ASCII subset for readable random keys (no O, 0, 1, I or l confusion)?
Every so often you have to enter a product activation code or similar.
If it's badly designed you end up trying to guess if you're looking at
zeroes or capital Os, or at ones, lowercase Ls or uppercase Is.
Such characters can often only be distinguished by humans by context, i.e.
their use in real words, and so shouldn't be used in random sequences.
Has anyone done any work to determine the subset of letters that most
people can uniquely distinguish in most reasonable fonts irrespective of
context (and assuming reasonable print quality)?
E.g. characters suitable for use in keys sent by plain text email - one
cannot know what font the end user will use to read the email - but it's
likely not too unusual. For printed material one obviously has more
control.
I'll point out before anyone else does that this question is similar to
(some might consider it a duplicate of) "OCR - most "different" or
"recognizable" ASCII characters?"
No comments:
Post a Comment