Ellis Pritchard

All FT code is MIT; see the top-level organisation.

Sounds useful.

Hey @bforchhammer, thanks for the kind words! Yes, this is something I have considered recently, when re-factoring the trace and span ids *away* from integers (there was too much repeated...

NB Tapper supports passing through 64-bit trace ids (which used to be the default for Zipkin, but some users ran out of these pretty quickly, e.g. Twitter!), so for received...

Yep, it's super simple, and only breaks one test in 64-bit mode ("the trace id is 128-bits"!). I'll probably just commit this, as it might be useful for other people:...

Hey there, the [Zipkin spec](https://github.com/openzipkin/zipkin-api/blob/7692ca7be4dc3be9225db550d60c4d30e6e9ec59/thrift/zipkinCore.thrift#L396) says that binary annotations should always be binary, i.e. a string, and no other value. There used to be a bug in Tapper that would...

I agree that this is a bit restrictive, but since any string can be made an atom, and these are converted to strings on encoding to a zipkin trace, you...

I see. Thanks, it would be nice to at least fix that inconsistency.

I'm getting this error when I've got custom scalars, when the custom scalar validation rejects the input: the nicely informative message from my custom scalar is not shown, but the...

NB the original error from validation is also reported in the result, so `ValuesOfCorrectType` is also being run, but this doesn't stop related errors also being generated from the cost...