I was looking to generate some special storage tables in my application based on a normalized string (think json string used as an id). At first I was using this item as an id, but the table has gotten so large that even reasonable queries are taking unreasonable amounts of time.
I was going to use this string as part of the database name, so I was looking at how to do that. It looks like I can, but the overwhelmingly common response is to "not do it" and "I'll regret it later."
I am fine changing up this id so I avoid special characters, but I'd really like to know why this is so taboo. I have found plenty of examples saying not to, but no one actually to explain why. The closest I have found to an answer is that "it will cause pain in the development cycle"
Particularly in situations where the database is pretty much entirely managed in code, is there a good reason?
I know this runs the risk of being opinion based, so I'd really like some quantifiable reasons.