On 6/17/13 10:08 AM, "[email protected]" <[email protected]> wrote:
>I am not an expert on elliptic curve cryptography so my analysis may go >wrong but the issue seemed to be that generated signature length can vary >because signature must be different and unique each time signing is done. That might explain part of it, but it doesn't explain why the length varies (not that you're wrong, just that that's a distinct issue. > Therefore the length can be shorter than expected which has to be taken >into account. I am not claiming that these changes work in all cases >because I tested code only with my simple test case. >Anyway I hope this helps to find a better solution for the problem. I'm doing a release very shortly, but you can appreciate that I need more time with this to think about including it, so it's not going to be applied right now. Can you please file an issue in Jira with your patch and your test case? If not, you've done enough, but I just want to make sure it doesn't get lost. -- Scott
