3 ms·
> In the talk about git in 2007, Linus said something about git being secure — that if someone would attempt to alter git history (as in attack that happened at
by lawl 13y ago
> In the talk about git in 2007, Linus said something about git being secure — that if someone would attempt to alter git history (as in attack that happened at Linux repo before), it would be instantly noticed because of hashes.
> Does this mean that compromising someone's git repo without anyone noticing can actually be done?
No. Brute forcing 4 bytes of the commit hash is something different than brute forcing 20 bytes. Because you know, exponentials.
By the way. That's pretty much what you're doing when you're mining bitcoin.
Edit: added parents comment, because deleted and i see no reason to delete this, others might ask the same question.
- golergka 13y agoThanks. I guess I just got so accustomed to working with shortened commit hashes in source control that I almost forgot that they have full versions.
- rakoo 13y agoOn a similar topic, this is the reason why some people urge gpg users to use long key ids (8 bytes) rather than short key ids (4 bytes) (see http://www.asheesh.org/note/debian/short-key-ids-are-bad-news.html http://www.asheesh.org/note/debian/short-key-ids-are-bad-new... for example)