Space Engineers Feedback

71
Votes
Ships are too delicate in collisions, and ore spawning is like having a warhead go off.
Two things that are all too often directly related. FIRST: Ships are too delicate. Heavy armor seems ok in terms of soaking up damage but barely slows the ship down at all when faced with voxels at speed. Literally any other part feels like it's made out of glass. Say you have a mining ship with drills on the front and are coming up to a rock. There's a certain speed where damage is ignored, maybe under 3 m/s or so. But what if server lags and it thinks you came in at 10m/s? The drills just pop and the contents dump out. Ship is not even slowed down, it's entirely possible for this little fender-bender to demolish the front half of the ship, and i think that's only because that's where the ship happened to stop, not because of a meaningful impact energy reduction between the heavy steel parts and the dirt. If you accidentally cut power or dampeners while a whole 5 meters in the air? DESTRUCTION. Any thruster or non-armor part that touches a voxel just dissappears as if it were a popped balloon of no mass at all. Is that connector magnet a bit strong? BOOM. Now you don't have a connector, problem solved. I think you see where this is going. Our creations have that industrial look but not the industrial durability. I've been on servers that run shield mod for this exact reason, because random little clangs and server lag spikes are the most lethal weapons in the game. While I won't presume to know a terrible lot about how the game engine handles physics, I am going to go out on a limb and say that I think the expectation is that several tons of heavy welded iron parts are much hardier than the dirt and rocks, and that the roles should be reversed, bumping the surface at reasonable speed should leave a crater on the ground not in the ship. Voxels seem to be considered immovable objects, but should not be calculated as such in this situation. The numbers need tweaked by a factor of (at least) 100. Not even kidding. A large ship hitting an asteroid and it has armor on the front, 2.5 meters of crumple zone. That should slow it down considerably, STOPPED if it was going any less than 20-30m/s. Depending on weight of course, a monster ship would take a lot to slow down. But something like that should stand a good chance of just powering through the whole asteroid like an icebreaker ship! The crumple zone is key, most clang seem to be zero-time movement events, and the engine assumes huge magnitudes of energy must have been injected to create them. This is obviously wrong. The movement energy should determine the outcome, not the other way around. Causing damage to the blocks or to the voxels has to use up this energy, AND buy us TIME. Also, the non-crumpling blocks REALLY need a visual way to display damage. SECOND: When an ore ball spawns (usually because drill or cargo popped) it might as well be a warhead going off. In my little mining ship example, if the connector pops and it was full of an ore, it literally guts the entire bottom of the ship out. It's ridiculous. ITEMS SHOULD VERIFY CLEARANCE BEFORE SPAWNING. This problem is bad at 1x inventory, the ore ball often physically larger than the container it was in, but at 10x it's just plain silly. Spawn it at 10 small rocks if you have to. Incidentally, when this happens over land it leaves a MASSIVE crater. So much energy?


AutoMcD shared this idea 06/09/17 02:13
AutoMcD 07/09/17 15:38
I would really like to see blocks that are intended for contact (drills, welders, landing gear, connector, merge, etc) have a dampening function built-in. Obviously the tools would be on the minor/stiff end of this, and the landing gear would be designed to absorb a lot more energy. If the impact is too much, they should crumple a bit like the armor does instead of pop. Look at this chart: http://www.enidine.com/CorporateSite/media/itt/Products/Theory2B.gif The red line is high magnitude shock, spread out over a short time. The orange line shows how dampener limits the maximum shock by spreading the energy out over more TIME. In case on the red line, in SE this time is pretty much instant since the voxels don't compress or anything like that. As the time duration approaches zero, the magnitude of energy spike becomes INFINITE. This is why the ability of a block to deform to absorb energy is such a huge deal. It's also why clang is so dangerous right now. I also think a "bumper" block would be amazing. Just something that resembles a giant button, designed to spread out impact force over time.
AutoMcD 11/09/17 18:14
I had another relevant thought. In a lot of situations where the ship touches the ground at speed, the ship should bounce and MOST of the energy continues forward. Currently it seems the same as hitting a wall. I'm sure any number of "shortcuts" could apply to prevent the calculations from becoming too heavy, such as using the average slope of a whole area of voxels or maybe applying the wheel physics. I'll leave it to you guys on that, but I do think it's key to apply the collision force perpendicular to the plane of contact. If the ground was flat and 95% of the momentum is forward, that's only 5% of the energy we should be dealing with, should theoretically be able to ride it out on friction with just the bottom of the ship roughed up a little. Even with wheels on flat land they just kinda pop off over a certain speed, and then when the frame touches it's a total loss.
Seabird 15/09/17 21:18
Perhaps for each point of crash damage taken by blocks on the grid, there should be a small about of braking thrust applied over the next 0.1 seconds. If that were not possible, at the very least can we have damage scale more linearly for low speed collisions, so that there isn't a point where suddenly your ship is destroyed for being 0.1m/s over the limit.
AutoMcD 16/09/17 00:07
Yeah, some approximation of the fact that deforming a huge cube of metal would take a lot of energy.. I'll let them sort out what method applies best to the game internals. I suspect it's more of an instant thing so maybe take it in smaller steps.. 10% deformation = subtract A amount of energy from velocity at B vector, and pound in C voxels at the point of impact. The missing voxels and altered trajectory create a little bubble of time until the next event, even if only milliseconds away. Slow moving ships do bounce off stuff somewhat realistically, I imagine most of what needs to be done is already there. It just needs adjusted?
KManALPHA 17/09/17 15:57
When the Oxygen generator gets dinged on my small ship a massive Ice chunk spawns [from resources inside], this can cause quite an issue with the ship structure. It also looks dumb. Maybe resources shouldn't always group/clump up?
andreykl 24/09/17 12:40
>When an ore ball spawns (usually because drill or cargo popped) it might as well be a warhead going off. +1, in one case I had a lot of containers on a "main ship", disconnected miner from it, miner slightly damaged an armor slope and whole "main ship" ship blows up. Technically in this case it wasn't only ores' flaw - armor deformed, somehow transferred impact to container and it destroyed container. So either armor should not deform non-deformable blocksor all blocks should deformable. >Any thruster or non-armor part that touches a voxel just dissappears as if it were a popped balloon of no mass at all. It is very easy to see in case of meteor showed. When the easiest one hits armor, it will form a dent at most, when it hits large engine, engine vanishes, even if said engine can soak more bullets then armor...
Radeth 28/09/17 11:26
I agree with the fact that voxels are too hard, not sure about metal armor being too weak (maybe just making the terrain softer would balance everything out) and I don't see it as a difficult modification to make engine side, probably just tweaking some values. The expected result of a crash landing shouldn't be a destroyed ship and a few deformation bumps along the trail, it should be a damaged ship with a long trench made by the chassis (think plane crashes). Although I'm not sure if making voxel easier to deform would have a negative impact on performance, as it would greatly increase the amount of voxels being deformed.
AutoMcD 28/09/17 14:46
Yeah, a crash landing plowing a trench would be a cool effect. Voxel performance a separate issue. I just don't want my ship to get REKT every time it bumps into something (or server lags).
amrylin1337 01/10/17 20:42
I agree that Ships feel too delicate at times. I watch some battles on Youtube though and wonder how the combatants manage to live so long....so there is a give and take. There are decent armor mechanics in SE but some stuff just seems a tad too thin perhaps.
AutoMcD 20/11/17 14:09
So, I have to say that things are greatly improved with this physics update. I accidentally left stone ejectors going on my miner when I docked to base that had stone, which usually results in the ejectors and the bottom of the ship being gutted. Not this time! And it was on dedicated server. When I undocked they sort of sprung off the ground all funny like they were loaded, but nothing broke! I have not tested if a popped drill or cargo spawned oreball blowing up the whole ship is still a thing yet. It wasn’t implicitly mentioned in the change log so I would still expect a lot of damage upon losing a full drill. The planetary collisions are greatly improved, the lack of phasing thru the voxel in itself gave the armor a lot more “meat”. But I still feel it is too weak of defense against voxels, the base numbers still need to be adjusted there. Keep up the good work, Keen!