I think it would make sense to add Nash to CCXT library.
Anyone feeling up for the challenge?
I think it would make sense to add Nash to CCXT library.
Anyone feeling up for the challenge?
Thanks! Didn’t think to check.
Is it from you? CI appears to be failing
Hi @Oldsport, I’m the guy that made the pull request.
I just did the first steps and was planning updates but first want to make everything works.
As Nash is using graphql this library does not supports is out of the box. Also adding the SDK is not an option cause the Javascript will be transpiled to python and PHP.
The CI is indeed failing because of some issue when transpiling to python. I will check with one of the devs how this can be solved.
Contributions a very welcome
I had actually just found you from your name and picture. Thanks for your work on this
Unfortunately my dev days are behind me, don’t think I’d be able to help.
We talked with CCXT team a while back, but it can’t have dependencies due to the transpiling like @laurens55 mentioned and some other auditing concerns. So it will/would require considerable engineering (that the community might have in long term available but we decided to focus elsewhere) to port all the cryptography to the CCXT library. We have other efforts in motion to allow easier migration to Nash. You should hear from that in a few weeks.
We have other efforts in motion to allow easier migration to Nash. You should hear from that in a few weeks.
@canesin Can you elaborate about the efforts that are done in order to improve integrations to Nash? I heard about a Rust project that can compile into other languages, like Python.
I’m currently trying to integrate Nash to Hummingbot, but I’m currently struggling with order placement because of all the cryptography operations needed (mainly how to fill the blockchainSignatures
field)
Thanks in advance for your answer.
Yes, @ethan and @Robert_nash (our research engineer) are working on a Rust client for Nash and @steffenix (our backend engineer) is working on generalizing the interfaces for several languages (including Python, NodeJS, Java and C#).
The issue is exactly what you mention - Nash is not a run-of-the-mill exchange API since it is actually powered by a lot of crypto. Is quite infeasible to ask the community alone to implement all the details of the MPC protocol in every integration, so we are building this so that it can be used as a library or compiled in portable format.
We are still a couple of weeks away of open sourcing this work. But should make life for people integrating the exchange much easier.
That’s good news, but I have to ask: what is preventing you to open source these works right now?
It is WIP, there will probably be some refactor before publishing.
I wasn’t speaking about publishing, just open sourcing these projects repos, so the Nash dev community could help, or at least give feedback.
But I guess you are a perfectionist, I can understand that…