Closed Barbarrosa closed 2 years ago
nice performance optimization! :) Do you know what the breaking change is in 6?
@arj03 There's some bounds checking to prevent NaN
and Infinity
values from being returned or used by accident, so an error will be thrown in those cases.
I'm wondering if varint is abandoned now b/c the maintainer hasn't commented on the PR for over a month. The maintainer's still working on other Github projects, though.
Latest release is a while ago. Maybe you can try and contact him on mail and see if he's willing to let you commit and do releases. I can help review.
I didn't get an email, but I sent a private message to an account where the maintainer appears to be active. At this point I doubt that I'm going to get a response. Do you have any thoughts about creating an alternate package or embedding the code here?
@arj03 Glad to see that you got an improvement in #37 by unrolling the loop. Did you have a chance to test with the Math.pow
calls pre-calculated, or do you think that gets optimized?
It is actually faster. Thanks :)
The new version includes bounds checking, which is nice. It would be best to wait for the upgrade until https://github.com/chrisdickinson/varint/pull/24 is merged so we can benefit from the
decode
performance optimization.