Clark Boylan

As a workaround to this I think you can manually edit the key config file to set `Le_Keylength` to `"2048"`. Neither the workaround or the PR have been tested locally...

It looks like the docker hub hosted image hasn't been updated since https://github.com/walthowd/husbzb-firmware/commit/0ab67453a1e552706c260a4c83a6ab7bfe988917 was committed to the repo. That commit renamed `ncp-uart-sw.ebl` to `ncp-uart-sw-6.6.3.ebl`. You can confirm this checking the...

I've run into this independently and done a bit of digging before I saw this issue and thought I would share a bit of what I've learned. Editing `setup.cfg` to...

One difference I've found is that `build` via `pep517` appears to fork a subprocess to run the `setuptools.build_meta.__legacy__`: https://github.com/pypa/pyproject-hooks/blob/v0.13.0/pep517/wrappers.py#L306-L313. Tox via pyproject-api appears to call the backend with the flag...

> Okay, so I've narrowed the hanging issue down to this line: > > https://github.com/tox-dev/tox/blob/12f6268a1800550e68db028c63db4a2813c1edab/src/tox/tox_env/python/virtual_env/package/pyproject.py#L164 > > If I comment that out, things pass successfully. That's a bit dumb though...

> > The expectation from `build` (and when you run `setup.py` explicitly) is that these happen as distinct processes with fresh memory spaces. > > What makes you think this...

> I'd be happy to accept a PR, though, that configures this behavior for those backends requiring this fresh subprocess behavior. PR for that is welcome. Do you think this...

Is this the commit that fixes things: https://github.com/ether/etherpad-lite/commit/a61f634586017dcadffd859820b66cd5916cef3a ? We also run a jitsi meet server with embedded etherpad documents and are hitting a similar issue (the errors are actually...