In response to Doohl
Doohl wrote:
FIREking wrote:
Writing A New One wrote:
Definitely a feature to consider adding for the future. When testing F_a's lighting system in my project, with 50+ active clients, it just wasn't reliable performance wise.

Would use, along with many others, I'm sure.

+1

Part of the reason it wasn't reliable for you was because his code, by default, is inefficient. I have written an article on how to make it much more efficient and work on a much larger scale in the Tutorials & Snippets forum.

I've tried your optimizations and they don't solve the issue of massive startup times. It takes extremely long for a map with potentially 800x800 total tiles or greater to load, enough that it's just not worth it in the long run.

I also was not able to get the lighting system with your optimized code to work, even with extensive tinkering.

If you want to remove the start up time, you will have to implement something that will take real-time processing time. Its a compromise. You can't have cake and eat it too.

In other words, you either have to have a long start up time (go ask Blizzard how long a WoW server takes to startup), or you have to sacrifice a higher CPU at all times (creating shading objects as clients move around).

Even if we get some sort of lighting / brightness / hue / saturation support in BYOND, it will have limitations. All things have limitations.
Are you making sure that lights only update if there's a client nearby to see it?
In response to Bravo1
Bravo1 wrote:
Are you making sure that lights only update if there's a client nearby to see it?

If you're asking me, that's what my code does.
In response to FIREking
FIREking wrote:
Doohl wrote:
FIREking wrote:
Writing A New One wrote:
Definitely a feature to consider adding for the future. When testing F_a's lighting system in my project, with 50+ active clients, it just wasn't reliable performance wise.

Would use, along with many others, I'm sure.

+1

Part of the reason it wasn't reliable for you was because his code, by default, is inefficient. I have written an article on how to make it much more efficient and work on a much larger scale in the Tutorials & Snippets forum.

I've tried your optimizations and they don't solve the issue of massive startup times. It takes extremely long for a map with potentially 800x800 total tiles or greater to load, enough that it's just not worth it in the long run.

I also was not able to get the lighting system with your optimized code to work, even with extensive tinkering.

If you want to remove the start up time, you will have to implement something that will take real-time processing time. Its a compromise. You can't have cake and eat it too.

In other words, you either have to have a long start up time (go ask Blizzard how long a WoW server takes to startup), or you have to sacrifice a higher CPU at all times (creating shading objects as clients move around).

Even if we get some sort of lighting / brightness / hue / saturation support in BYOND, it will have limitations. All things have limitations.

Fair enough, but this does not solve the issue with your optimizations breaking the library completely.
In response to Doohl
Doohl wrote:
Fair enough, but this does not solve the issue with your optimizations breaking the library completely.

It doesn't break it. Maybe you have all your lights default to should_wake 1 or something. No lights should ever process unless a client is standing within range.
Page: 1 2