YouTube has a limit on how long YouTube video titles can be.
The upload.py script knows how to check for that and keep us from uploading that file.

But it seems like it is better caught at the split.py step or maybe a step before that, when we download the JSON.
Because when we find out at the upload.py step, we need to modify the JSON, and possibly rename the already split mp4 files before we can proceed.

Here's an example of a talk with a too-long title that triggered this issue.
https://www.socallinuxexpo.org/scale/21x/presentations/optimal-approaches-real-time-machine-learning-apache-spark-kubernetes-best