Live data from Hacker News

Please donate to keep Network Time Protocol up – Goal 1k

ntp.org

161–170 of 242 posts

Re: Please donate to keep Network Time Protocol up – Goal 1k

#161
post #78

Rather than money one can donate NTP servers to the pool. [1] It can be a fun learning exercise in setting up a stable stratum-2 time server. One can create graphs from the optional logs. Why bother? Many of the rabbit holes one could venture down in learning to set up a stable time server can also benefit application servers in terms of latency, responsiveness, learning how to get clients to share resources and so m…

One of the really nifty things about having a stratum-1 time server on-site (because... reasons) is those graphs. You can very readily see the subtle temperature-dependence of timing crystals. At the facility I was at there was a large cycle every day during the week and then smaller cycle on each weekend day. Our HVAC system didn't heat/cool the building as much on the weekend when no one was there so the temperature swing -> frequency swing was smaller.

Really drives home one of my favourite half-jokes: every sensor is a temperature sensor; some of them measure other things too.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#162
post #78

Rather than money one can donate NTP servers to the pool. [1] It can be a fun learning exercise in setting up a stable stratum-2 time server. One can create graphs from the optional logs. Why bother? Many of the rabbit holes one could venture down in learning to set up a stable time server can also benefit application servers in terms of latency, responsiveness, learning how to get clients to share resources and so m…

We ran a public NTP server for many years. Then, details hazy, but I think there was a UDP amplification vulnerability that was exploited which upset our transit provider so we took it down. Might be fun to try again though.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#163

This was submitted with a title that doesn’t match the page and is not even accurste (Please donate to keep Network Time Protocol up) is not correct. This donation page is not for the public NTP pool. It’s for the NTP Project organization and their web page. All of the angry comments from people who think NTP will stop working if the donation bar doesn’t get to $1000 are misinformed. Also note that the bar isn’t upda…

I see it at $2,130 fifteen minutes after your comment.

Yes, it’s updating now. Ctrl-F this thread and search for ‘365’ and you’ll see that it was not changing for several hours despite donations rolling in.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#164

Earlier quoted context omitted.

Those trillion dollar companies are running a lot of public NTP server infrastructure. Here is a list of public NTP addresses from different big tech companies: https://gist.github.com/mutin-sa/eea1c396b1e610a2da1e5550d94... > Trillion dollar companies depend leech on it Are you confusing the NTP Foundation (the group asking for donations) with NTP the protocol or the NTP software itself? This donation request isn’t…

It would hurt the big companies though because the employees of those big companies rely on Ntp. It may not directly impact them but it’s better than letting them continue to rake in billions and then never support the core foundational tech. Maybe letting Ntp fail will wake up some of the employees of other companies to the absolute sad state of the software world.

> Maybe letting Ntp fail will wake up some of the employees of other companies to the absolute sad state of the software world.

Big companies run their own NTP servers (which you can use for free) and do not use the reference implantation.

There is nothing to “fail” in this project which would cause big companies to have infrastructure problems.

The saddest state of the software world is that some people here have convinced themselves to cheer for this project to fail because they don’t understand that big companies do not depend on this project that is asking for donations.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#165
post #92

Earlier quoted context omitted.

Who's running ntpsec?

The major Linux distributions replaced ntp with ntpsec. A better question would be who is still running ntp. I know about FreeBSD and NetBSD. ntpsec as a project seems to be doing ok. They are releasing new versions, fix reported issues, accept patches, and develop the code publicly. While ntp still has a huge list of acknowledged but unfixed CVEs.

Which distributions use ntpsec?

Re: Please donate to keep Network Time Protocol up – Goal 1k

#166

Earlier quoted context omitted.

It would hurt the big companies though because the employees of those big companies rely on Ntp. It may not directly impact them but it’s better than letting them continue to rake in billions and then never support the core foundational tech. Maybe letting Ntp fail will wake up some of the employees of other companies to the absolute sad state of the software world.

> Maybe letting Ntp fail will wake up some of the employees of other companies to the absolute sad state of the software world. Big companies run their own NTP servers (which you can use for free) and do not use the reference implantation. There is nothing to “fail” in this project which would cause big companies to have infrastructure problems. The saddest state of the software world is that some people here have co…

From the previous comment.

> Google’s, for example, uses time-smearing to handle leaps. This is different than the standard and therefore you shouldn’t mix Google’s leap-smearing NTP system with NTP servers that don’t leap smear.

So we can use or not?

If we can then well good, then there will be no problem then if the funding fails.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#167

Earlier quoted context omitted.

I see it at $2,130 fifteen minutes after your comment.

Yes, it’s updating now. Ctrl-F this thread and search for ‘365’ and you’ll see that it was not changing for several hours despite donations rolling in.

I suspect someone is updating that bar by hand :V

Re: Please donate to keep Network Time Protocol up – Goal 1k

#168

Earlier quoted context omitted.

> Maybe letting Ntp fail will wake up some of the employees of other companies to the absolute sad state of the software world. Big companies run their own NTP servers (which you can use for free) and do not use the reference implantation. There is nothing to “fail” in this project which would cause big companies to have infrastructure problems. The saddest state of the software world is that some people here have co…

From the previous comment. > Google’s, for example, uses time-smearing to handle leaps. This is different than the standard and therefore you shouldn’t mix Google’s leap-smearing NTP system with NTP servers that don’t leap smear. So we can use or not? If we can then well good, then there will be no problem then if the funding fails.

> If we can then well good, then there will be no problem then if the funding fails

The funding is not for the NTP servers or pools. Please read the actual page and all of the comments trying to explain that the HN title is a lie. NTP will not “go down” if this project fails.

You can use Google or Amazon’s NTP servers if you’d like. Just be aware of how they represent leap seconds differently.

Re: Please donate to keep Network Time Protocol up – Goal 1k

#169

Earlier quoted context omitted.

Yes, it’s updating now. Ctrl-F this thread and search for ‘365’ and you’ll see that it was not changing for several hours despite donations rolling in.

I suspect someone is updating that bar by hand :V

Probably yes, it's still at 2130

Re: Please donate to keep Network Time Protocol up – Goal 1k

#170
post #89

Earlier quoted context omitted.

Even a single highly paid employee of a FAANG, many of whom frequent this very forum, could pay for a year without breaking the bank.

They would rather downvote you than giving away $100

Because it is factually wrong, which would have taken seconds to discover.
Post reply on HN