Timeline for If and why is it bad to scrypt a bcrypted password?
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 28, 2018 at 20:52 | comment | added | Squeamish Ossifrage | Collision resistance is not relevant to password hashing. | |
| May 28, 2018 at 16:43 | comment | added | VincBreaker | @gaazkam That is only true when you take pre-image-attacks into account, but when you are able to create collisions for bcrypt, you are as well able to create a collision for scrypt(brcypt) (or the other way arround) so you actually decrease collision resistance. | |
| May 28, 2018 at 9:21 | comment | added | gaazkam | This is, of course, a very good answer, have my upvote. However, it still fails to address the primary argument for stacking scrypt with bcrypt: Suppose that either is cracked, you still have the other one to protect you. However, if you stick to either and that one is cracked, you're screwed. | |
| May 27, 2018 at 23:18 | comment | added | Maarten Bodewes♦ | Agreed. The only reason to use bcrypt is to decrease the amount of cache misses I suppose. If you're going to spend RAM and memory access to scrypt anyway, you might as well stick to it. | |
| May 27, 2018 at 22:06 | history | answered | Future Security | CC BY-SA 4.0 |