JFSOCC wrote:Looks and sounds pretty cool. Those nebulas are amazing. I even turned off F.lux in order to appreciate them more
How's your workload been? Stress, anxiety to deliver, impulse to oversell how amazing it's all been?
I'm really glad that you seem to have got your groove back.
I've got a question that's a little outside of this update: I'm curious about ship design. You're designing all sorts of algorithms to procedurally generate an impressive amount of content; will the players be able to override/bypass generation algorithms in order to design their own ship/station aesthetic?
PS: I'm damn impressed by the planetary generation! If that is "bad" then I'm curious what good is going to look like.
Workload is manageable, especially with the 'low pressure' weekly updates. This is a nice groove indeed. I feel driven to get good work done, but not so stressed that I feel I need to make every week look like a new pot of gold at the end of another rainbow. If I can keep this balance up, I think we are in good shape
To the latter question: 1. From a modding perspective, absolutely. 2. From a vanilla game perspective, a ship/station editor is a planned feature so that players can override the PCG models. A simple such editor was shown during the KS campaign. A better one was in progress during the RTB week where I wiped out
Flatfingers wrote:... I'm still concerned that LuaJIT might not be able to handle everything, and it sounds like doing multi-system AI (plus smooth local graphics) performantly is still the acid test. I think you're being smart and responsible allocating a little time most weeks to Option Z just in case you do discover that LuaJIT can't cope. What I'm trying to communicate is the perspective that there is enormous value in finding out LuaJIT's practical limits NOW, because that news will be very upsetting. The sooner you rip that bandage off, the better for everyone it will be.
Or maybe it can be confirmed that LuaJIT can be persuaded to handle every reasonable thing you throw at it. Great! Again, the sooner you find that out, the better -- it will be a huge weight off your shoulders, and you can reallocate the time you felt you had to spend on Option Z development back into pure Limit Theory coding.
As am I, and I agree with all that you said. I'm trying to get to that place of being able to push LJ's limits. I have days where I'm very confident in the perf and that I'll be able to work around (using the practical Josh mindset) issues that do arise. I also have days where it still makes my stomach a little queezy to think that I'm writing tight loops in a dynamic language, JIT or no JIT. Only time will tell -- until then, I continue to emphasize doing things in the C core where possible as well as exploring Option Z (and ways in which Option Z could work together with LJ).
Flatfingers wrote:I say this with all seriousness: I could not be prouder of you.
It does mean a lot.