> what's the best way to generate "safe" license keys? In the library (like the kind with books) field where I work, one identifier standard was devised that i

> what's the best way to generate "safe" license keys? In the library (like the kind with books) field where I work, one identifier standard was devised that intentionally has alternating letters and digits, with never more than two letters in a row. Explicitly for the intention of avoiding the possibility of any meaningful words (that might end up being offensive or just off-putting in an undesirable way.) It does make the identifiers longer for the same entropy/byte width, compared to a more normal BASE-X with an alphabet. Which mattered to me when they were going to be used in a URL, although probably doesn't for a license key. I personally in my projects stopped using this system for a more straightforward "Ascii-85" like encoding (which can contain coincidental meaningful words), because it was more convenient. The particular system the library community was using [https://n2t.net/e/noid.html https://n2t.net/e/noid.html] was, I still think, over-complicated for at least my needs, but the alternating letter/number schema seems attractive to me now and perhaps worth slightly more characters in identifiers and slightly more complex algorithm for creation than a simple base-x encoding.