Return to “Dev Logs”

Post

Re: [Lindsey] Friday, December 15, 2017

#47
Flatfingers wrote:
Sun Dec 17, 2017 3:10 am
Wowzers. To clarify: are these being generated by our old friend, Mr. PRNG? Or are they hand-designed for now?

If procedural, that's fantastic -- it implies that the universe of ship types we encounter in LT v1.0 will be terrifically diverse.

I do have two pieces of feedback at a conceptual level, both of which are related to kostuek's comment: wings make sense for atmospheric flight.

1. Forward-facing wing surfaces that are flat are an offense to physics. :D If it really is a "wing," implying that this thing could fly well in an atmosphere, then there probably needs to be some logic to insure that the "front" part of any wing is an edge where two surfaces join.

......
Spoiler:      SHOW
Image
"A sufficiently advanced technology is indistinguishable from magic."
- Arthur C. Clarke
Post

Re: [Lindsey] Friday, December 15, 2017

#48
lobosan wrote:
Sun Dec 17, 2017 11:53 am
Could there be distinct pallets of ship design algorithms to represent different "space cultures" (with different values/goals/philosophies expressed in the design)?
It may make recognizing ship classes through visual shorthands more difficult though ...
That's the plan. Ship classes will have their own unique algorithms, so style shouldn't interrupt their visual distinction. Size is going to be a big factor in ship class.

nobb wrote:
Sun Dec 17, 2017 4:07 pm
That super impressive! Especially if you didn't cheery pick the result! Would you consider generating a video where you change the parameter at random and generate a new ship like every 10 second ? That a pretty good way to visualize the breadth of the possibility space of a generative algorithm, and it would allow to see if there is any bias repetition or something not explored. Also it should be pretty low effort to generate automatically, and of course it would be super cool!
I'm not going to do the streaming idea as Hyperion proposed, at least not until we're trying to generate release-level hype. But it's certainly a cool idea to keep in my back pocket. A video or gif might be fun for a devlog in the near future. But I'd rather spend more time writing the algorithms than making fancy ways to present them, so it totally depends on how much time that consumes.

Talvieno wrote:
Sun Dec 17, 2017 5:43 pm
Nope, they cannot morph as of now... and I really don't think that's at all likely to be a feature Lindsey implements, either. :P
+1, for the reason stated above.
Ship Inspiration Pinterest!! (send me stuff)

"You’ve got to work on something dangerous. You have to work on something that makes you uncertain. Something that makes you doubt yourself... because it stimulates you to do things you haven’t done before. The whole thing is if you know where you’re going, you’ve gone, as the poet says. And that’s death."
- Stephen Sondheim
Post

Re: [Lindsey] Friday, December 15, 2017

#50
Some questions

How well can the engine handle lots of visual diversity on the screen at different times? Will 1000 unique snowflakes on screen will be far slower than 1000 identical ships?
Are you planning on making a Damage, Debris, & Deformation function in the algorithms that is applied when the pew pew makes the boom boom?
Can ships connect with each other and break away from each other? Like so:
Spoiler:      SHOW
Image
I think it would be pretty cool if some races' "carrier" was just a bunch of little ships snapped together and a big ass engine and repair bay attached at the back, and the ships just break off into smaller segments as needed, and where a heavy fighter is just 2 fighters stuck together :p
Image
Challenging your assumptions is good for your health, good for your business, and good for your future. Stay skeptical but never undervalue the importance of a new and unfamiliar perspective.
Imagination Fertilizer
Post

Re: [Lindsey] Friday, December 15, 2017

#51
How well can the engine handle lots of visual diversity on the screen at different times? Will 1000 unique snowflakes on screen will be far slower than 1000 identical ships?
That's a question for Josh. Logically speaking, 1000 unique ships would require more resources; both 1000 unique ships and 1000 identical ships ought to take similar render times, but if anything is faster it would be the identical ships. From what Josh has said previously, I'm guessing probably not by much.
Are you planning on making a Damage, Debris, & Deformation function in the algorithms that is applied when the pew pew makes the boom boom?
I don't think there are plans for dynamic soft-body damage & deformation systems. That's outside the scope of LT. The texture may change based on how much damage a ship has taken. When a ship is destroyed, the model will probably change somewhat, but not dynamically. Ideally it would fall to pieces; I think the LT engine is currently set up well for that, although the detail on the pieces/how well they relate to the original ship/etc may be minimal.
Can ships connect with each other and break away from each other? Like so:
This is not in the scope of LT. :P The current shipgen algorithms no longer use CSG. I don't think Josh has plans to re-implement a CSG algorithm.
I think it would be pretty cool if some races' "carrier" was just a bunch of little ships snapped together and a big ass engine and repair bay attached at the back, and the ships just break off into smaller segments as needed, and where a heavy fighter is just 2 fighters stuck together :p
This is not going to happen unless plans drastically change. :P It will probably be moddable to an extent, though. Carrier takeoffs and landings will probably not be "actual" landings either; rather, you will "enter" the carrier, the screen will fade out, and then fade back in showing your ship landed. Also keep in mind that carriers were a stretch goal, so they'll be added post-release.
Have a question? Send me a PM! || People talking in IRC over the past two hours: Image
Image
Image
Post

Re: [Lindsey] Friday, December 15, 2017

#52
Thanks to batching, (might be using the wrong term here) older systems would be faster with 1000 identical objects. Thanks to modern hardware, 1000 unique objects that all fit in memory will be perfectly okay.
Provided we have a fair bit of texture reuse, we shouldn't even come close to capping out a low-middle end modern laptop video card's memory (1GB).

Mesh damage is hefty, so is procedural morphing of a mesh while live. If the mesh is static there are optimizations that can be applied, and you don't need to recalculate collision data.
Adding damage layers to textures however, is much faster. Still not a trivial task to implement properly, but I can see that being done. With damage effects like fires, sparks, smoke, it should look good.
(And yes, I understand physics doesn't allow fire or smoke in space, but for pretties, they fit well)

Connecting ships is trivial provided a hierarchical model of the view-space. So if it isn't in the base game, expect it in mods. I would love to see carriers with visible and uncovered docking spaces for fighters. Mostly so that we could build an RTS inside of LT. :V (Homeworld Theory!)
Last edited by Silverware on Mon Dec 18, 2017 2:21 pm, edited 1 time in total.
°˖◝(ಠ‸ಠ)◜˖°
Toba - A Development Dump
Post

Re: [Lindsey] Friday, December 15, 2017

#53
Distant wrote:
Mon Dec 18, 2017 9:30 am
Here's my Video Dev Log. Enjoy :)

https://youtu.be/QKcK0DtCB3s
YAY!! I created a YouTube account just to answer comments lol


Thank you Talv for answering questions! +1 to his answers.
Ship Inspiration Pinterest!! (send me stuff)

"You’ve got to work on something dangerous. You have to work on something that makes you uncertain. Something that makes you doubt yourself... because it stimulates you to do things you haven’t done before. The whole thing is if you know where you’re going, you’ve gone, as the poet says. And that’s death."
- Stephen Sondheim
Post

Re: [Lindsey] Friday, December 15, 2017

#54
Today's ships 12.18.17
Added more curve warps, more surface warps, and a new hull shape.

No more devlogs this year, so be sure to keep up with this thread for daily screenies. Also, would like to make a correction on when the next devlog is. My Dec 1 devlog said the first post of the new year will be on January 12. I have decided instead to post on January 6th. Yay, earlier devlog! I also corrected this on the original post.
Ship Inspiration Pinterest!! (send me stuff)

"You’ve got to work on something dangerous. You have to work on something that makes you uncertain. Something that makes you doubt yourself... because it stimulates you to do things you haven’t done before. The whole thing is if you know where you’re going, you’ve gone, as the poet says. And that’s death."
- Stephen Sondheim
Post

Re: [Lindsey] Friday, December 15, 2017

#56
Keep in mind that designing beautiful asymmetrical ships - procedural algorithms aside - is a lot harder than designing symmetrical ones. I don't think asymmetrical ships are currently planned, but I wouldn't rule them out completely.
edit: This is confirmed. Asymmetrical ships are not planned.


Great ships, Lindsey! They're taking shape rapidly! :thumbup:
Have a question? Send me a PM! || People talking in IRC over the past two hours: Image
Image
Image
Post

Re: [Lindsey] Friday, December 15, 2017

#57
Hey just wanted to contribute with something. Some people say that fighters in LT don't need to be streamlined because they are flying in space. I used to think that as well, but recently i read the "star carrier" series by Ian Douglas, and in it he made a few interesting points about space fighter shapes. In those books fighters accelerate to near C speeds, at these speeds the fact that space is a near, and not total, vacuum makes a lot of difference. That happens because at relativistic speeds even scattered atoms of hydrogen can create a lot of resistance, so streamlined designs, like atmospheric fighters, do make sense. And from the videos of LT I saw, the time it took to travel the distances between bodies, if scaled to real star systems, the ships would have to be moving at those speeds.
Build a man a fire, and he'll be warm for a day. Set a man on fire, and he'll be warm for the rest of his life.
Post

Re: [Lindsey] Friday, December 15, 2017

#60
N810, I'm familiar with the swept-forward concept. It's pretty cool; it's also super-dependent on computer software... which, as a one-time programmer, makes me a little nervous. :D

But that's not what I meant when I was talking about atmospheric craft needing wings whose leading elements aren't flat surfaces.

Here's a good example of what caught my eye:


Image


Theoretically, you could make something like this fly, just as you could make a brick fly... but in either case you're making the engines work harder than they have to for atmospheric flight because a flat leading surface interferes with aerodynamic lift.

Again: for spacecraft, I will quote the guy at the end of Buckaroo Banzai: "So what? Big deal." For a spaceship in normal (i.e., significantly slower-than-light-speed) flight, you could make a ship that's a wing, or a cube, or anything in between -- its "flight" characteristics will depend on how its engines can move its mass distribution.

But if you're going to design a ship that looks like it could be efficient in atmospheric flight, you wouldn't undercut your design by imposing a low-efficiency wing design. And a flat leading edge is low-efficiency.

"Edge" is sort of a misnomer, actually; a wing doesn't have to come to a sharp leading or trailing edge. Because most aircraft are meant to operate at subsonic speeds, their wings (airfoils is the technical term) use a rounded shape. True sharp leading edges are low-drag but tend to significantly increase the risk of stalling, so they're typically only seen on aircraft that are meant to fly in the supersonic regime.

In general, then, what I'm suggesting is that if it's a flying thing that has bits that look like wings, the front of those wings shouldn't be flat. How many polys are required to achieve that look, I leave to the graphics experts. ;)

lmaluko wrote:
Mon Dec 18, 2017 9:01 pm
Hey just wanted to contribute with something. Some people say that fighters in LT don't need to be streamlined because they are flying in space. I used to think that as well, but recently i read the "star carrier" series by Ian Douglas, and in it he made a few interesting points about space fighter shapes. In those books fighters accelerate to near C speeds, at these speeds the fact that space is a near, and not total, vacuum makes a lot of difference. That happens because at relativistic speeds even scattered atoms of hydrogen can create a lot of resistance, so streamlined designs, like atmospheric fighters, do make sense. And from the videos of LT I saw, the time it took to travel the distances between bodies, if scaled to real star systems, the ships would have to be moving at those speeds.

This is an interesting notion. With due respect to Douglas, it seems to me that if you're flying at only a very small fraction of c, you don't need a specialized shape to minimize the damage caused by hitting gaseous particles.

On the other end, at near-relativistic speeds, it seems unlikely that any shape is safe -- you'd require some kind of projected shield that prevents any kind of matter from slamming into your ship's hull. So if all shapes are equally unsafe, again, you might as well use any shape you want. (And then there's the problem of controlling a ship's flight when it's moving faster than any brain, organic or electronic, can think, requiring probabilistic navigation planning... but that's a separate thread. :) C.J. Cherryh's books do a good job of describing what this might be like.)

So on balance, I'm still in the "ack, no flat forward wing surfaces!" camp.

Online Now

Users browsing this forum: No registered users and 1 guest

cron