Update documentation and remove some unused code

CHANGELOG.md

@@ -8,6 +8,7 @@ All notable changes to this project will be documented in this file.
 ### Removed
 
 * GitHub sponsors functionality
+* experimetnal `async-evolution` feature
 
 ### Fixed
 

docs/src/concepts/pragmatic/problem/objectives.md

@@ -8,8 +8,9 @@ of multi objective.
 ## Understanding multi objective structure
 
 A multi objective is defined by `objectives` property which has array of array type and defines some kind of "hierarchical"
-objective function where priority of objectives decreases from first to the last element of outer array. Objectives inside
-the same inner array have the same priority.
+objective function. Here, priority of objectives decreases from first to the last element of outer array. So, these objectives
+are optimized in lexicographical order. Objectives inside the same inner array have the same priority and are processed
+considering Pareto optimality.
 
 
 ## Available objectives
@@ -42,7 +43,9 @@ constraints such as time windows. The objective has the following optional param
 * `minimize-tours`: minimizes total amount of tours present in solution
 * `maximize-tours`: maximizes total amount of tours present in solution
 * `minimize-arrival-time`: prefers solutions where work is finished earlier
-* `fast-service`: prefers solutions when jobs are served early in tours
+* `fast-service`: prefers solutions when jobs are served early in tours. Optional parameter:
+  *  `tolerance`: an objective tolerance specifies how different objective values have to be to consider them different.
+      Relative distance metric is used.
 
 ### Job distribution objectives
 
@@ -74,9 +77,9 @@ There are four work balance objectives available:
 Each objective has optional parameters defined by `option` property:
 * `threshold`: a target coefficient of variation (scale invariant statistical measure of dispersion) value which specifies
 desired minimum balancing level. All values below threshold are considered equal which helps the search algorithm to
-optimize conflicting objectives.
+optimize conflicting objectives. This similar to ε-Constraint Method optimization (in some degree).
 
-It is recommended to set both option values to guide the search towards optimum for conflicting objectives, e.g. cost
+It is recommended to set `threshold` option to guide the search towards optimum for conflicting objectives, e.g. cost
 minimization and any of work balance.
 
 An usage example:

rosomaxa/Cargo.toml

@@ -11,10 +11,6 @@ homepage.workspace = true
 repository.workspace = true
 edition.workspace = true
 
-[features]
-default = []
-# an experimental asynchronous evolution targeting better performance on multiple CPUs
-async-evolution = ["tokio"]
 
 [dependencies]
 rand.workspace = true
@@ -24,7 +20,6 @@ rand_distr = "0.4.3"
 
 [target.'cfg(not(target_arch = "wasm32"))'.dependencies]
 num_cpus = "1.16.0"
-tokio = { version= "1.37.0", features=["sync", "rt"], optional = true }
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 getrandom = { version = "0.2.15", features = ["js"] }

rosomaxa/src/evolution/strategies/mod.rs

@@ -2,11 +2,6 @@
 
 use super::*;
 
-#[cfg(feature = "async-evolution")]
-mod async_iterative;
-#[cfg(feature = "async-evolution")]
-pub use self::async_iterative::{AsyncIterative, AsyncParams};
-
 mod iterative;
 pub use self::iterative::Iterative;
 

vrp-cli/Cargo.toml

@@ -14,7 +14,6 @@ edition.workspace = true
 [features]
 default = ["vrp-core", "csv-format", "scientific-format"]
 
-async-evolution = ["vrp-core/async-evolution"]
 csv-format = ["csv"]
 scientific-format = ["vrp-scientific"]
 py_bindings = ["dep:pyo3"]