Are these commands tested?
Every one. Each command on this page was executed against the Rendobar API on the same source file, and the page records what it produced: encode time, output size and billed cost. 28 runs, $0.066088 to measure.
Every command here was run, not copied. Each page records what the command produced on a known file, and most of them also run the version people try first, so you can see what going wrong actually looks like.
Of the 8 naive commands we ran, only 1 failed outright. The other 7 succeeded while producing something worse, which is the failure mode nothing in a pipeline catches.
Every one. Each command on this page was executed against the Rendobar API on the same source file, and the page records what it produced: encode time, output size and billed cost. 28 runs, $0.066088 to measure.
Because "don't forget this flag" is an assertion and a measurement is evidence. We ran the naive version of each recipe as well. Only 1 of 8 failed outright; the rest succeeded while producing something worse, which is the case no error handler catches.
No. Every command here ran through the API against a URL, with no local FFmpeg binary and no server to manage.
Yes. Each is a single API call with the command as a parameter, so the only thing that changes is the input URL.