Quinn Slack

Thanks for reviewing this. It doesn't address any of those 3 bullet points you raised; good catches. I'll revise the patch. --- BTW, the master test failure (which, to be...

OK, I've rebased the patch against current master and addressed @cryptix's issue. Responses below to the points @jmoiron raised. Also, @jmoiron, let me know if you think the reflection code...

In the [CHD paper](http://cmph.sourceforge.net/papers/chm92.pdf) I see that `n/m` determines (at least in part) the probability of finding a collision-free hash:  So the solution might just be to bump up...

It seems this library chooses `n=len(keys)` and then lets `m=n/c` (where `c=2`). However, the CHD paper's Simple Example (Section 4) chooses `m=len(keys)` (and defines `m` to be "the number of...

Whoops, sorry. I'll take a look at the right paper.

Agreed. I think this is a quick fix. I will take a look this weekend. If anyone else wants to submit a PR in the meantime, I would happily merge...

Yes, this would be great. Overall I would like to overhaul goreturns soon and make it shell out to goimports/gofmt and (probably) pass along flags to future-proof it. A PR...

Thanks. Good call. I'd happily accept a PR that implements this.

Thanks! I think there is some other underlying issue, though. Posted a troubleshooting question at https://github.com/sqs/goreturns/issues/50#issuecomment-438527392.

Hmm, this should not be happening. What is the output of `git remote -v` in your `goreturns `directory?