mfr-itr

Maybe then return a `Future`? See #80 Another problem is that you can call `.error` and such multiple times. I'm thinking something like that for readdir: ``` let mut reply...

Hum, right XD

Oh, OK, I thought it only kept the last snapshot on local, that makes sense. Did you check if `btrfs send` correctly handles unreliable connections? Are the garbled backups kept...

By unreliable network, I mean that the ssh connection is closed like every 5min (for example, by disconnecting/reconnecting the wifi network), so that no upload can complete without at least...

So the upload process will start again?

Hum, did not know that was possible. At first glance it seems good enough, not sure if it would be sufficient though... Edit: my bad, I thought you meant it...

Yeah, but it must be written at runtime. Better safe than sorry! Also, we must into account that the function can be called multiple times or after some data was...

A quick way to do that would be to generate an output compatible with https://github.com/brendangregg/FlameGraph#2-fold-stacks, and `flamegraph.pl` could be used to generate the graph. A native way to do that...

This works for me: ```rust fn print_trace(s: flame::Span, prefix: &str) { let my_prefix = format!("{};{}", prefix, s.name); println!("{} {}", my_prefix, s.delta); s.children.into_iter().for_each(|c| print_trace(c, &my_prefix)); } for span in flame::threads().into_iter().flat_map(|t| t.spans.into_iter())...

It produces the same output as `stackcollapse-*`, so I run `cargo run | flamegraph > out.svg`. `flamegraph` combines the frames if they have the same path. It takes a long...