Hacker News new | more | comments | ask | show | jobs | submit login
How the Government Shutdown Warps Time Itself (typesandtimes.net)
70 points by psuter 35 days ago | hide | past | web | favorite | 69 comments

Why don't they use the Canadian National Research Council bulletin?


In Canada if a budget isn't passed it triggers an election. We don't have government shutdowns.

I like this idea a lot. I think the US, we like elections to come in certain cycles to give the incumbents enough time to actually accomplish something, but I have always admired the parliamentary system where you can call an election at various points.

I think the US Constitution either explicitly precludes this kind of system or could be construed to preclude it.

It’s been abused a lot to call elections when advantageous for the incumbent. A decade old law made for more regular term durations but there are still loopholes for the ruling party to unilaterally declare elections.

Failure to pass a budget is an automatic no confidence vote still which I think works very well.

An other method to get a budget passed is to allow any representative to submit a budget and let which ever budget has most votes pass. That is what Sweden use.

Portugal just uses the budget from the last year unchanged.

If you do, you probably should use https.


There's probably some really weird MITM attack possibilities here if people have automated systems fetching and using these files. Also, following the link from the article, it would seem the US government's official source is unencrypted FTP?

Yes very true. In fact the NRC should probably set up an HTTPS redirect and HSTS.

No need. House seats are up for election every two years for this reason.

I guess the US government will reopen in 2020 then.

Such an informative post; thanks for sharing this!

Also, and somewhat sad and just plain crazy: some functions of NIST are not considered essential!?! It seems to me someone in the gov. is not clearly reporting who and what functions are truly essential to the higher ups in the gov. Then again, maybe they have, but the higher ups ignore them. sigh

Nevertheless, cool post.

*could possibly affect an edge case of utc leap second distribution if Another shutdown happens during a leap second year (this is not one), AND lasts over 6 months. I get that it acks the clickbaity title at the beginning, but that coy bs doesnt help.

> Because TAI is not based on the Earth’s rotation, it’s not ever-so-slowly changing. It’s the measure of time against which UTC’s watch is occasionally correct. That correction is called a leap second: a 61st second that is sometimes added to particular minutes in UTC, like the very last minute of December 31, 2016. As of January 2019 there have been 27 such leap seconds inserted.

This seems backwards. Perhaps I'm misunderstanding what the author is trying to say, and another poster could help clarify?

If UTC is based on the earth's rotation and TAI is based on cesium-atomic timekeeping, and UTC is slowing down, and TAI is "the measure of time against which UTC's watch is occasionally corrected," then how can UTC have leap seconds? Delaying for an additional second on December 31st would only make it one second even slower, exacerbating the problem.

I don't see how adding a delay to the slower clock is going to put it in sync with the faster one.

Is it instead that UTC is TAI, plus leap seconds to slow it down to keep it in sync with the rotation of the Earth?

It's the other way around - UTC is based on TAI plus an integer number of seconds offset, and this offset is adjusted (by the insertion of leap seconds) to keep UTC roughly in sync with the Earth's rotation.

The delay is added to make UTC slow. The Earth is the slow clock, UTC ticks at the same pace as TAI, so sometimes we add a leap second to UTC to sync it with the Earth.

The license stuff seems weird without further context: even if they can't copy that file, what's the hold up to just "encode" the information from the bulletin on their own and putting it into their file?

Wouldn't that make it a derivative work, legally speaking?

> Wouldn't that make it a derivative work, legally speaking?

IANAL, but:

I think not, facts are not under copyright. If you only copy facts (and avoid copying style), then you are not creating derivative work.

Furthermore, works created by federal employees in the course of their duties are not eligible for copyright in the U.S. [0] The U.S. does not have an equivalent of crown copyright. The U.S. government can only own copyrights transferred to them by private individuals or organizations, or state and local governments which may or may not hold copyrights on their works.

[0]: https://www.usa.gov/government-works

We still can't migrate to the metric system. Let's work on that first!

I don't want to be the person with a metric house in a country of imperial builders.

But I digress it's 80 past 45 and still haven't had breakfast :-)

So if I am reading this right people who use NTP aren't affected. Is that right or does OpenNTPD depends on tzdata ?

Leap seconds are the most stupidest thing ever. It's time to abandon UTC and start using what professional astronomers use: UT1

UT1 is a disaster for many purposes. It is, quite simply, a measure of the Earth's current angle in space. This is ideal for pointing an Earth-based telescope. Trouble is, we can't reliably predict how fast this will tick. You may think leap seconds are bad, but how about if you simply can't know the current time without having a current measurement of the position of quasars in the sky? The length of a second isn't even constant.

TAI matches up with the atomic second, and UTC is that plus enough leap seconds to approximate UT1.

It all sucks, but this is reality. You can't fix reality. You'd have to tell the Earth to obey an atomic clock.

> It all sucks, but this is reality. You can't fix reality. You'd have to tell the Earth to obey an atomic clock.

The question might be super naive, but I wonder how useful/workable a time-system would be that's completely decoupled from celestial body movements? Isn't Swatch Internet Time something like that?

Gonna be really interesting once humanity becomes interplanetary or even interstellar. Will we have whole different calendars and time zones for different planets and solar systems?

You'd imagine an internet time (or civilisation time) used for timekeeping in interplanetary protocols, synced (relatively) well to Earth- or wherever the centre of human Civilisation ends up being. Because clock speed changes due to relativity, your Civilisation Time counter would have to take into account your position and velocity relative to earth to not slip- while syncing often enough to counteract This way you can have relatively stable time measurement, even when time strictly speaking passes faster for some computers than others.

Time slows down as you go faster. The people leaving earth are the ones accelerating, so they're the ones who travel faster into the future- less time passes for them. If we colonise stars around us, the people going to the further stars (assume all ships leave from earth) will lose more time: Civilisation Time will be ahead of the time they measure on their regular clocks.

Maybe we could use pulsars as clocks? I think they pulse pretty regularly.

I think you've wrapped your head the wrong way around this. The stupidest thing ever is the idea that a country's government can be shut down as an extortion method. It's a concept so alien to me that I cannot fathom the logic of the person who came up with the idea.

Nobody "came up with" the idea. It's the consequence of a principle of the American Constitution, wherein only Congress can appropriate money. This is a bedrock principle, because it provides a check on the Executive: The President can yell all he wants, but unless Congress appropriates money to pay for whatever it is he is yelling about (in this case, a border wall), he doesn't get it.

For observers outside the US, the idea of "the national government shutting down" probably seems way way more alarming that it does to most Americans. Almost all basic services in the US are handled by states and municipalities. This includes garbage collection, policing, firefighters, schools, local tax collection, road maintenance, the majority of prisons, most courts, most water and power infrastructure, building inspections and permitting, and many more.

Even airports, the site of a lot of hand-wringing during the shutdown due to the TSA (security screeners) and air-traffic controllers being federal employees, are usually (always?) municipally or state-owned. I have no idea of the legality of these municipalities or the airlines themselves stepping in to pay the ATCs or screeners, but it doesn't seem totally bonkers.

Thanks for elaborating.

The Swedish tradition is that you cannot vote to stop a budget, you can only vote for another proposal. That is, if there is only one proposal, you only allow 'yes' or 'abstain' votes, and if there are several, you do a series of votes to eliminate the lesser proposals until only one stands, and then only allow 'yes' or 'abstain' again.

I hope that explains why I find it so strange.

Another peculiarity we use to avoid stand-offs (it has always worked up until last election) is negative parliamentarism, where the prime minister is tolerated, not elected by the parliament. What it means is that a majority of MPs have to vote "no" to not elect a prime minister, which means that a lot of parties vote "abstain" (we had such a vote yesterday where the PM was tolerated with the numbers yes: 115, no: 153, abstain: 77, absent:4). This has historically meant that minority governments have been able to flourish.

Thanks for explaining this to us non-USians.

So is the current problem is because the president can veto congresses appropriations? In a functioning government this seems sensible so the executive has a symmetric check on congress, but currently it gives the president unwarranted leverage.

Congress can override a Presidential veto, but it requires a 2/3 majority, which neither party has by itself (due to parliamentary rules, it's slightly more complicated than that in ways that I don't totally understand, but that's the high-school-civics-version). The President isn't actually necessary to reopen the government: Congressional Democrats and Republicans could come to a deal that got them to the 2/3 mark if they wanted to, even over the President's objections. Whether or not that will happen in this cases is anyone's guess.

The problem is threefold:

1) it's not a true shutdown because the military still get paid. There is a builtin rightwing fudge of what gets deemed "essential".

2) Most governments have a similar possibility of not being able to pass a budget; but, in the Westminster system, this forces fresh elections. The people get to decide on what compromise is necessary to resume business.

3) Businesses should not make their employees into creditors. Failing to make payroll is usually a sign that a business is either unethical or nearly bankrupt. This screws over people badly, e.g. https://twitter.com/shawndgoldman/status/1085733207469891585

"The NASA Postdoctoral Program contract runs out of money tomorrow. This means its fellows - some of the best and brightest space scientists in the world - will go without pay starting Friday. These folks are contractors, which means they aren't covered by the backpay bill."

(There was also some suggestion that not being paid would automatically terminate their J-1 visas, but I can't find that now)

>The problem is threefold:

>1) it's not a true shutdown because the military still get paid. There is a builtin rightwing fudge of what gets deemed "essential".

The military still gets paid because congress was able to agree to pay for that. It's certainly possible, for congress to be unable to agree to a DoD budget. Separating the budget process into a series of sub-budgets allows for continuity in areas where there's room for consensus.

> 2) Most governments have a similar possibility of not being able to pass a budget; but, in the Westminster system, this forces fresh elections. The people get to decide on what compromise is necessary to resume business.

Presumably, the government is shutdown until the votes are certified? And how long do the newly elected representatives have to form a budget, before a second election would be triggered? Regardless, for better or worse, the US system does not allow for snap elections, and that would be a big change, especially given the design of different term lengths for different offices.

> 3) Businesses should not make their employees into creditors.

Yes, this is pretty awful. I would be raising some hell if my employer expected me to work without paying me contemporaneously.

>Presumably, the government is shutdown until the votes are certified?

This can be solved by setting the budget deadline some time before the expiration of the past budget. You have until July 1 to pass a budget for 2020; if there isn't one, we hold elections in August for your replacement, giving them several months to pull together something workable.

Presumably, the government is shutdown until the votes are certified?

No. I'm not an expert in the details, but there are rules about how government services continue to operate during the period that Parliament is dissolved. The key principle, however, is that it continues to operate.

Most of the military is still being paid because the DoD isn’t one of the unfounded agencies. The coast guard, however, is currently working without pay, because they’re part of commerce.

What has been the best argument for it that you've found?

Leap seconds is just the time abstraction being leaky: the Earth simply doesn’t have a constant rotation speed.

Certainly seems more practical to correct for this once in a while rather than having a variable length second.

We could easily define a time system that is simply "number of seconds since 00:00:00 January 1, 1970" (other than epoch being different, we have this with GPS and TAI time, but the network effects for UTC are strong) and used that as our base time system. Leap seconds could then be added at higher layers simmilar to how leap hours are. Instead evryone has to deal with leap seconds weather they care about the orientation of Earth or not.

Also, at least for computer applications, a variable lenfth second is easier to deal with than a leap second.

Yes, you can easily create a constant rate counter, but no: it wouln't keep time.

To keep time you do care about the position of the sun because that defines the time of day, humanity's base for the experience of time.

If you will never care about what the time is in your application, by all means, count clock cycles, otherwise: UTC is the standard for keeping time.

That brings up an interesting question: Is that the current, best, or only basis for time?

I'd argue that many relevant time concepts are more framed by social conventions than the sun.

Most of us are more interested in concepts like "when does work start", "lunch time", and "when the store closes", than "when does sunrise occur" these days. In a pre-industrial society, those might be closely synched, but less so today.

Those are important events in the day, but they don't serve as a basis for defining time.

When does work start? Well, it starts when the hands on the clock points just so, there really isn't any more fundamental definition. But clocks are tools for measuring time, they don't define it.

If your clock slows nothing particular happens to other clocks, and time marches on unslowed. But sunset will bring night whether you notice it or not, and that's a pretty definite marker of the passage of time.

Time is a well defined physical property that has nothing to do with the posistion of the sun.

We already have time localizations, time zones, daylight savings time, and leap days all handled by layers above UTC (actually, all in the timezone layer). Leap seconds should be part of that layer, not the UTC layer.

Once we get into relativistic effects, there is more complication to the lower layers, but we (mostly) don't need to deal with that.

Just like the amount of mass in a gram doesn't change based on location, time, or any other variable it's very important to have some measurement of time that is as constant and precise as possible. This helps us conduct science, where many types of experiments take place in less than a second. The measurements taken today should be readable decades or centuries from now without having to adjust for the slowing of Earth's rotation.

Now, maybe this unit of time needed for scientific purposes doesn't need to be the same one used for daily life.

> maybe this unit of time needed for scientific purposes doesn't need to be the same one used for daily life.

Don't think the unit is the problem, it's the offset. And when it comes to that, we already have a virtualization layer which assures alignment with the sun for social purposes: time zones. No need to add another one over TAI.

Except that’s not entirely the story of time zones. They’re political divisions. Some of them match up with the sun better than others. For instance, the sunset time gets pretty weird in LV over the winter.

> They’re political divisions

Which I think is perfectly fine (if not necessary) for that layer. The human body doesn't care for extra seconds, and science doesn't either. So who do we do the "coordination" in UTC for? If - over hundreds of years - the time zone drifts off into weirdness, it can be changed politically. What more do you need?

Oh. In UT1 there number of seconds in a day is constant, so it's value is defined by the rotation of the earth which is slowing. UTC is better than UT1 in that regard. Maybe UT1 is fine for humans, but really humans just assume ut1.

Time zones need to be defined relative to some reference time, and UTC is (sort of) the least impractical reference.

UTC adds little of value on top of TAI, except some headache for clock synchronization. So I argue for basing timezones directly on TAI, as I see TAI as the least impractical reference.

I beg to differ, Dst is the stupidest time related thing ever. From a software development perspective it is a tremendous pain in the ass and is stupid. From the non software aspect it is just stupid and annoying.

I submit the Japanese imperial succession and the necessary changes to traditional calendars as another contender...

Leap seconds are a lot worse for software development, because Unix time does not include DST adjustments, but does include leap second adjustments. This means that if you use unix time in an application, you can just use a library at the input/output edge and not worry about it, but leap seconds can screw up anywhere that you use a timestamp.

From a non-software perspective it kills people.

Also, if you're a parent, it's not great for sleep training children. It just messes with everyone.

Can you elaborate on how it kills people??

Traffic deaths spike following the DST starting (and to a lesser extent, ending) date- thought to be due to disrupted sleep schedules.


There is apparently an increase in heart-attacks right after the transition, attributed by some to the loss of an hours sleep...


You mean TAI, not UT1.

In my opinion it's possible the Democrats will cave. The strategic situation is that Democrats care about the country, and Trump does not. However, not enough Americans are convinced of the latter fact. Once they are, the only rational decision is to cave - and then remove him from office as quickly as possible.

The logical play is that any wall is a long-running project. They can promise Trump anything to get him to sign the budget, and then immediately unwind their efforts once the smoke clears.

I honestly want to see them push a budget with something like "Canadian style health care plus a wall" to see if he bites.

I don't think they could negotiate something as big as a completely new health system in time to stop government services from collapsing...

> The strategic situation is that Democrats care about the country, and Trump does not. However, not enough Americans are convinced of the latter fact.

I'm not an American, just an immigrant. Historically this isn't about caring, it's just about politics.

Normally it really isn't about caring for the overall good of country. Politicians competing for voters all care about their country, or at the very least have to act as if.

In this case, it is different, because Trump cares more about his "win" in this issue than about the country. Or even about winning reelection, because everything is pointing to public opinion being strongly against what he is doing right now.

However, the nation is hurting badly under the shutdown, and it is going to get worse. Even so Trump is blamed by a majority, and thus "politically" losing more, at some point the democrats will have to concede that the only way to stop the damage is to give in.

But by then, there will be no doubt in most voters' minds that Trump fundamentally doesn't care. And this makes him unelectable for a larger majority and may even pierce his senate-armor against impeachment.

> at some point the democrats will have to concede that the only way to stop the damage is to give in.

Which would be the point where your 'they care more' hypothesis would become true.

So are they just wanting to hurt him politically right now?

Right now they think they are playing the game "who loses more, politically", in which the president is losing bigly.

Or at least they act as if they believe the president cares about his reelection chances. But as more and more time passes, and he loses more and more support, it will become obvious to even some of his supporters, that he will burn down the country to get his way. And at that point he may get his wall, but he loses everything else.

I don't care about him all that much. As you imply, he'll do whatever he wants to get whatever he wants.

I care about things moving forward. And if we assume that he's not about that, then we need to start moving our gaze to those who are the actors that should be acting correctly and having things move forward. They're not doing that.

I'm not an American. I think your constitution basically sucks, with the mitigating factor that it was very practical and relatively democratic compared to other states at the time it was written.

One of the huge criticisms that I have is that you can't get rid of an administration. Impeachment seems to be virtually impossible, whatever Trump does or did. And even then you get a vice president who ran on the same ticket and profited from the same criminal behavior.

In Germany, we can get rid of the chancellor within days or weeks, depending on the procedure. We also can't have split governments sabotaging everything. Every vote counts the same etc.

I have to wonder, if something like VLC can exist, which blatantly flies in the face of copywright law unchallenged, why do so many open-source projects get ridiculously anal about license issues? The problem here appears to be that the license (CC-BY 4.0) is only specified in the README file, not in the timezone data file itself. Like, do they really think that it only applies to the README, and if they use the data they're going to get sued? I get that it's important to cover one's ass, but if it gets to the point where software breaks as a result, it's gone much too far.

I wonder if the right solution here is to have some kind of body provide insurance against frivolous litigation over licensing issues for open-source projects, so that this kind of stupidity doesn't arise.

Edit: It appears I was mistaken about the legal status of VLC: it is about software patent licenses, not software copywright licenses.

VLC is not American so doesn’t have to care about American software patents. Licenses are different from patents.

>which blatantly flies in the face of copywright law unchallenged

How do you figure? Patent law isn't copyright law, and doesn't apply equally around the world.

>why do so many open-source projects get ridiculously anal about license issues

Because it's their code and people are using it without compensation and in ways not amenable to them? I think you'd be pretty "anal", too, if someone took your software that you worked hard on, without wanting to contribute their improvements and without any other form of compensation.

x264 basically funds VLC. Don't steal other people's code.

Stealing code isn't at issue here, the scenario is that developers are not incorporating new timezone data because the open-source license for that data is only printed the README, not in the timezone data file itself. This is the sort of thing I'm calling anal.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact