Noel Maersk

Does the same happen if you load the key from a file (using `load_key()`)? Are you sure the key/secret are valid? Did you previously use them, especially with a different...

Did you use the same key/secret pair when trying both Coinigy and `krakenex`? The former may be using a different `nonce` scheme - `krakenex` uses epoch time, multiplied by a...

The `nonce` that's sent with a request has nothing to do with time. It's a counter. One might just as well use `0, 1, 2, 3...` (starting off with a...

Actually - sorry! - reopening with an edited, more verbose title; so that there's a clear indication of an unsupported use pattern. The issue would likely be brought up again...

@pietrosperoni That seems to me like an overly-complicated way of dealing with unsynced time, in a scenario where keys are still being reused. (I find this inappropriate from a security...

What @haslo said. ^_^ There's also [an example API provided by Kraken themselves](https://github.com/krakenfx/kraken-wsclient-py/blob/master/kraken_wsclient_py/kraken_wsclient_py.py), (linked in [the WS API docs](https://www.kraken.com/features/websocket-api)). It's possible there'll be no need to fiddle with this package,...

The issue is now [documented](https://python3-krakenex.readthedocs.io/en/v0.1.4/#krakenex.API).

See [`pykrakenapi`](https://github.com/dominiktraxl/pykrakenapi) (via https://github.com/veox/python3-krakenex/issues/40#issuecomment-346184410) for a higher-level library that has this.

I've been live-testing this since the PR was opened. An undesirable quirk that's surfaced with this is that, should a query fail even after `self.retries` attempts, the Python stack trace...

> Have you noticed this inconsistency in the Spread data (using Krakenex or not)? Yes, this is a known behaviour of upstream. See https://github.com/veox/python3-krakenex/issues/112#issuecomment-454334977 for details. > Does Krakenex take...