- Completely removed both `log` and `pretty_env_logger` and replaced
with `tracing` and `tracing_subscriber` where necessary.
- Converted all `log::info!(...)` et al. statements to just use the
shorthand macro i.e. `info!`. This was mostly to make renaming easier.
- authc no longer uses reqwest
- image only supports PNG
- replace routille with tiny_http
- several other dependencies
- cargo upgrade
- following improvement was measured on R7 1700X:
before:
- cargo build: 3076.73s user / 4:45 total / 589 dependencies
- cargo test: 6118.38s user / 7:30 total / 959 dependencies
after:
- cargo build: 2680.54s user / 4:05 total / 480 dependencies
- cargo test: 5351.81s user / 7:04 total / 791 dependencies
- added xMAC94x to CODEOWNERS for Cargo.toml, he will protect them from now on and hit people with evil looks ;)
For anything in worldgen where you use a HashMap, *please* think
carefully about which hasher you are going to use! This is
especially true if (for some reason) you are depending on hashmap
iteration order remaining stable for some aspect of worldgen.
Turned a lot of for loops into for_each loops, which should be easier
for LLVM to optimize currently. Also updated almost all the non-erosion
stuff in WorldGen to run in parallel (and take advantage of the cache,
in the case of TownGen), and hopefully improved performance somewhat for
chunk generation as well.
- soil production (currently disabled).
- debris flow erosion (combined with regular stream power law).
- flow computation using multiple receivers.
- filling strategy during drainage network calculations.
Also tweaks a variety of other aspects of erosion.
Currently we only do this when no players are in range of the chunk. We
also send the first client who posted the chunk a message indicating
that it's canceled, the hope being that this will be a performance win
in single player mode since you don't have to wait three seconds to
realize that the server won't generate the chunk for you.
We now check an atomic flag for every column sample in a chunk. We
could probably do this less frequently, but since it's a relaxed load it
has essentially no performance impact on Intel architectures.
