On Tue, Mar 10, 2026 at 01:02:13PM +0700, hahahahacker2009 wrote: > Vào CN, 8 thg 3, 2026 va??o lu??c 19:14 Crystal Kolipe > <[email protected]> ??ã vi???t: > > > > On Sun, Mar 08, 2026 at 12:49:30PM +0100, Peter N. M. Hansteen wrote: > > > On Sun, Mar 08, 2026 at 11:44:15AM +0000, hmjsp wrote: > > > > disable ntpd? why? > > > > > > See https://marc.info/?l=openbsd-bugs&m=177296357231841&w=2 > > > > Just to make it clear to anyone reading the archives in the future, the > > suggestion to disable ntpd was a joke and a form of irony. > > > > Unfortunately the original two messages were posted to different lists, > > (-bugs > > and -misc), so it's entirely possible that this could be missed by casual > > readers of just one list. > > > > There is _no serious suggestion_ to disable ntpd. > > > > You are their friend and have met them face to face? > > I will provide some context about the suggestion to disable NTP.
You are posting to an OpenBSD mailing list. Within the OpenBSD development community, there is no serious suggestion to disable ntpd. There is widespread agreement on this. Accurate timekeeping is important for various reasons, which is why ntpd is one of the few services enabled by default. If somebody came up with a genuine reason and demonstrable use-case for disabling it, that might have the potential to be an interesting discussion. On the other hand, general waffle and FUD based on random opinions with different contexts, limited knowledge, outdated knowledge, or things of this nature is _not welcome_. The -misc mailing list is receiving a considerable amount of rubbish these days, a lot of which is coming from people who clearly do not understand the security models and concepts that OpenBSD implements. This is putting a burden on the people who write the code, and reducing the utility of the mailing list in general, because the useful information becomes diluted. Suggestions and commentary about 'security' from outside the OpenBSD development community are often based on different threat models and different implementations of protocols. What works in one place might be completely irrelevant or counter-productive in OpenBSD. Furthermore, a lot of the commentary on the wider internet is based on limited, incorrect, outdated, or incomplete knowledge of the issues and yet still passed off as being complete and authoritative. Asking questions and educating people is useful. Hand waving and making loud and outrageous claims that OpenBSD or it's contributors are getting something wrong, (without providing a comprehensive write-up of why you think that is the case), is not useful or welcome. To give you a specific example, you quoted: " Instead of NTP, you can connect to a trusted website over a secure connection " [ ... ] " and extract the current time from the HTTP header " That completely ignores the fact that accurate timekeeping is _not_ just about 'setting the clock'. Ntpd does far more than just set the system clock. It analyses the rate of drift and adjusts the clock rate of the local system to _maintain_ accurate time. So even if your intentions are good, by encouraging people to do this instead of using the default ntpd setup, you would be reducing the accuracy of their system clock.
