We re-implemented the library using latest C++ enhancements, and made it fully compatible with boost.asio toolkit
This means that you benefit form :
- asynchronous high level primitives (ex: async_read, async_read_until ....)
- TLS layererd over UDT using latest OpenSSL (Yes, TLS, not DTLS)
- extendible framework so that you can use futures or coroutines to layer your code as you wish
- Everything that make it soo cool to work with asio.
We try to be fully compliant with existing UDT standard library, based on current trunk source code, but we do not support everything... yet (e.g. we do not support messaging API).
The code is in it's early stage, and it is part of a even bigger framework : Secure Socket Funneling. UDT is planned to be used instead of TCP on future release of SSF for even better performances.
Warning :
We encountered timer resolution issues on windows that impact overall performance on UDT. This issue is currently discussed with boost.asio author. We encourage you to apply the proposed patch before testing the library.
There is room for optimization. At this point, the library is functionnal, but young.
Any questions / remarks or feedbacks are welcome, and we're fully open to contributors.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi everyone,
We were inspired by this library and we tried to make it compliant with the boost.asio toolkit.
Boost Asio based UDT
We re-implemented the library using latest C++ enhancements, and made it fully compatible with boost.asio toolkit
This means that you benefit form :
- asynchronous high level primitives (ex: async_read, async_read_until ....)
- TLS layererd over UDT using latest OpenSSL (Yes, TLS, not DTLS)
- extendible framework so that you can use futures or coroutines to layer your code as you wish
- Everything that make it soo cool to work with asio.
We try to be fully compliant with existing UDT standard library, based on current trunk source code, but we do not support everything... yet (e.g. we do not support messaging API).
The code is in it's early stage, and it is part of a even bigger framework : Secure Socket Funneling. UDT is planned to be used instead of TCP on future release of SSF for even better performances.
Warning :
We encountered timer resolution issues on windows that impact overall performance on UDT. This issue is currently discussed with boost.asio author. We encourage you to apply the proposed patch before testing the library.
There is room for optimization. At this point, the library is functionnal, but young.
Any questions / remarks or feedbacks are welcome, and we're fully open to contributors.