This is the official SDK for Hyperledger Indy, which provides a distributed-ledger-based foundation for self-sovereign identity. Indy provides a software ecosystem for private, secure, and powerful identity, and the Indy SDK enables clients for it. The major artifact of the SDK is a c-callable library; there are also convenience wrappers for various programming languages and Indy CLI tool.
All bugs, stories, and backlog for this project are managed through Hyperledger's Jira in project IS (note that regular Indy tickets are in the INDY project instead...). Also, make sure to join us on Hyperledger's Rocket.Chat at #indy-sdk to discuss. You will need a Linux Foundation login to get access to these channels
If you have just started learning about self-sovereign identity, here are some resources to increase your understanding:
-
This extended tutorial introduces Indy, explains how the whole ecosystem works, and how the functions in the SDK can be used to construct rich clients: Indy-SDK Getting-Started Guide
-
Hyperledger Indy Working Group calls happen every Thursday at 8amPT, 9amMT, 11amET, 4pmBST. Add to your calendar and join from any device: https://zoom.us/j/232861185
-
A recent webinar explaining self-sovereign identity using Hyperledger Indy and Sovrin: SSI Meetup Webinar
-
Visit the main resource for all things "Indy" to get acquainted with the code base, helpful resources, and up-to-date information: Hyperledger Wiki-Indy.
-
You may also want to look at the older guide that explored the ecosystem via command line. That material is being rewritten but still contains some useful ideas.
The major artifact of the SDK is a C-callable library that provides the basic building blocks for the creation of applications on the top of Hyperledger Indy. It is available for most popular desktop, mobile and server platforms.
A set of libindy wrappers for developing Indy-based applications in your favorite programming language. Indy SDK provides libindy wrappers for the following programming languages and platforms:
Indy CLI is the official command line interface that helps Indy developers and administrators.
Libnullpay is a libindy plugin that can be used for development of applications that use the Payments API of Indy SDK.
Libvcx is a c-callable library built on top of libindy that provides a high-level credential exchange protocol. It simplifies creation of agent applications and provides better agent-2-agent interoperability for Hyperledger Indy infrastructure.
This library is currently in an experimental state and is not part of official releases.
A set of libvcx wrappers for developing vcx-based applications in your favorite programming language.
Indy SDK provides libvcx wrappers for the following programming languages and platforms:
These wrappers are currently in experimental state and it is not part of official releases.
For the main workflow example check VCX Python demo.
Dummy Cloud Agent is simple implementation of VCX compatible Cloud Agent. The main purpose of this implementation is VCX testing, demos and documentation of VCX protocol. There is an parallel work on providing reference implementation of Agent2Agent protocol in indy-agent repo. We plan to migrate to this protocol and implementation soon.
Short, simple tutorials that demonstrate how to accomplish common tasks are also available. See the docs/how-tos folder.
- Write a DID and Query Its Verkey
- Rotate a Key
- Save a Schema and Cred Def
- Issue a Credential
- Negotiate a Proof
- Send a Secure Message
The Indy SDK release process defines the following release channels:
master
- development builds for each push to master branch.rc
- release candidates.stable
- stable releases.
Please refer to our release workflow for more details.
It is recommended to install the SDK packages with APT:
sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 68DB5E88
sudo add-apt-repository "deb https://repo.sovrin.org/sdk/deb xenial {release channel}"
sudo apt-get update
sudo apt-get install -y libindy
{release channel} must be replaced with master, rc or stable to define corresponded release channel. Please See the section "Release channels" above for more details.
- Go to https://repo.sovrin.org/windows/libindy/{release-channel}.
- Download last version of libindy.
- Unzip archives to the directory where you want to save working library.
- After unzip you will get next structure of files:
Your working directory
include
...
lib
indy.dll
libeay32md.dll
libsodium.dll
libzmq.dll
ssleay32md.dll
include
contains c-header files which contains all necessary declarations
that may be need for your applications.
lib
contains all necessary binaries which contains libindy and all it's dependencies.
You must add to PATH environment variable path to lib
. It's necessary for dynamic linkage
your application with libindy.
{release channel} must be replaced with master, rc or stable to define corresponded release channel. See section "Release channels" for more details.
See wrapper iOS install documentation.
- Go to
https://repo.sovrin.org/android/libindy/{release-channel}
. - 3 architecture are supported as of now arm,arm64 and x86.
- Download latest version of libindy.
- Unzip archives to the directory where you want to save the
.so
files. - After unzip you will get next structure of files:
Your working directory
include
...
lib
libindy.so
libindy_shared.so
libindy.a
include
contains c-header files which contains all necessary declarations
that may be need for your applications.
lib
contains three types of binaries.
-
libindy.so
- This is a shared library which is statically linked with all the depenedencies. You dont need to sidelaod other dependencies like zmq, sodium and openssl to android app if you use this. -
libindy_shared.so
- This is pure shared library. It is not dynamically linked to its dependencies. You need to sideload the binaries with its dependencies. You can download the needed pre-built dependencies from here- Rename this library to
libindy.so
before loading it into the app. This will help you in having the compatibility with existing wrappers.
- Rename this library to
-
libindy.a
- This is a static library, which is compiled with NDK.
{release channel} must be replaced with rc or stable to define corresponded release channel. See section "Release channels" for more details.
Note :
-
[WARNING] This library should be considered as experimental as currently unit tests are not executed in the CI phase.
-
We are using the NDK16b because it is the last NDK to have support for
gnustl_shared
stl. gnustl_shared is deprecated in latest NDK. gnustal_shared is needed because the dependencies are compiled using gnustal_shared and you will get build errors if more than one type of stl while compiling.
Pre-built libraries are not provided for MacOS. Please look here for details on building from source for MacOS.
Note: After building libindy
, add the path containing the library the LD_LIBRARY_PATH
and
DYLD_LIBRARY_PATH
environment variables. This is necessary for dynamically linking
your application with libindy
. The dynamic linker will first check for the library in
LD_LIBRARY_PATH
if the library in your application doesn't include directory names.
If the library in your application does include any directory name, then dynamic
linker will search for the library in DYLD_LIBRARY_PATH
(not LD_LIBRARY_PATH
)
so we recommend you set both variables to be safe.
Pre-built libraries are not provided for RHEL-based distributions. Please look here for details on building from source for RHEL-based distributions.
After successfully compiling libindy
, you will need to add the path containing libindy.so
to the
LD_LIBRARY_PATH
environment variable. This is required for your application to link to
libindy
.
- Ubuntu based distributions (Ubuntu 16.04)
- RHEL based distributions (Amazon Linux 2017.03)
- Windows
- MacOS
- Android
Note:
By default cargo build
produce debug artifacts with a large amount of run-time checks.
It's good for development, but this build can be in 100+ times slower for some math calculation.
If you would like to analyse CPU performance of libindy for your use case, you have to use release artifacts (cargo build --release
).
To test the SDK codebase with a virtual Indy node network, you can start a pool of local nodes using docker:
Note: If you are getting a PoolLedgerTimeout error it's because the IP addresses in cli/docker_pool_transactions_genesis and the pool configuration don't match. Use method 3 to configure the IPs of the docker containers to match the pool.
Start the pool of local nodes on 127.0.0.1:9701-9708
with Docker by running:
docker build -f ci/indy-pool.dockerfile -t indy_pool .
docker run -itd -p 9701-9708:9701-9708 indy_pool
Dockerfile ci/indy-pool.dockerfile
supports an optional pool_ip param that allows
changing ip of pool nodes in generated pool configuration.
You can start the pool with e.g. with the IP address of your development machine's WIFI interface so that mobile apps in the same network can reach the pool.
# replace 192.168.179.90 with your wifi IP address
docker build --build-arg pool_ip=192.168.179.90 -f ci/indy-pool.dockerfile -t indy_pool .
docker run -itd -p 192.168.179.90:9701-9708:9701-9708 indy_pool
To connect to the pool the IP addresses in /var/lib/indy/sandbox/pool_transactions_genesis (in docker) and the pool configuration you use in your mobile app must match.
The following commands allow to start local nodes pool in custom docker network and access this pool by custom ip in docker network:
docker network create --subnet 10.0.0.0/8 indy_pool_network
docker build --build-arg pool_ip=10.0.0.2 -f ci/indy-pool.dockerfile -t indy_pool .
docker run -d --ip="10.0.0.2" --net=indy_pool_network indy_pool
Note that for Windows and MacOS this approach has some issues. Docker for these OS run in their virtual environment. First command creates network for container and host can't get access to that network because container placed on virtual machine. You must appropriate set up networking on your virtual environment. See the instructions for MacOS below.
If you use some Docker distribution based on Virtual Box you can use Virtual Box's port forwarding future to map 9701-9709 container ports to local 9701-9709 ports.
If you use VMWare Fusion to run Docker locally, follow the instructions from this article and add the following lines to /Library/Preferences/VMware Fusion/vmnet8/nat.conf:
# Use these with care - anyone can enter into your VM through these...
# The format and example are as follows:
#<external port number> = <VM's IP address>:<VM's port number>
#8080 = 172.16.3.128:80
9701 = <your_docker_ip>:9701
9702 = <your_docker_ip>:9702
9703 = <your_docker_ip>:9703
9704 = <your_docker_ip>:9704
9705 = <your_docker_ip>:9705
9706 = <your_docker_ip>:9706
9707 = <your_docker_ip>:9707
9708 = <your_docker_ip>:9708
9709 = <your_docker_ip>:9709
where <your_docker_ip> is your Docker host IP.
Docker machine needs to be rebooted after these changes.
The following wrappers are tested and complete. There is also active work on a wrapper for Go; visit #indy-sdk on Rocket.Chat for details.
- An explanation of how to install the official command line interface for that provides commands to manage wallets and interactions with the ledger: Indy CLI
The documents that provide necessary information for Libindy migration. This document is written for developers using Libindy 1.3.0 to provide necessary information and to simplify their transition to API of Libindy 1.4.0.
- We'd love your help; see these instructions on how to contribute.
- You may also want to read this info about maintainers and our process.
- We use developer certificate of origin (DCO) in all hyperledger repositories, so to get your pull requests accepted, you must certify your commits by signing off on each commit. More information can be found in Signing Commits article.
- Libindy implements multithreading approach based on mpsc channels.
If your application needs to use Libindy from multiple processes you should keep in mind the following restrictions:
- Fork - duplicates only the main thread. So, child threads will not be duplicated. If any out-of-process requirements are possible, the caller must fork first before any calls to Libindy (otherwise the command from a child thread will hang). Fork is only available on Unix.
- Popen - spawns a new OS level process which will create its own child threads. Popen is cross-platform.