boxes but in the same vein. Basically systems that are long, thin and have a lot of turns are weaker than systems that are blobby and sit in one place. Trying to do multiple lines with each touching through 1 conduit will work, but will drop the overall effectiveness of the system (shields, reactor, thrusters) due to multiple groups.

 

[Dire Venom]

I like. But can we just fix breakoff? Then a spaghetti ship gets hit once and looses lots of parts.

Unfortunatly most players computers couldn't handel it, and if youve seen what happens when a docked reactor gets loose......

 

[Reilly Reese]

Unfortunatly most players computers couldn't handel it, and if youve seen what happens when a docked reactor gets loose......

"Fix"

 

[Dire Venom]

"Fix"

You mean optimise the game or something else ???

 

[Reilly Reese]

You mean optimise the game or something else ???

As in fix it so it works...

Not that hard to comprehend. :/

 

[Dire Venom]

As in fix it so it works...

Not that hard to comprehend. :/

The game in it's current state could not handel parts breaking off for medium to large sized ships, especialy if they caused collisions as well.
The number of break off checks that would need to be made, especialy with million block ships would be insane.
Unless the whole game was rebuilt in a new engine it simply would not be viable beyound ships of a small size for performance reasons.

Large ships near planets already can cause unbearable lag, collisions between titans can crash servers imo. Imagine a thousand of those happening every second as a ship breaks apart XD

 

[Az14el]

Break off would be an awesome fix if it didn't work by generating tons of potentially clippy entities

I like the idea of needing conduits between systems groups, as well as a power cost for conduits themselves, not too sure about required thickness, just protecting them against penetration weapons on its own would make some thickness useful while trading off vs power cost

I do wonder about the general surfaces/vertexes count being overall higher on servers this way (even though it should certainly nip more extreme s;v ratio designs in the bud), but I suppose that's to be expected from systems 2.0 either way

How is the Occlusion Culling setting (config only one) going as far as CPU load these days?
I've stopped using it since getting little/no performance increase with it for most larger ships even though i have a 4.2ghz thread speed but it has been a while now

 

[Jojomo]

For you 7 people who've "liked" the OP, you'd have done more to help the suggestion if you stick with the majority and "agreed" instead...

EDIT: whoops, I'm one of the 7! Fixed now (Likes were the majority when I rated...)

 

[Panpiper]

The game needs (rather desperately) to reduce collision checking, not increase it with more breakaway chunks.

Some sort of system that rewards clumping may work, like a 5-10% bonus in block effectiveness for every identical block placed on a block face. That will encourage square or chunky rectangles for systems. Ship shape could be more or less accommodated by judicious arrangement of those resulting blocks. 10% might be too large, giving too much incentive to stay with nothing but cubed systems. But 5% might be insufficient to deter strings.

I don't know, just throwing it out there.

 

[Jojomo]

Hmm, this is a very good point. I'll have to think on how to fix this. Anyone have any suggestions for how to fix using systems as conduits to avoid distance debuffs?

Statistically we could still expect removing a single random shield block to effectively remove 50% of shields in that case, so it's still very different to/much better than current physically unconnected systems where removing a single block would still leave the shields almost entirely unaffected.

I love the idea of physically connected systems, but I wonder whether even with creative programming that performance might suffer a bit? Or perhaps the difference will be insignificant relative to other processing....

For reactors at least, as PanPiper already suggested, I think a bonus for tight grouping would be good. Reactors would tend towards spheres/cubes, which I don't think is a bad thing.
[doublepost=1508378684,1508377740][/doublepost]Damn, I just thought of another way to dodge part of this suggestion - to avoid using thick conduits you break them up into smaller conduits with system blocks.

So instead of a conduit that's 11 blocks long and needs to be 2 blocks thick, you use length 5 conduit that only needs to be 1 block thick, then one shield block, then another 5 conduit blocks...
[doublepost=1508378884][/doublepost]

I like the extra depth that it adds, but I dislike the method being used to get that as there's plenty of frustration if you have to do it all manual. There's most likely a middle ground where the frustration part is eliminated and it's not a lengthy process to just get it all to work (not efficient till you start knowing how to better build ships).

How about an auto-conduit tool in build mode? Select two blocks and it draws a straight line of conduit between them, and as an added benefit auto-calculates how thick they need to be and places blocks accordingly...

 

[Dire Venom]

Some great discusion being throwen around!

Just another take on conduits:
Instead of being required to connect to systems, they power systems in an AOE around the conduit. The thicker the conduit, the larger the area. The conduits must be attached to a reactor, and act sort of like blood vessels. E.g they do not need to touch the blocks, merly be near them.

Just think of the heart as the reactor and the blood vessels as the conduits!

Here is a badly drawn example:
(Green-no power loss, Orange little power loss, Red noticble power loss.) Conduits provide power in an AOE determined by thickness. Smaller conduits leak power.

Just like nuterients and oxyegen can be absorbed through a couple of cell walls, so could energy be absorbed through surrounding blocks!

A penalty or reduction could be applied to long lines, (as suggested above with 'resistance'), with wide conduits having little to no power loss while smaller conduits 'bleed' power out of the system.

Therefore having systems wrapped around conduits of various sizes would be the optimal build, with smaller conduits traveling out from the larger stems to power smaller areas of the ship, or simply having a really fat conduit running through the center of your ship : ).

Spaghite ships would no longer beable to function, as long lines of conduits would bleed out massive ammounts of power resulting in it being dead in the water. Making the conduits thick enough not to bleed power would singificantly increase mass and similary leave the whole thing dead in the water (as well as not being a spagheity ship anymore).

_____________________________________________________

What this suggestion accomplishes:

-Makes spaighti ships unable to power themselves due to power bleedout.
-Cutting off conduits to a section will depower blocks in that section (e,g cutting a ship in half does eactly what you think it would do)
-Ships will have a 'nervous system' of conduits which are vital for keeping all areas of the ship powered.
-Uses current mechincis and suggestions to form an effective solution.
-Conduits and 'power grids' would be an awesome tool for RP focussed players.
-Relaticly low on calculations. The AOE powering could be accomplished a number of ways, including system "efficiency".
-Does not require hooking up each induvidual block group.

_____________________________________________________

What do you guys think:? This is a combination of this discusion and I hope it might help us achieve a solution.
Additional protection for the conduits would likely need to be implimented (e.g conduits have a damage resistance dependent on your ships current HP%? Thus you will likely want to damage a ship before you can easily cut a ships conduits. Trying to use conduits as armour would still result in significant power bleed out and should be discouraged by the mechinics.

EDIT: Each joined conduit block would add to the total reactor/conduit HP pool. This pool would serve to act as shielding for the conduits/reactor. Adding more conduit blocks on would also server to better protect your key energy systems (like adding shield blocks).

 

[Lecic]

Instead of being required to connect to systems, they power systems in an AOE around the conduit.

You know, this is a really good idea. Not only does it fix the issue of hooking up small filler systems in gaps, but it also fixes the issue of using systems in place of conduits to get around the thicker conduit problem.

Just one problem I see- what prevents me from just making a blob of conduits at the center of my ship that has its AOE cover the whole ship?
I think a good idea would be to make the AOE scale logarithmically so that you still need to get conduits near the systems to power them and can't just make a giant ship covering blob.

 

[Dire Venom]

You know, this is a really good idea. Not only does it fix the issue of hooking up small filler systems in gaps, but it also fixes the issue of using systems in place of conduits to get around the thicker conduit problem.

Just one problem I see- what prevents me from just making a blob of conduits at the center of my ship that has its AOE cover the whole ship?
I think a good idea would be to make the AOE scale logarithmically so that you still need to get conduits near the systems to power them and can't just make a giant ship covering blob.

Quite true.
Sounds like a good fix to me.
-Cyclinders volume (or if its hollow) surface area scales exponientialy vs the cylinders radius.
-Cycliners AOE scales in a increasing fashion with the radius (width of cylinder).

(And yes, I love pictures, this is a cylinder^^)

Thus to increase the volume of the AOE you have to add more and more blocks for each extra block of AOE, making large conduits block intensive, but in return having no excess power loss.

Bah my brain is dead but I'm sure you get the point XD Would this be better as a seperate thread:?

 

[Jojomo]

Some great discusion being throwen around!

Just another take on conduits:
Instead of being required to connect to systems, they power systems in an AOE around the conduit.

It's a nice idea that gives a good result, but this would be even more computationally intensive than conduits being physically connected.
It would require real time distance checking of every system block against every conduit block in the ship, and real time recalculating of conduit ranges.
I don't know if that's significant relative to all the other calculations going on, but I suspect it possibly might be.

 

[Dire Venom]

It's a nice idea that gives a good result, but this would be even more computationally intensive than conduits being physically connected.
It would require real time distance checking of every system block against every conduit block in the ship, and real time recalculating of conduit ranges.
I don't know if that's significant relative to all the other calculations going on, but I suspect it possibly might be.

Good points. I was simply offering up a potiential example of how it could work (not expecting it to be implimented as is), I bet it could be done in a less intensive way with further disscusion and thought..
Maybe it could be more like the stabilizor mechanic except inverted, and instead of checking each block, it simply checks the location of the various system groupings, if within (max conduict range, easily calculated) move to the next system grouping etc. Not quite there yet but I think a solution could be found in this concept.

If the stabilizor mechanic can work without performance issues I'm sure a similar systems for conduits could work as well.

 

[Jojomo]

I think the suggestion in the OP is great, I'd love to see it all happen, but the more I think about the less feasible I think point #2 is in practice.
I don't see a way to stop people avoiding it, and even if they couldn't, I think checking the cross-section thickness all the way along the length is a difficult piece of code to get correct and robust, plus also computationally cheap at the same time (I could be wrong though, if there's an easy way other coders please comment!)

Although I like it, I think for practical reasons of making it work in a computer game, #2 might need to be dropped (happy to be proved wrong).

I think #1 and #3 by themselves should be enough to kill spaghetti meta though, providing reactor efficiency scales up with tight/dense block groupings (i.e. causing people to want to make their reactors spheres/cubes, not lines).

Turn rate being more punishing on distance (as opposed to mass) from CG would also help. (I'd really love to know details of exactly how SM calculates turn rate...any possibility of that being shared Lancake?)

 

[Lancake]

Break off was mentioned, but it's not a realistic option. While we're sure we can make it work quite well in most cases by expanding upon it and improving it where we go, there would be always some edge cases where it goes wrong.
Cutting a 10 million block ship in half comes down to instantly removing 5 million blocks of one ship, and spawning in a separate entity with that amount of blocks. That's a lot of data and both actions are a serious impact on the server, physics have little to do with it in this case.

Maybe one day it will be improved/fixed to work flawlessly up to a size, could still be useful for competitions or events where ship limits are put in place...but it's definitely not now, break off is a complicated system and will require some effort to bring it up to par again.

Performance issues aside, it would also change the game entirely as every constructed ship would most likely fall apart in some way or another. Building experience is different too as you need to make sure everything is connected. It's most likely not something everyone would want.

(This is no invitation to further discuss break off, just a clarification ^^)
[doublepost=1508407929,1508407279][/doublepost]

How about an auto-conduit tool in build mode? Select two blocks and it draws a straight line of conduit between them, and as an added benefit auto-calculates how thick they need to be and places blocks accordingly...

While that would improve the experience, not every line can be a straight connection. There would be plenty of ship examples where either the ship shape, or existing systems get in the way of that line.
The game could "try" to curve the line out of the way but I doubt it would work well, or at least how you would expect it to do.

 

[Jojomo]

While that would improve the experience, not every line can be a straight connection. There would be plenty of ship examples where either the ship shape, or existing systems get in the way of that line.
The game could "try" to curve the line out of the way but I doubt it would work well, or at least how you would expect it to do.

No, but it's some automation, reducing the manual aspect of it. For non-straight sections players could even "map" where they wanted the conduit by laying some blocks as "waypoints", and then forming straight line sections between them as individual straight connections. I think it would work very well. Simple, intuitive, fast.

 

[Lancake]

It's a nice idea that gives a good result, but this would be even more computationally intensive than conduits being physically connected.
It would require real time distance checking of every system block against every conduit block in the ship, and real time recalculating of conduit ranges.
I don't know if that's significant relative to all the other calculations going on, but I suspect it possibly might be.

Depends on how it's made, but if every block change requires the whole thing to be adjusted, then it could definitely be a concern.

---

I've got a question about something I've read before, and what I'm seeing being suggested here.
Some dislike "forced design" yet for me that is always been "encouraging you to build this way" and also, being part of the game mechanic itself. Every rule you add to a game, is forcing one design over the other. Sure, there's always the choice where usually the most efficient one (the forced design) is also the most vulnerable to weapon damage.

While I think that's completely acceptable, others did not seem to agree with me as forced design is being brought up multiple times as a bad thing.
The question would be then, is "forced design" a bad or a good thing? and if it depends, when is it an acceptable thing to have and when is it not?
And what makes these suggestions different to not be considered a bad thing? Both OP's suggestion and the one Dire Venom mentioned, forces you to build conduits or else you get nothing. You need to create these weak points as without them, the ship would simply not work at all. There's no encouragement here, you're truly forced.
Both suggestions would also eliminate floating (functional) parts of a ship, as it would either result in a physical connection going from one part to the other which is an easy to see/target weak-point. Or it would require you to make giant/thick conduits to get your AoE power far enough which would be extremely inefficient.

 

[Jojomo]

Depends on how it's made, but if every block change requires the whole thing to be adjusted, then it could definitely be a concern.

I don't see that being required - you'd have to move an entire system to need to redo a conduit from scratch (which would be a quick little task with a conduit tool like I mentioned)

I've got a question about something I've read before, and what I'm seeing being suggested here.
Some dislike "forced design" yet for me that is always been "encouraging you to build this way" and also, being part of the game mechanic itself. Every rule you add to a game, is forcing one design over the other. Sure, there's always the choice where usually the most efficient one (the forced design) is also the most vulnerable to weapon damage.

While I think that's completely acceptable, others did not seem to agree with me as forced design is being brought up multiple times as a bad thing.
The question would be then, is "forced design" a bad or a good thing? and if it depends, when is it an acceptable thing to have and when is it not?
And what makes these suggestions different to not be considered a bad thing? Both OP's suggestion and the one Dire Venom mentioned, forces you to build conduits or else you get nothing. You need to create these weak points as without them, the ship would simply not work at all. There's no encouragement here, you're truly forced.
Both suggestions would also eliminate floating (functional) parts of a ship, as it would either result in a physical connection going from one part to the other which is an easy to see/target weak-point. Or it would require you to make giant/thick conduits to get your AoE power far enough which would be extremely inefficient.

I realise that question wasn't for me specifically...
I think some people here yell "forced design" for any design aspects they don't like/are new/don't fit their goals/start with a letter in the first half of the alphabet instead of the second.... The rules of design that they do like are just as forced, but aren't labelled that way by them.

Design rules are essential. Just like sport, design with no rules would be chaotic, fun for about 30 seconds, and then incredibly boring. Design is (and should be) all about compromise. E.g. "I want to move my stabilisers over here for better efficiency, but that will eat into space allocated for weapons/make my ship heavier than I want/make the ship too large to fit through my warp gate/decrease my turn rate/etc/etc/etc....

Also, although I'm no psychologist nor marketing expert, perhaps Schine could use different terminology when describing things - e.g. instead of stabilisers "reducing efficiency" the further they are from the ideal distance away, perhaps they could have "increased efficiency"/"given bonus efficiency" as they got closer to the ideal separation distance... same thing with different words, one with a positive spin, the other negative. You see the same sort of thing continuously in advertising and politics every day, for good reason.