The new job manager allows you to run in parallel on your own machine, the old one didn't. The new one can submit jobs to a remote cluster, before you had to copy the script by hand, manually, and copy the results back, and write your own PBS script, etc. These are huge improvements that users have specifically asked for, for many years, and could not be implemented in the old Job manager.
The ability to run a simple script locally should be roughly the same as in the old one. There have certainly been some bugs in the preview versions, for sure, and we'll do our best to fix any new ones that come up, but once you learn the new approach I'm sure you will appreciate it. Else, do feel free to provide your concrete input on possible improvements.
As for not wanting to save Python scripts, I both agree and disagree. In general, I don't think you would ever not want to save your script for a serious calculation - the script is the only complete record of what the calculation actually is, and the only way to rerun it in a new version or with modified parameters. Sadly, in the past we have had several support cases where users have problems with calculations, but didn't save the script, so we can't help. Saving your script before you run is always the best practice.
I do see situations where you run small test scripts and perhaps it's not necessary to save them, but the only hassle now is that you have to delete some unwanted files in the end. Just call them test1.py and in the end it's easy to identify what to clean up.
I didn't understand the point about "automatically assigning names", since there is no concept of "names" in VNL... The output file is named x.log if your input file is x.py, that's true, but that's pretty logical and makes it easy to keep track of things, doesn't it? Otherwise there is no naming that can't be modified by the user.
But again - all constructive feedback is most welcome, and we'll keep improving it. However, comments like "please make some improvements" are pretty useless, to be honest.
