Conversation

My unpopular opinion is that homoglyph attacks in software code should be handled by normalizing identifiers such that "foo" and "fοο" (second identifier uses GREEK SMALL LETTER OMICRON) are the same variable. Moreover, var_name and varName should be normalized to be identical

Quote Tweet

Björkus "Fix It Or Die Trying" Dorkus

@__phantomderp

Nov 1, 2021

Good morning to everyone except the people blaming Unicode for Homoglyph attacks.

Show this thread

zaratustra the phyrric fiasco

@zarawesome

Nov 1, 2021

Replying to

@mcclure111

What about ladle / Iadle / iadle

Replying to

stop displaying code in sans serif fonts lol :(

Replying to

and

Decent monospace fonts have substantially different glyphs for I / I and 0 / O. fonts.adobe.com/fonts/source-c Even the non-monospace Source Sans Pro has a slashed zero OpenType option and l is visibly different but nearly as much as it is for the monospace one meant for programming.

Replying to

and

I would tend to consider monospace conventionally a third category separate from serif/sans-serif

Replying to

and

fonts.adobe.com/fonts/source-s is the non-monospace sans serif variation and you can see l still looks visibly different than I. It's too bad slashed zero is opt-in for the proportional variation of the font. Not really sure why someone would prefer having that be less distinct.

Replying to

and

otf.show/zero?font=Sour

9:42 PM · Jan 6, 2022Twitter Web App