P2058 Make std::random_device Less Inscrutable

[wg21bot]

P2058R0 Make std::random_device Less Inscrutable (Martin Hořeňovský)

[brycelelbach]

Prague 2020-02 LEWGI Minutes

P2058R0 Improve std::random_device

Chair: Bryce Adelstein Lelbach

Champion: Martin Hořeňovský

Minute Taker: Peter Bindels

Start Review: 2020-02-10 11:33

I'm not sure what decisions you would make based on the return value of is_predictable. It'd be nice to see some examples of how this is used.

Is std::random_device ever suitable for use in applications that need secure random numbers?

Authors should get Walter's input on this.

Needs a better name.

This probably has to be VERY conservative if it's constexpr; what if you're running on an older OS kernel or older hardware that has bugs?

Start Polling: 11:49

POLL: We should promise more committee time to pursuing P2058R0 (make std::random_device less inscrutable), knowing that our time is scarce and this will leave less time for other work.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
3	6	5	0	0

Attendance: 15

# of Authors: 1

Author Position: SF

That has consensus.

POLL: Deprecate std::random_device::entropy even if we provide no replacement or alternative.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
2	1	4	1	0

Attendance: 15

# of Authors: 1

Author Position: SF

That has no consensus.

POLL: Knowing what we know right now, implementations should only have std::random_device::is_predictable be false if std::random_device is cryptographically secure.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
3	1	1	3	1

Attendance: 15

# of Authors: 1

Author Position: SF

That has no consensus.

End: 12:05

CONSENSUS: Further discussion in LEWGI is needed this week.

[brycelelbach]

Prague 2020-02 LEWGI Minutes

P2058R0 Improve std::random_device

Chair: Nevin Liber

Champion: Martin Hořeňovský

Minute Taker: Marco Foco

Start Review: 2020-02-11 13:38

This probably has to be VERY conservative if it's constexpr; what if you're running on an older OS kernel or older hardware that has bugs?

Start Polling: 13:56

POLL: Deprecate std::random_device::entropy.

Strongly For	Weakly For	Neutral	Weakly Against	Strongly Against
3	6	5	0	0

Attendance: 19

# of Authors: 1

Author Position: SF

That has unanimous consent.

End: 13:56

CONSENSUS: We will not pursue P2058R0 (Improve std::random_device), but bring a proposal for deprecating std::random_device::entropy.