IT Jack of all trades, developing and maintaining Mumps/Intersystems systems since the 90's.
This looks so much easier than when I had to do it and is very welcome.
At the time I moved it to a personal email so I didn't have to do it again changing companies.
This means it should actually be possible to keep it linked to wherever you are working.
I appreciate the comments, some things to think about.
If you have planned tables and data properly uniqueness should not be an issue and it looks like the performance consideration is unlikely to be a problem in most cases.
Being able to check the raw data easily with a standard global naming is, for me at least, a benefit of not using the hashed names.
However, knowing these reasons is very helpful in planning and identifying those edge cases.
Is there anything else I have missed that would make this more of an issue?
Great overview, I have had plenty of discussions with IT departments who don't understand that Linux, and Windows, will use available resources, especially for IO, and it is only an issue if it is persistent or unusual, and obviously if there are performance/stability issues.
Being able to provide metrics over time and proving it is "normal" usually helps.