Skip to main content

Develop services

Prerequisites​

This doc assumes that you already have a fluence project, if not, fluence init and follow the prompts.

What is service and why is it needed​

A Marine service is a bunch of wasm modules linked together by a config. It is the computation part of the network. When deployed to a Peer, a service is accessible from Aqua.

Basic example​

To create a sample service, use fluence service new, like that:

bash
$ fluence service new services/some_service --name some_service
bash
$ fluence service new services/some_service --name some_service

This will create a service structure at services/some_service:

bash
services/some_service
├── modules              <- modules for this service
│   └── some_service
│       ├── Cargo.toml   <- rust crate config
│       ├── module.yaml  <- module configuration file
│       └── src
│           └── main.rs  <- rust source of main module of the service
└── service.yaml         <- service configuration file
bash
services/some_service
├── modules              <- modules for this service
│   └── some_service
│       ├── Cargo.toml   <- rust crate config
│       ├── module.yaml  <- module configuration file
│       └── src
│           └── main.rs  <- rust source of main module of the service
└── service.yaml         <- service configuration file

The default source, main.rs, is pretty much a hello world:

rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();                       // some required stuff, just keep it here
pub fn main() {}                          // will be called one module is started
#[marine]                                 // macro makes function appear in this module interface
pub fn greeting(name: String) -> String {
    format!("Hi, {}", name)
}
rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();                       // some required stuff, just keep it here
pub fn main() {}                          // will be called one module is started
#[marine]                                 // macro makes function appear in this module interface
pub fn greeting(name: String) -> String {
    format!("Hi, {}", name)
}

This service some_service now consists of only one module. You can run this service locally in a repl to test. Deployment to the network will be covered by other doc pages (link).

Repl testing​

Run repl using fluence CLI. It accepts service name or path to the directory with service.yaml

sh
$ fluence service repl some_service
sh
$ fluence service repl some_service

You will see some hints and setup information, as well as prompt to execute a command:

bash
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = 3b0aaf21-a499-4092-b2ef-890620b285c5
elapsed time 71.005084ms
1> 
bash
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = 3b0aaf21-a499-4092-b2ef-890620b285c5
elapsed time 71.005084ms
1>

The most useful commands are help, call, interface, and the commands have shortcats. Here is the output of the interface command:

bash
1> i
Loaded modules interface:
exported data types (combined from all modules):
<no exported data types>
exported functions:
some_service:
  func greeting(name: string) -> string
bash
1> i
Loaded modules interface:
exported data types (combined from all modules):
<no exported data types>
exported functions:
some_service:
  func greeting(name: string) -> string

The output says that module has no non-default data structures in interface, and has a greeting function, exported from some_service module that takes a string as an argument and returns a string.

Using information about interface it can be called:

bash
2> c some_service greeting "World"
result: "Hi, World"
 elapsed time: 5.047333ms
bash
2> c some_service greeting "World"
result: "Hi, World"
 elapsed time: 5.047333ms

That is the simplest way to execute the service and look if it works. To exit repl press ctrl + C.

More complicated examples​

After basic example works its time to go try more features:

  • structs in interfaces
  • mounted binaries and filesystem access
  • multi-module services

Structs in interfaces​

#[marine] functions can have numbers, strings, vectors and custom structs in signatures. To be used in signature, a structure must:

  • be marked with #[marine] macro
  • have only numbers, strings, vectors or other #[marine] structs as fields

Lets add a structure and a new function in our service:

rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub struct SomeStruct {
    number: i32,
    string: String,
    vector: Vec<i32>
}
#[marine]
pub fn process_struct(arg: SomeStruct) -> SomeStruct {
    SomeStruct {
        number: arg.number + 1,
        string: format!("{} {}", arg.string, 1),
        vector: vec![arg.vector[0]; 2],
    }
}
#[marine]
pub fn greeting(name: String) -> String {
    format!("Hi, {}", name)
}
rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub struct SomeStruct {
    number: i32,
    string: String,
    vector: Vec<i32>
}
#[marine]
pub fn process_struct(arg: SomeStruct) -> SomeStruct {
    SomeStruct {
        number: arg.number + 1,
        string: format!("{} {}", arg.string, 1),
        vector: vec![arg.vector[0]; 2],
    }
}
#[marine]
pub fn greeting(name: String) -> String {
    format!("Hi, {}", name)
}

And then use fluence service repl some_service again to run it. Then, interface command will show how interface changed:

bash
1> i
Loaded modules interface:
exported data types (combined from all modules):
data SomeStruct:
  number: i32
  string: string
  vector: []i32
exported functions:
some_service:
  func process_struct(arg: SomeStruct) -> SomeStruct
  func greeting(name: string) -> string 
bash
1> i
Loaded modules interface:
exported data types (combined from all modules):
data SomeStruct:
  number: i32
  string: string
  vector: []i32
exported functions:
some_service:
  func process_struct(arg: SomeStruct) -> SomeStruct
  func greeting(name: string) -> string

The structure added as data SomeStruct, and process_struct appeared in the exported function section. Lets call the function then. But how to represent the structure? The answer is as JSON. Repl interprets function arguments as JSON, so it must be a single json value (string, number, array or object). The top-level array or object is always interpreted as a container for function arguments, to handle functions with more than one argument. So, if function has only one argument and it is an array or an object, it is required to wrap it with []. So, here is the call:

bash
2> c some_service process_struct [{"number": 1, "string": "test", "vector": [4]}]
result: {
  "number": 2,
  "string": "test 1",
  "vector": [
    4,
    4
  ]
}
 elapsed time: 10.09025ms
bash
2> c some_service process_struct [{"number": 1, "string": "test", "vector": [4]}]
result: {
  "number": 2,
  "string": "test 1",
  "vector": [
    4,
    4
  ]
}
 elapsed time: 10.09025ms

Also, an array can be used instead of object:

bash
3> c some_service process_struct [[1,"test", [4]]]
result: {
  "number": 2,
  "string": "test 1",
  "vector": [
    4,
    4
  ]
}
 elapsed time: 211.125µs
bash
3> c some_service process_struct [[1,"test", [4]]]
result: {
  "number": 2,
  "string": "test 1",
  "vector": [
    4,
    4
  ]
}
 elapsed time: 211.125µs

Function imports​

A service can consist of more than one module. A module can import functions from other modules. This is done using extern "C" block wrapped with marine macro, and some attribute. Lets try:

First, create a new module:

bash
fluence module new services/some_service/modules/new_module
bash
fluence module new services/some_service/modules/new_module

Then, add it to service.yaml of previously created some_service

yaml
# yaml-language-server: $schema=../../.fluence/schemas/service.yaml.json
# Defines a [Marine service](https://fluence.dev/docs/build/concepts/#services), most importantly the modules that the service consists of. For Fluence CLI, **service** - is a directory which contains this config. You can use `fluence service new` command to generate a template for new service
# Documentation: https://github.com/fluencelabs/fluence-cli/tree/main/docs/configs/service.md
version: 0
name: some_service
modules:
 facade:
   get: modules/some_service 
 other:                      # new line. "other" is an arbitrary chosen name, just make it unique
   get: modules/new_module   # new line. here is the path to the dir with module.yaml
yaml
# yaml-language-server: $schema=../../.fluence/schemas/service.yaml.json
# Defines a [Marine service](https://fluence.dev/docs/build/concepts/#services), most importantly the modules that the service consists of. For Fluence CLI, **service** - is a directory which contains this config. You can use `fluence service new` command to generate a template for new service
# Documentation: https://github.com/fluencelabs/fluence-cli/tree/main/docs/configs/service.md
version: 0
name: some_service
modules:
 facade:
   get: modules/some_service 
 other:                      # new line. "other" is an arbitrary chosen name, just make it unique
   get: modules/new_module   # new line. here is the path to the dir with module.yaml

Update new module's to make it more meaningful:

rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub fn get_hello() -> String {
    format!("Hello")
}
rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub fn get_hello() -> String {
    format!("Hello")
}

And import and use this function in facade module:

rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub struct SomeStruct {
    number: i32,
    string: String,
    vector: Vec<i32>
}
#[marine]
pub fn process_struct(arg: SomeStruct) -> SomeStruct {
    SomeStruct {
        number: arg.number + 1,
        string: format!("{} {}", arg.string, 1),
        vector: vec![arg.vector[0]; 2],
    }
}
#[marine]
pub fn greeting(name: String) -> String {
    let hello = get_hello();
    format!("{}, {}",  hello, name)
}
#[marine] 
#[link(wasm_import_module = "new_module")] // use the name of .wasm file. In this case it is from [[bin]] section of Cargo.toml of module. "host" is reserved here.
extern "C" {
    fn get_hello() -> String; 
}
rust
#![allow(non_snake_case)]
use marine_rs_sdk::marine;
use marine_rs_sdk::module_manifest;
module_manifest!();
pub fn main() {}
#[marine]
pub struct SomeStruct {
    number: i32,
    string: String,
    vector: Vec<i32>
}
#[marine]
pub fn process_struct(arg: SomeStruct) -> SomeStruct {
    SomeStruct {
        number: arg.number + 1,
        string: format!("{} {}", arg.string, 1),
        vector: vec![arg.vector[0]; 2],
    }
}
#[marine]
pub fn greeting(name: String) -> String {
    let hello = get_hello();
    format!("{}, {}",  hello, name)
}
#[marine] 
#[link(wasm_import_module = "new_module")] // use the name of .wasm file. In this case it is from [[bin]] section of Cargo.toml of module. "host" is reserved here.
extern "C" {
    fn get_hello() -> String; 
}

Now check using repl:

bash
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣽
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = cdc9af5b-3931-48fe-a7d4-f18f695b01bb
elapsed time 95.499791ms
1> i
Loaded modules interface:
exported data types (combined from all modules):
data SomeStruct:
  number: i32
  string: string
  vector: []i32
exported functions:
new_module:
  func get_hello() -> string
some_service:
  func process_struct(arg: SomeStruct) -> SomeStruct
  func greeting(name: string) -> string
2> c some_service greeting "fluence"
result: "Hello, fluence"
 elapsed time: 10.474334ms
3> 
bash
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣽
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = cdc9af5b-3931-48fe-a7d4-f18f695b01bb
elapsed time 95.499791ms
1> i
Loaded modules interface:
exported data types (combined from all modules):
data SomeStruct:
  number: i32
  string: string
  vector: []i32
exported functions:
new_module:
  func get_hello() -> string
some_service:
  func process_struct(arg: SomeStruct) -> SomeStruct
  func greeting(name: string) -> string
2> c some_service greeting "fluence"
result: "Hello, fluence"
 elapsed time: 10.474334ms
3>

Mounted binaries​

To unlock some options not possuble in pure wasm, exists a mounted binaries API - a way to call programs on host machine from wasm services. It is similar to the importing functions, the only differences are that wasm_import_module must be "host" and signatures are predefined.

Lets use the curl tool from the host system in our project. First, add import the binary and use it:

rust
use marine_rs_sdk::MountedBinaryStringResult;
/*... skipped code from previous snippets ...*/  
#[marine]
pub fn call_curl(url: String) -> MountedBinaryStringResult {
  curl(vec![url])
}
#[marine]
#[link(wasm_import_module = "host")]
extern "C" {
    fn curl(cmd: Vec<String>) -> MountedBinaryStringResult;
  
}
rust
use marine_rs_sdk::MountedBinaryStringResult;
/*... skipped code from previous snippets ...*/  
#[marine]
pub fn call_curl(url: String) -> MountedBinaryStringResult {
  curl(vec![url])
}
#[marine]
#[link(wasm_import_module = "host")]
extern "C" {
    fn curl(cmd: Vec<String>) -> MountedBinaryStringResult;
  
}

Function name should match the name in the config, arguments should be Vec<String> and return value either MountedBinaryStringResult or MountedBinaryStringResult. The difference is that the latter transform stdin/stdout to String internally, while the other keeps it Vec<u8>

Then add the mounted binary to the config <project_root>/services/some_service/modules/some_service/module.yaml:

yaml
version: 0
type: rust
name: some_service
mountedBinaries:
    curl: /usr/bin/curl
yaml
version: 0
type: rust
name: some_service
mountedBinaries:
    curl: /usr/bin/curl

And now it is runnable:

bash
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣽
Making sure service and modules are downloaded and built... ⢿
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = c3c4f345-c5e9-492c-bf58-4c2424d710c6
elapsed time 90.803042ms
1> c some_service call_curl "google.com"
result: {
  "error": "",
  "ret_code": 0,
  "stderr": "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n                                 Dload  Upload   Total   Spent    Left  Speed\n\r  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0\r100   219  100   219    0     0   1269      0 --:--:-- --:--:-- --:--:--  1351\n",
  "stdout": "<HTML><HEAD><meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<TITLE>301 Moved</TITLE></HEAD><BODY>\n<H1>301 Moved</H1>\nThe document has moved\n<A HREF=\"http://www.google.com/\">here</A>.\r\n</BODY></HTML>\r\n"
}
 elapsed time: 198.43175ms
bash
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣽
Making sure service and modules are downloaded and built... ⢿
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = c3c4f345-c5e9-492c-bf58-4c2424d710c6
elapsed time 90.803042ms
1> c some_service call_curl "google.com"
result: {
  "error": "",
  "ret_code": 0,
  "stderr": "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n                                 Dload  Upload   Total   Spent    Left  Speed\n\r  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0\r100   219  100   219    0     0   1269      0 --:--:-- --:--:-- --:--:--  1351\n",
  "stdout": "<HTML><HEAD><meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<TITLE>301 Moved</TITLE></HEAD><BODY>\n<H1>301 Moved</H1>\nThe document has moved\n<A HREF=\"http://www.google.com/\">here</A>.\r\n</BODY></HTML>\r\n"
}
 elapsed time: 198.43175ms

Call Parameters​

Peers provide additional information about what's happening to services. This includes where the service is, who initiated the call, where are arguments from. It is known as call parameters and structure definitions are here:

rust
pub struct CallParameters {
    /// Peer id of the AIR script initiator.
    pub init_peer_id: String,
    /// Id of the current service.
    pub service_id: String,
    /// Id of the service creator.
    pub service_creator_peer_id: String,
    /// PeerId of the peer who hosts this service.
    pub host_id: String,
    /// Id of the particle which execution resulted a call this service.
    pub particle_id: String,
    /// Security tetraplets which described origin of the arguments.
    pub tetraplets: Vec<Vec<SecurityTetraplet>>,
}
pub struct SecurityTetraplet {
    /// Id of a peer where corresponding value was set.
    pub peer_pk: String,
    /// Id of a service that set corresponding value.
    pub service_id: String,
    /// Name of a function that returned corresponding value.
    pub function_name: String,
    /// Value was produced by applying this `json_path` to the output from `call_service`.
    pub json_path: String,
}
rust
pub struct CallParameters {
    /// Peer id of the AIR script initiator.
    pub init_peer_id: String,
    /// Id of the current service.
    pub service_id: String,
    /// Id of the service creator.
    pub service_creator_peer_id: String,
    /// PeerId of the peer who hosts this service.
    pub host_id: String,
    /// Id of the particle which execution resulted a call this service.
    pub particle_id: String,
    /// Security tetraplets which described origin of the arguments.
    pub tetraplets: Vec<Vec<SecurityTetraplet>>,
}
pub struct SecurityTetraplet {
    /// Id of a peer where corresponding value was set.
    pub peer_pk: String,
    /// Id of a service that set corresponding value.
    pub service_id: String,
    /// Name of a function that returned corresponding value.
    pub function_name: String,
    /// Value was produced by applying this `json_path` to the output from `call_service`.
    pub json_path: String,
}

It is accessible through marine_rs_sdk::get_call_parameters function. Here is an example:

rust
#[marine]
pub fn call_parameters() -> String {
    let cp = marine_rs_sdk::get_call_parameters();
    format!(
        "init_peer_id: {}, service_id: {}, service_creator_peer_id: {}, host_id: {}, particle_id: {}, tetraplets: {:?}",
        cp.init_peer_id,
        cp.service_id,
        cp.service_creator_peer_id,
        cp.host_id,
        cp.particle_id,
        cp.tetraplets
    )
}
rust
#[marine]
pub fn call_parameters() -> String {
    let cp = marine_rs_sdk::get_call_parameters();
    format!(
        "init_peer_id: {}, service_id: {}, service_creator_peer_id: {}, host_id: {}, particle_id: {}, tetraplets: {:?}",
        cp.init_peer_id,
        cp.service_id,
        cp.service_creator_peer_id,
        cp.host_id,
        cp.particle_id,
        cp.tetraplets
    )
}

After adding it to a service, it can be tested. Call parameters can be set manually in repl, as a second json value after arguments:

sh
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣾
Making sure service and modules are downloaded and built... ⣾
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = c35b8c5f-a903-4676-8d72-0fd7b2277cf0
elapsed time 95.512125ms
1> c some_service call_parameters [] ["a", "b", "c", "d", "e", []]
result: "init_peer_id: a, service_id: b, service_creator_peer_id: c, host_id: d, particle_id: e, tetraplets: []"
 elapsed time: 6.426833ms
sh
$ fluence service repl some_service
Making sure service and modules are downloaded and built... ⣾
Making sure service and modules are downloaded and built... ⣾
Making sure service and modules are downloaded and built... done
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Execute help inside repl to see available commands.
Current service <module_name> is: some_service
Call some_service service functions in repl like this:
call some_service <function_name> [<arg1>, <arg2>]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    
Welcome to the Marine REPL (version 0.19.1)
Minimal supported versions
  sdk: 0.6.0
  interface-types: 0.20.0
app service was created with service id = c35b8c5f-a903-4676-8d72-0fd7b2277cf0
elapsed time 95.512125ms
1> c some_service call_parameters [] ["a", "b", "c", "d", "e", []]
result: "init_peer_id: a, service_id: b, service_creator_peer_id: c, host_id: d, particle_id: e, tetraplets: []"
 elapsed time: 6.426833ms