[Permissions] Pairing code issue

[boyars]

An interesting nuance (and annoyance) that I have spotted this year when I generate pairing codes is that Canvas seems to generate the codes randomly and often they contain a capital I (the letter between H and J) and/or a lower case l (the letter between k and m) and/or the number 1. Due to the default font used by Canvas (and most web browsers) being sans-serif, all 3 of these characters are easily confused with each other causing a nuisance with giving codes to parents (too many issues of what letter is that). Typically, I have to copy the code, paste it into a word processor and then change the font to a serif font (like times new roman) in order to determine which characters where generated. I would suggest that the code generation algorithm be modified to specifically exclude these 3 characters (and maybe the letter O and the number 0), as these characters can look similar on a lot of fonts and lead to confusion when giving codes to parents. As an example, I know that Southwest Airlines specifically excludes the numbers 0 and 1 when generating their alphanumeric confirmation numbers (so if you see one of those in your number, you know it is a letter), as they can easily be misread as letters (their codes are not case sensitive, hence I suggest the additional step here of excluding two letters that are easily confused). I do recognize the need for security, but reducing the number of characters that could be used in a code from 62 to 56 still leaves a large enough character set that there would be a minimal reduction in security.