Provide predefined alphabets containing only characters that are clearly distinguishable when read & spoken

[matthewadams]

Motivation: to provide alphabets that produce ids that can be both read & spoken without ambiguity.

Please provide alphabets that contain only those characters whose spoken & written forms are clearly distinguishable and unambiguous when spoken. Default, I suppose, would be en_US, but it'd be bonus to provide alphabets for many locales in both case-sensitive and case-insensitive flavors (for locales that support casing).

As an en_US example, the characters M & N would be removed because they sound too similar when spoken, as would S & F, and so on.

1 & l ("el") as well as 0 & O should be excluded because they look too similar in many locales.

My suggestion for en_US (case-insensitive) might be: AEGHJKQRUVWXYZ246789

[CyberAP]

@matthewadams how do we determine what's easy to speak and what's not?

For things that are confusing when reading we have nolookaliakes.

[Blue-smoke007]

I'm applying to solve the problem of creating alphabets for IDs that are easy to read and speak without confusion. My approach involves filtering out visually or phonetically similar characters (like 1 and l, or M and N), starting with en_US and supporting other locales. I’d build a script to generate safe alphabets for both case-sensitive and case-insensitive use. With experience in usability-focused development and attention to detail, I’m well-suited for this task.