avatar/pfp: Alixesque
banner: Mermaid Elizabeth

 

• high school dropout (proud)
• professional computer toucher (remorseful)
• my organic brain was replaced with a NEC V810 when i was 8
• love and kindness 'til we die, baby

 

THIS WEBSITE IS FREE BUT IF YOU LIKE IT YOU CAN PAY FOR IT gif: @westfailia


what are the actual odds of an md5 collision occurring in two sequences of data of identical length? I'm wondering how to detect duplicate bitmaps in this comic chat toolkit without needing to actually compare every last byte every time (and keep that data buffered)

I assume it's "pretty rare" but I am acknowledging here that you absolutely could mess this library up on purpose if you want to, but like. don't do that then I guess? you'd just mess up your own files.

... but I am wondering if anyone could really ever conceivably trip a collision by accident ...


You must log in to comment.

in reply to @nago-'s post:

MD5 is the most commonly used and when compared to the other two hash functions, it represents the middle ground in terms of hash collision risk. In order to get a 50% chance of a hash collision occurring, there would have to be over 5.06 billion records in the hub.[11]

it seems pretty unlikely unless you're engineering an attack