Message 403400 - Python tracker

Message403400

Author	christian.heimes
Recipients	Mark.Shannon, christian.heimes, erlendaasland, gvanrossum, lemburg, methane, rhettinger, serhiy.storchaka, vstinner
Date	2021-10-07.12:23:08
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1633609388.8.0.629366471255.issue29410@roundup.psfhosted.org>
In-reply-to

Content
SipHash is a cryptographically secure PRF, not a cryptographic hashing algorithm, https://www.python.org/dev/peps/pep-0456/#siphash I'm strongly against using a different algorithm when the rest of the world is using either SipHash-2-4 or SipHash-1-3. A minuscule performance gain is not worth the risk. If users like to use an alternative algorithm, then they are free to do so. PEP 456 gives them a hook point to override the algorithm on build time of CPython.

Content

SipHash is a cryptographically secure PRF, not a cryptographic hashing algorithm, https://www.python.org/dev/peps/pep-0456/#siphash

I'm strongly against using a different algorithm when the rest of the world is using either SipHash-2-4 or SipHash-1-3. A minuscule performance gain is not worth the risk.

If users like to use an alternative algorithm, then they are free to do so. PEP 456 gives them a hook point to override the algorithm on build time of CPython.

History
Date	User	Action	Args
2021-10-07 12:23:08	christian.heimes	set	recipients: + christian.heimes, lemburg, gvanrossum, rhettinger, vstinner, methane, Mark.Shannon, serhiy.storchaka, erlendaasland
2021-10-07 12:23:08	christian.heimes	set	messageid: <1633609388.8.0.629366471255.issue29410@roundup.psfhosted.org>
2021-10-07 12:23:08	christian.heimes	link	issue29410 messages
2021-10-07 12:23:08	christian.heimes	create