Conversation
|
Fixed the formatting problems. |
This requires upgrading secp256k1 as well, because otherwise they have conflicting dependendencies no the cc crate.
|
I saw this failure, too, but as far as I can tell it is intermittent. It doesn't seem like this would be related to this PR, since this test doesn't even seem to be using ring. |
|
its because we have a probabilistic test. It fails from time to time. |
|
We actually try to get rid of ring library all-together. Too many other libraries are using it with different versions and it creates a lot of compatibility problems. Since we expect Curv to be integrated into other systems it happened to us more than once that one of the other libraries used ring from a different - not latest - version. |
|
@oleiba if you have inputs to add |
|
We have a similar issue: we want to use ring for other things as well, but we prefer to stay on the latest version. I don't know what timeline you have in mind for replacing ring, but in the meantime (and since I worked on this PR anyway) it would seem nicer if you rely on the latest one. |
|
Make sense. |
|
All of our stack is aligned on this specific version of ring: We use ring for hash function so it is not very critical for us. |
|
https://crates.io/crates/merkle merkle also depends on ring, removing ring is not enough |
|
Dependabot is a great way to stay up to date, I use it for many of my projects. |
|
I removed ring in this pr: #42 |
|
Can I close this PR? (ring is no longer a dependency for this library) |
|
I'll close it. |
This requires upgrading secp256k1 as well, because otherwise they have conflicting dependendencies no the cc crate.