Return to “General”

Post

Re: Remaking Limit Theory - From the ground up

#77
man theres a lot of posts here (very encouraging btw :thumbup: ), forgive me if some of what I say is a bit of a repeat/parrot of some stuff other people have mentioned already.

I think that it would be important for the project to follow as much of the procedural vision of LT as possible, I see it as a bit of a trade off. The more procedural the game, the less custom assets/tweaking will be required, theres a reason why the big AAA games have teams of hundreds of employees building big worlds because everything needs to be hand crafted and balanced/tweaked accordingly which also makes it quite riggid (with less emergent possibilties). The flip side is that we'll need to create new tech to get the procedural stuff right and fit it in with the more conventional aspects which is itself quite the task as we all know from josh's blogs, the answers are not imediately obvious.

As for engine/framework/build source from scratch dilemma, im full team C/C++ (my skills are negligible and I have a long way to go before I could contribute meaningfully, so maybe a bit of a dunning kruger opinion here :lol: ) entirely because of how important I think it would be to get the procedural stuff right and accomplishing that with the sorts of scale we'd all like to see will (as far as im aware) require a solution that is incompatible (or at least extremely difficult) with prebuilt engines like unity or unreal however, IF using a pre built engine is something that most of us would like to do, I personally recommend using unreal or godot because its open source (and I have a fair amount of experience with unreal engine :D ) but I think we could end up having to write so much engine work around stuff, it could likely end up being ultimately easier if we just started something from scratch.

On the subject of licensing/open source etc. I think the best thing to do would be to do it though something like github and make it all completely open, with GNU or MIT licence or something, to not only insure longevity of the poject, but part of what I find so engaging about LT, is the technology was so interesting. If this was completely open, which would remove the money and business insentive out of the picture so any potential complications and drama related to that baggage down the line wont become a problem (i recently had a personal project of mine blow up over this crap, so im trying to be a more cautious about that in future).

Also, keeping everything open like that, would allow people who would be interested in simulation research, the opportunity to use the project to experiment with a what essentually could be the digital equivalent of studying an ant colony. If python scripting for example was implemented in a particular way, then im sure someone would like to try to plugin tensorflow at some point and see what they could do with it. Im not suggesting that we should be building a game with the intention that it uses the most complicated and over engineered cutting edge whatever, but if money is never the goal, then I can only see benefits to keeping things as open as possible.

last thought: vulkan recently announced adding some kind of ray tracing support (not nividia RTX specific either), so potentially more incredible lighting? :think:
Post

Re: Remaking Limit Theory - From the ground up

#78
Talvieno wrote:
Mon Mar 30, 2020 7:50 pm
Haron, read through your post (quite excellent and informative)! I agree with most of what you say, and it mostly makes sense, but this part stood out to me in particular regarding Rust:
Haron wrote:
Mon Mar 30, 2020 3:35 pm
Cons:
- Relatively hard language to start with. Learning curve is pretty steep for the most of developers.
- Not so much developers knowing Rust at the moment.
In my opinion this is a very, very strong argument that more or less precludes Rust from being usable for this project, as there would be very few developers actually working on it, which means slow progress and a great loss of interest from most individuals involved.
I completely understand this. And it also a sad situation: when people choose between well known but old tool and better but newer one they usually choose former and have to deal later with all issues introduced by it. :(
But as I said I understand your point of view.
Haron wrote:
Mon Mar 30, 2020 3:35 pm
RE: Custom engine: If Josh couldn't finish his engine in four years despite it being a full-time passion project he got paid for, I don't think we're going to manage to write our own engine as a ramshackle group of coders working on this as a voluntary hobby project in our spare time. Harsh to put it like that, I suppose, but we don't want to fall into the same trap Josh did. Whatever we decide on using must either:
  1. Be very easy to learn
  2. Already have a good number of people here that know how to use it
  3. Both of the above
Completely agree with you. There is no need to start with the custom engine. But still it should be possible to switch engine later if selected one won't satisfy project requirements.
So as the option it can be Godot engine and C++ as main language.

By the way I just got an idea to use modules compiled to Wasm (WebAssembly)? This way it will be cross platform and it has pretty fast execution speed when jitted. And most of top programming languages can be compiled in Wasm nowadays.
Post

Re: Remaking Limit Theory - From the ground up

#79
Fingus wrote:
Tue Mar 31, 2020 4:58 am
As for engine/framework/build source from scratch dilemma, im full team C/C++ (my skills are negligible and I have a long way to go before I could contribute meaningfully, so maybe a bit of a dunning kruger opinion here :lol: ) entirely because of how important I think it would be to get the procedural stuff right and accomplishing that with the sorts of scale we'd all like to see will (as far as im aware) require a solution that is incompatible (or at least extremely difficult) with prebuilt engines like unity or unreal however, IF using a pre built engine is something that most of us would like to do, I personally recommend using unreal or godot because its open source (and I have a fair amount of experience with unreal engine :D ) but I think we could end up having to write so much engine work around stuff, it could likely end up being ultimately easier if we just started something from scratch.
I know from personal experience that procedural mesh generation is entirely possible, and relatively easy (as easy as proc-mesh generation gets, anyway) within Unity. Aside from performance, I don't see any particularly glaring issues with Unity in regards to this project. Like I said, I don't have any real experience with Godot, so I can't speak for it, but if it's too bare bones, it might also mean that we have to do a modification ourselves to meet our goals.

There's also as Tal said the issue of the community's familiarity with the engine, and language. I'm native C# and a relatively young programmer with little other experience. I've also only worked in Unity game engine wise. I could certainly learn Godot and C++, and I wouldn't be averse to it, but it's worth noting the time sink and potential loss of contribution. If we can get away with using Unity, it'd probably be the best option, but again: performance is the question.

Also because if I'm going to contribute, I'd be balancing this and my actual game. Adding C++ on top of that means I'm not certain how I'd fare, not for lack of trying I'm sure.
<Detritus> I went up to my mom and said "hey... do you feel like giving five dollars to black lives matter?" and she laughed and said no :v <Black--Snow> my life does matter though ~~ added by Hema on Jun 11 2020 (2770)
Post

Re: Remaking Limit Theory - From the ground up

#83
insta wrote:
Tue Mar 31, 2020 10:00 am
So a question turned up in the general brain storming doc.

Do we consider warp lanes/trade lanes a core feature that should be in? I'd say yes but I wouldn't mind some other in-system fast travel mechanism (as long as the systems are large enough to rquire one :crazy: ).
I would love to have trade lanes, or warp lanes, or accelerators, or sling shots , or ether with lower density ...
Post

Re: Remaking Limit Theory - From the ground up

#84
insta wrote:
Tue Mar 31, 2020 10:00 am
So a question turned up in the general brain storming doc.

Do we consider warp lanes/trade lanes a core feature that should be in? I'd say yes but I wouldn't mind some other in-system fast travel mechanism (as long as the systems are large enough to rquire one :crazy: ).
I see no reason not to. Features don't have to be a release completeness to be implemented. Unless your question is "Should it be replaced", which is a more difficult one to answer. :V
<Detritus> I went up to my mom and said "hey... do you feel like giving five dollars to black lives matter?" and she laughed and said no :v <Black--Snow> my life does matter though ~~ added by Hema on Jun 11 2020 (2770)
Post

Re: Remaking Limit Theory - From the ground up

#86
My hope/desire is that resulting game will be configurable enough via modding that a lot of wish list features can be added by the community. Not having made a mod-able game before, I don't know how easy it would be to expose internal functions via an API. For example, something clever enough to import /create a system gate, rotate it, add a glowing interior, link it to another system gate, manage which is the out-going and in-coming sides, and allow transits. Or, allow the creation of an alien vessel from a different set of parts rather than the default .
Post

Re: Remaking Limit Theory - From the ground up

#87
zircher wrote:
Tue Mar 31, 2020 3:27 pm
My hope/desire is that resulting game will be configurable enough via modding that a lot of wish list features can be added by the community. Not having made a mod-able game before, I don't know how easy it would be to expose internal functions via an API. For example, something clever enough to import /create a system gate, rotate it, add a glowing interior, link it to another system gate, manage which is the out-going and in-coming sides, and allow transits. Or, allow the creation of an alien vessel from a different set of parts rather than the default .
As someone that has made moddable systems (as part of a gameplay mechanic) it's as simple as keeping content in editable files in the game folders. More or less: as long as you don't hide things in the code, you've already succeeded.

Making that properly performant is a different story though.
Have a question? Send me a PM! || I have a Patreon page up for REKT now! || People talking in IRC over the past two hours: Image
Image
Image
Post

Re: Remaking Limit Theory - From the ground up

#88
Talvieno wrote:
Tue Mar 31, 2020 3:17 pm
I would certainly consider them a core feature, but I know some people disagree.
I would agree.
Josh's reasoning was so we could enjoy the view of space and the AI in the "living" world
And being able to ride on a rail, and view the system et al, is a lovely way to spend some time, whilst figuring out the next chapter in my space opera :)
YAY PYTHON \o/

In Josh We Trust
-=326.3827=-
Post

Re: Remaking Limit Theory - From the ground up

#89
FormalMoss wrote:
Tue Mar 31, 2020 3:42 pm
I would agree.
Josh's reasoning was so we could enjoy the view of space and the AI in the "living" world
And being able to ride on a rail, and view the system et al, is a lovely way to spend some time, whilst figuring out the next chapter in my space opera :)
What if the AI constructed them? Lay out a few on startup, and build out a network as you go. Maybe specialized exploration ships can lay out their own rails or something?
Post

Re: Remaking Limit Theory - From the ground up

#90
Bitrage wrote:
Tue Mar 31, 2020 3:43 pm
FormalMoss wrote:
Tue Mar 31, 2020 3:42 pm
I would agree.
Josh's reasoning was so we could enjoy the view of space and the AI in the "living" world
And being able to ride on a rail, and view the system et al, is a lovely way to spend some time, whilst figuring out the next chapter in my space opera :)
What if the AI constructed them? Lay out a few on startup, and build out a network as you go. Maybe specialized exploration ships can lay out their own rails or something?
OOh, lovely idea.. but the Ai would be limited by the available mines in the current solar system.. unless one of them got all sneaky, and discovered - dun dun DUN.. a WORMHOLE!
Full of exotic ore
YAY PYTHON \o/

In Josh We Trust
-=326.3827=-

Online Now

Users browsing this forum: No registered users and 1 guest

cron