MarkSwanson

https://github.com/rust-lang/cargo/issues/9521 I'm not sure if I should place this here. Just trying to help with keeping track...

Fyi f64 supports an accuracy of ~= 1.0 x 10 ^ 16. Folks doing scientific work need around 1.0 x 10 ^ 40.

I tried BigRational but it just doesn't fit - I can't always specify my numbers as a fraction. Of the various arbitrary precision libraries out there the num-decimal crate provides...

The problem with most of those (I'm not familiar with rug) is that the number type they provide does not support Copy. Here's why it's bad, and how I worked...

I took a stab an implemented support for f256 using the quad library [email protected]:MarkSwanson/uom_num_decimal.git branch=qd_f256 cargo test shows 67 warnings. It breaks f64 and f32 a tiny bit. I left...

Cannot build in Ubuntu 12. Error: ... git ls-files --others -i --exclude-standard | xargs rm -f || true rm_f /home/maswanso/src/build-couchdb/build/bin/autoconf rm_f /home/maswanso/src/build-couchdb/build/bin/autoreconf rm_f /home/maswanso/src/build-couchdb/build/bin/autoheader rm_f /home/maswanso/src/build-couchdb/build/bin/autom4te rake aborted! Command failed...

Nice to see the roadmap. Especially interested to see replication specifics!

Also, /etc/nsswitch.conf: hosts: files dns /etc/resolv.conf: nameserver 127.0.0.1

Thanks for looking into it. I found it odd the first few sendto bytes were different - recvfrom too. If I can format things better, or do different tests please...

It does, and this may be the tricky part because it relies on an OUT_DIR env var supplied by rustc. For example, on my system the source file winds up...