Markus
Markus
No problem, take the time you need. There's no push from my side.
Glad it got merged and thanks for your review and your finishing up! Unfortunately I couldn't review your changes (couldn't find the time), but I'm sure it works out like...
> I'm not happy with the current strategy of raising an error, and catching it with a rescue_from because that enqueues another Execution record AFAIK this happens with https://github.com/nickelser/activejob-traffic_control, too....
I think the PR is ready now. What do you think? If you merge and release this, it should be a bigger version change (maybe even 1.0.0??).
Don't use a string as a key, it must be an instance of `OpenSSL::PKey::RSA`. You can use `OpenSSL::PKey::RSA.new(KEY)` in your example.
I think it should work with a long `acquisition_timeout` so it will wait for that time? You probably should amend `acquisition_delay` then, too, or else it might try to acquire...
I did some tests with different `cpu-used` `6` and `3`: both finished the encode. Tried `2` again: crashed. Looks like a bug in aomenc then? Edit: Starting encode with `cpu-used...
Failed with `cpu-used 1`, too, so seems to be something below `3` that causes the crash. I'll see if I can reproduce it without av1an.
> an out-of-memory situation [...] Check the system log to see if an oom event occurred I lowered the worker count to 6 to make sure it was no out...
yeah, would be great to have totalterminal to cover the fullscreen-app. So your terminal is always with you, no matter what fullscreen app you're working on. Might be hard to...