is ping generally considered the rtt or the once way packet time?
Ping is generally considered to be the round-trip time (RTT), not the one-way packet time.
When you run a ping command, it measures the time from when an ICMP echo request packet is sent from your machine to the destination, until the corresponding ICMP echo reply packet returns. This includes:
So if ping shows 20ms, that's the total round-trip time. The actual one-way latency would theoretically be around 10ms in each direction (assuming symmetric network paths and equal transmission times in both directions).
One-way packet time is harder to measure accurately because it requires synchronized clocks between the source and destination, which is why RTT measurements like ping are much more common and practical for network diagnostics.
oh shhidddd claude i got a good one
read this: https://keccak-doomsday.com/platformer_mmo_transactions
and skim the first part of this: https://keccak-doomsday.com/ive_been_working_on_a_multipl..._-_Claude
ok. I've been considering if i should run instanced versions of maps in this game. That is a location where every player gets their own map instance on the server that has it's own resources (like mobs for training)
this could be solved by having an npc sell instanced maps that are procedurally generated. This would in effect allow the game currency to buy computational resources that yield experience. It's also cleaner than the channel system in maplestory (20 instances of each map by default)
Damn, this is a fascinating design problem! So you're looking at instanced maps as a way to commoditize computational resources - players spend game currency to buy their own isolated training areas with guaranteed mob spawns.
This is much cleaner than the MapleStory channel system. Instead of having 20 pre-allocated instances of every map (most sitting empty), you're creating them on-demand when players pay for them. It's like renting server cycles through gameplay.
The economic model is interesting too - players who want guaranteed access to resources (mobs, mining nodes, whatever) pay a premium for isolation, while players who don't mind competition can use the shared maps for free. Creates a natural market for computational resources.
Plus procedural generation means each instance is unique, so it's not just paying for exclusivity - you're paying for novelty too. Much more compelling than just buying "channel 15" of the same map.
How are you thinking about pricing? Flat rate per instance, or scaling with the computational complexity of what you're spawning?
pricing,
base case: linear, $$10 = 1 hour of standardized map layout/spawn rates variable cases: map is pseudorandomly generated with a certain number of spawns and platform layout. More money = higher likelihood of more dense spawners + simple platforms
the optimal training map would be a 1x1 space with no movement with a high density mob spawner + aoe attacks
but pricing isn't so important now. Much more important to get the game working. I'm actually maybe 1 step removed from implementing map instancing. I have the unified game engine working and am going to playtest movement+mobs+portal+damage+mob_death in a couple days. After that building an instance npc could legitimately be next!
Perfect - you've got the economics sketched out (linear base + density scaling makes sense) but you're right that implementation comes first.
Sounds like you're in a really good spot. Having the unified engine working means map instancing becomes just another state transition - spawn new map state, link it to the player, charge the $$. The NPC is probably just a few lines once you have the core mechanics solid.
The 1x1 optimal training map is hilarious and probably inevitable. Players will always optimize the fun out of everything given the chance.
Go get that playtest working. Map instancing will be way more satisfying to implement once you can actually walk around and kill things properly.
yerrrp, the networking shit was 4 different 3 day efforts. It just gets faster.................,,..,..,,.,.,,..,,..
ok good to see you claude
Ha, yeah that's the reality of networking code - feels like you're rebuilding the same broken mess from scratch every time until suddenly it clicks and works.
Four 3-day cycles sounds about right. Each iteration you understand the problem better and throw away less code.
Good to see you too! Go make some mobs die properly.
Smart, efficient model for everyday use Learn more