Time to party like it's 2002: Acura and Honda car clocks knocked back 20 years by bug

Owners of older Acura and Honda vehicles marked the new year by revisiting 2002, a consequence of a bug affecting the cars' clock software. With the onset of 2022, Acura and Honda customers began reporting on various online forums that their dashboard display clocks reset to January 1, 2002 when they restarted their vehicles …

[b0llchit]

Always date and time

There are always problems with date and time in computers. It is one of the most difficult things to get right. And when you get it right, you probably got it wrong anyway, somewhere.

More errors pop up with more fancy stuff you want. And then, not appreciating the date/time complexity is the real mother. Also, management does not appreciate the complexity of date/time and cannot fathom why it takes so long to program that damned clock, which results in forced coding shortcuts. Not taking the junior programmer into account, that is.

The company not acknowledging the problem, despite ample evidence, is a problem in and of itself. Then, after the 7 month self-repair timeout, you will get a car that skips the third gear. August in 2022 is the month where it has been determined that the third gear is no longer necessary and it is so programmed into the car software. After 5 months you will get the third gear back while Mondays will skip second gear on odd seconds. Management has declined any comment on that feature.

[W.S.Gosset]

Re: Always date and time

> There are always problems with date and time in computers. It is one of the most difficult things to get right.

Oh yeah. At both micro and macro levels.

Macro-level recommendation: for datetime-range slicing&dicing/analysis (or doing so with extreme performance reqts), I can recommend kdb+. Time&Date tools which were just very sane, simple, yet correct, which were actually a pleasure to use. As I've posted previously: extraordinarily sweet (dis) aggregation esp.re time periods

.

_{Also seen while searching for that comment here: Conway's Game-of-Life in k (kdb+ language):}

life:{3=a-x*4=a:2{+x-2{|1_0+':x,1#x}/x}/x}

_{Absolutely nothing to do with date-time slicing. I just boggled just now, is all.}

[druck]

Re: Always date and time

There are always problems with date and time in computers. It is one of the most difficult things to get right. And when you get it right, you probably got it wrong anyway, somewhere.

Very true. I knocked up a quick Python script to make a graph from an rsyslog file using datetime.util.parse to turn the month and day in to a full date, and thought was working perfectly until I looked at the graph on new years day. Instead of showing the last 2 days data, it showed a year with data from 1st Jan 2022 and 31st Dec 2022. Quickly changed it to use datetime.strptime with a calculation of the correct year.

[John Brown (no body)]

Re: Always date and time

"Also, management does not appreciate the complexity of date/time and cannot fathom why it takes so long to program that damned clock, "

To make them aware of the complexity of time and date setting etc on technologically advanced hardware is easy. Just force them to use a 1980's VCR as their clock/calendar/reminder system for a week! The more up to date once were capable of setting daily and weekly recurring start/end times so great for regular meetings including durations.

[Stuart Castle]

Re: Always date and time

It's always date and time.. The fun and games I've had when writing code that has to plug in to different systems, a lot of which seem to have different date/time formats.. Be much easier if everyone stuck to one..

Hell, one of the servers I deal with now uses the English date format (or whatever one is selected in the interface) for all user facing pages, but on any programming interfaces (API and Inventory import CSV) requires the date to be in the form yyyy-mm-dd hh:mm:ss. I wouldn't mind that, but that fact is buried somewhere in one of the guides. It's not mentioned on the server, even in the error message you get if you get the date format wrong (it just says "Wrong format").

If I get time, might see if I can write a script that imports the CSV, corrects the date then imports the data to the server.. There is other information I'd like to upload that isn't supported by the standard CSV import mech.

[Mage]

Re: Always date and time

Stupidest in world is USA mm-dd-yy

Best is yyyy-mm-dd

[TRT]

Re: Always date and time

Are those yyyy digits Julian, Gregorian, Islamic, Hindu, Jewish, Zoroastrian or something else?

[Dante Alighieri]

ISO 8601

obligatory xkcd

[W.S.Gosset]

Re: Always date and time

> If I get time

Ah ha ha hahahahahaaaaaaa I see what you did there

[John Brown (no body)]

Re: Always date and time

"Hell, one of the servers I deal with now uses the English date format (or whatever one is selected in the interface) for all user facing pages, but on any programming interfaces (API and Inventory import CSV) requires the date to be in the form yyyy-mm-dd hh:mm:ss. I wouldn't mind that, but that fact is buried somewhere in one of the guides. It's not mentioned on the server, even in the error message you get if you get the date format wrong (it just says "Wrong format")."

Isn't that standard practice though? Use a "standard" internally then "localise" it for the users? It does rather sound as though the error trapping and reporting could be improved though.

[JimboSmith]

Re: Always date and time

Reminds me of when the MD (of the small firm I was employed by) was an hour late for work after the clocks went forward. He blamed his iPhone not updating itself and left an hour late to compensate. We said at the time it was unfortunate given how much he was paying a week for his iPhone & contract. He was forced to apologise to the member of staff he’d yelled at the previous week when they were late. The staff member in question had suffered a flat battery on their phone.

[Absolute Cynic]

Re: Always date and time

I turned up 1 hour early for a meeting. I had checked the time of the meeting the evening before. However I had checked while waiting for my flight back from Iceland - who are 1 hour behind the UK.

My "smart" phone had automatically converted the meeting time to the local time. Maybe it should include the time zone.

[mevets]

Re: Always date and time

Have you ever looked at room full of computer programmers/engineers/title-of-the day and thought to yourself .... " I wonder if any of them would know what to do with a date?"

So, hardly surprising.

[The Oncoming Scorn]

I wonder if any of them would know what to do with a date?

An actual physical social encounter with the gender of ones preference, the fruit or an algorithm?

[Anonymous Coward]

Re: I wonder if any of them would know what to do with a date?

A camel, some palm fruit, and the 29th of February 2001

[W.S.Gosset]

Re: Always date and time

This needs and requires a thundering herd of massive laughing upvotes.

[Eclectic Man]

Satellites?

How can this possibly be a problem with the satellites? Were that the case then every system relying on those satellites would have the same problem.

Maybe they used the same programmers as MicroSoft: https://www.theregister.com/2022/01/04/microsoft_date_comparison_tweet/

[David 132]

Re: Satellites?

Yep this is definitely a Honda-side problem.

To steal an XKCD joke, if I may: Epoch Fail!

[Eclectic Man]

Re: Satellites?

Upvoted for such a groanworthy play on words.

[David 132]

Re: Satellites?

You’re too kind. But give Randall Munroe the credit, not me.

[Dante Alighieri]

linky

obligatory xkcd

[Gene Cash]

Re: Satellites?

I think they're just using the Microsoft-style "baffle 'em with bullshit" strategy.

[Anonymous Coward]

Re: Satellites?

> I think they're just using the Microsoft-style "baffle 'em with bullshit" strategy.

That's one explanation. Another is: no IQ test needed to become a Honda dealer.

[Doctor Syntax]

At least I don't have that problem with my car clock. It just runs a little slow. Tens of £k for a car and they fit a clock that's less accurate than a cheap wristwatch.

[David 132]

I have to say that the digital clock in my 1993 Nissan farm truck keeps perfect time. Takes a licking and keeps on ticking, as the old advertising slogan had it.

[Eclectic Man]

Are you sure that it isn't that the world is wrong and your car has the 'correct' time?

[David 132]

Ah, the Wonko the Sane approach?

[TRT]

Well the observer's local time is relative to the speed with which one got there...

[short a sandwich]

You are Admiral Boom AICMFP

[W.S.Gosset]

> my car clock. It just runs a little slow.

That's not the clock, that's you; you're driving too close to c.

To ensure your clock runs the same speed as the restoftheworld's, you can either:

* drive slower, or

* move inland

[Anonymous Coward]

You have to be going exactly 55mph to go backwards in time

[Is It Me]

88mph I believe...

[Stuart Castle]

Maybe not if the software controlling the speedo was written by the same people who wrote the software controlling the clock.

[David 132]

What if you’re Sammy Hagar?

[W.S.Gosset]

> You have to be going exactly 55mph

In which direction, though?

Or are you suggesting relativity is scalar, not vector?

:D

(There's actually a rather large observation, in that.)

(I'll mention that the core insight of Einstein's original thrust derived from extrapolating to what I call a Zero Point.)

[ThatOne]

> they fit a clock that's less accurate than a cheap wristwatch

Unless your car's clock is some complicated networked job, I guess that, like for the cheap wristwatches, it all depends on the dreaded "cheap component lotto": If you're lucky your watch has exactly-on-spec components and is thus as precise as the technology allows. If not, it's way off specs and will tend to drift dramatically.

My car's clock is quite precise, it only drifts mere seconds a year, I guess I was lucky here. (Don't know other people with that same model though, so I can't really make comparisons.)

[Mage]

Wristwatch

Partly an adjusted wristwatch with metal back always can be more accurate than the same parts in a car because the wrist helps regulate the temperature.

[spold]

Customer disservice...

Yeah we will fix it in 7 months... meanwhile don't waste our time

[Skiron]

Re: Customer disservice...

I don't understand that comment from Honda at all.

Why 8 months (after 243 days)? That doesn't sort of fit any time related maths - maybe after 6 months, or if it was a leap year after Feb. 29th, yea, but after 7 months ?

[Pascal Monett]

Seems to be a wet finger estimate to me.

You know, you wet your finger, find out which way the wind is blowing now and announce a number.

Honestly, if Honda knows that the issue will resolve itself in 7 months, then they know what the issue is and can explain what the hell is going on.

Otherwise, they're just pulling numbers out from where the sun don't shine.

[gotes]

Seems about as likely as their claim that the problem is with the satellite.

I guess Honda could have launched their own sats to supply navigation features to vehicles, but it seems like a huge waste of resources considering there are already constellations in place which do the same thing and actually know what year it is.

[TRT]

The building supervisor where I used to work could tell you down to the minute how long to go before he retired. I suspect something along these lines, when it becomes someone else's problem.

[Anonymous Coward]

Re: but after 7 months?

Apparently Honda's psy-ops people say that seven months is the time it will take for almost all owners to get used to the problem and stop griping about it. :-)

[Anonymous Coward]

Re: Customer disservice...

They probably don't have the staff that worked on that particular system anymore and no documentation. Not uncommon.

[Anonymous Coward]

Re: Customer disservice...

More likely, they bought the software off the shelf from some fly-by-night Chinese company which is long gone since...

[mevets]

Re: Customer disservice...

Is it possible that "... by the end of August", they meant by the time your car shows the date as "August 2022", which might be in human calendar August, 2042?

Either way, I applaud Honda adopting the customer service model of phone and utility companies.

[Chewi]

Ford C-Max

My 2008 Ford C-Max seems to follow the Mayan calendar. It doesn't let you manually set any date after 2012, as it just wraps back round to 2000. Ford evidently didn't think this car would last longer than 5 years! You can work around it by setting the time and date to just before midnight on December 31st and letting it naturally roll over a few times, but you have to be really careful not to touch the year. Bloody annoying.

[John Brown (no body)]

Re: Ford C-Max

Clearly a fault that was present during manufacture, therefore covered by consumer protection law both now and then. Same applies to the Hondas in the article.

Age of the product does matter of course, but something like a car and it's built in device ought to be expected to last many more years than a cheap TV or HiFi,

[.stu]

I remember in a previous role one day all the touchscreens on our customers' machines failed. After a bit of decompiling I found that the mouse driver scanned through the BIOS looking for a specific pattern of byte values, which it then used as an offset to read the serial port settings. Of course on that particular day the date matched the pattern, and just happened to be located a few bytes before the pattern the driver was looking for. We told the customers that we would push out a fix overnight, although we might have omitted to tell them it would happen again in 20-odd years...

[NullDev]

2009 Civic

I have an '09 Civic and can confirm this issue. I thought maybe the CMOS clock battery in the head unit finally died, but then I remembered it was designed to sync with the GPS sats. Maybe I just need to press the "any key" to continue?

[Skiron]

Re: 2009 Civic

If it was the CMOS battery you could have put it correct manually after starting each time - or can't you even do that?

[ThatOne]

Re: 2009 Civic

> the CMOS clock battery in the head unit finally died

That CMOS battery is just a backup, your car has its own main battery after all, so you'd need to have both batteries die at the same time to lose time. And I guess you would had noticed if your car battery died...

Also batteries don't resurrect after 8 months. It's definitely a "unrecognized data" type issue, something about encoding January 1st to August 1st, 2022.

I just wonder if it also happens to Hondas elsewhere in the world, all I hear here is about the USA?

[Sparkus]

The GWRO 'problem'....

has been documented for years. There was at least one article on it right here at El Reg.......

[StuartMcL]

And Furuno!

Similar problem with Furuno marine GPS products on 2 Jan 2022. Many of the models have no fix.

https://www.furuno.co.jp/en/news/notice/notice_category.html?itemid=753&dispmid=965

[Pascal Monett]

Re: And Furuno!

That's interesting. Furuno makes satnavs for boats. Seafaring boats are typically (but not always) bought by people who have a lot of money - or, at least, who want other people to think they have a lot of money.

In any case, those people are generally not inclined to appreciate that something on their precious boat is not working as it should, and they generally make sure that the provider of said knicknack knows about it.

It's too much of a coincidence to not imagine that the bug on Furuno is not the same bug on Honda. The fallout from this is going to be interesting to watch.

[Anonymous Coward]

Re: And Furuno!

I would have pheased that as "Seafaring boats are typically (but not always) bought by people who have" a strong interest in not hitting the rocks.

[ThatOne]

Re: And Furuno!

> Seafaring boats are typically (but not always) bought by people who have a lot of money

Yeah, like crabbers and other fishermen. All stinking rich, aren't they...

Seriously, your world view is a little restrictive here: For every rich yacht owner there are a hundred definitely not rich people working on boats for a living, and using the very same equipment. And even among sailboat owners, there is a vast majority who isn't wealthy. I have met a lot who live on their boat, it's the only thing they own. They just value the liberty it gives them.

[Anonymous Coward]

Re: And Furuno!

They all have a fix, buy our newest model.

[Skiron]

Re: And Furuno!

This note from that page made me laugh:

3. If you want to return the date data to normal for the models with * (asterisk), you can replace the GPS receiver inside the device or update the software (both for a fee).

[Paul Crawford]

Re: And Furuno!

Again, they are blaming the GPS week roll-over which seems odd as it is not happening on the date that occurs. A bit of bullshit to try and cover up other poor software coding?

You would think that companies as big as Honda, Furuno (sales about £0.5B per year) etc, could afford a GPS simulator to actually test that the date/time/navigation works for 30+ years ahead?

Could be an interesting court case if the USA Honda owners bring a class action and they have to give details of the error and original testing performed.

[The Dogs Meevonks]

2008 Honda Accord EX With Satnav

It's also affecting my Accord EX with built in Satnav... but not in the same way. My clock is 1hr out and the time cannot be adjusted at all... It's not due to BST or anything as turning it off does nothing, manually adjusting the time does nothing.

Only started happening after Jan 1st.

[Bod]

Re: 2008 Honda Accord EX With Satnav

2008 Civic EX and noticed it was an hour out. After messing with the settings I realised it was showing Summer Time. Changed to Winter and clock is fine. If I hit reset it goes back to summer.

Hadn't checked the date as didn't realise this was the probable cause until now. I thought it might have a messed up timezone. Even if it thinks it's 2000, surely it should still think it's winter?

Curious about the August reset though.

[Dwarf]

GPS week rollover

GPS week rollover is a real thing. Using a 1K field for week number seems quite short sighted.

So this will affect all GPS kit, but there are well known ways to offset the dates given that we've not figured how to travel back in time yet.

I can't see why manufacturers don't just provide the ability to set the required 20 year offset in their user interfaces, rather than assume that their product won't last 20 years.

I guess this is where the Honda claim comes from - they handled the event badly / not at all. I'd guess that the devs didn't even know about the rollover problem or some delivery manager defined it as outside of the MVP, hence it never got done, even though displaying the correct information should clearly be a requirement.

Here's an article from a well known GPS provider on how to handle this BLOX GPS week number rollover workaround

[david 12]

Re: GPS week rollover

So, it GPS requires a 19.6 year adjustment, and Honda, like many other suppliers, has made a 20 year adjustment.

When the error goes from 1024 GPS week to 20 years (about 7 months), the 'correction' will be correct, and the receiver will stop resetting.

(Just guessing)

[SkippyBing]

Re: GPS week rollover

That the design isn't perfect for something it wasn't intended for isn't really short sighted. When GPS was being specified a 1K field was more than enough given the memory limitations of the time and that if you know about it the roll over isn't an issue. It also doesn't effect the original purpose which was letting the US military know where it is.

[Eclectic Man]

Re: GPS week rollover

!!! WARNING NAZI GRAMMAR / SYNTAX PEDANT ALERT !!!

I hope you mean "affect", rather than "effect".

Either that or you missed the 'joke alert' icon.

[SkippyBing]

Re: GPS week rollover

Little of column A, little of column B...

[David 132]

Re: GPS week rollover

At risk of validating Muphry’s Law (not to mention invoking Godwin’s), may I point out that it’s usually written as “grammar nazi” not “nazi grammar”?

I mean, the latter evokes mental images of a classroom full of jackbooted stormtroopers, repeating by rote “for you ze war is over”, “it is DIE Britischer Swine, nicht DER Britischer Swine”, “i before e except in Heil” etc etc…

[John Brown (no body)]

Re: GPS week rollover

"It also doesn't effect the original purpose which was letting the US military know where it is."

Other than the fact military equipment is harfd to change or update because it has strict specifications and so what used to be modern hi-tech equipment is often still in use 20 or more years later with the same specs. Mind you, I'm basing that on how the UK military works and just assuming the US military will work to similar standards. With their competing and duplicative branches, that may be entirely wrong of me.

[dvd]

Re: GPS week rollover

The US military is still using B52s that were developed in 1952 (70 years ago) and are currently investigating the feasibility of upgrading their engines to extend their lives even further.

The US military likes to get their money's worth from their kit too.

[David 132]

Re: GPS week rollover

Indeed. They are perfectly seriously discussing keeping the airframes updated and flying past 100 years old.

[Death Boffin]

Re: GPS week rollover

In the 80's I was still repairing military radio equipment from the Korean war. Parts were still available in the military supply system even after 30 years. B-52's have a timeline closer to 60 years.

[The Oncoming Scorn]

Re: GPS week rollover

Ohhh you worked at RACAL too?

[SkippyBing]

Re: GPS week rollover

I mean there have been two GPS rollovers since the system begin in the early 80s and no one's got lost because of it, so genuinely not a problem.

[David 132]

Re: GPS week rollover

no one's got lost because of it

…that we’ve heard of…

[doublelayer]

Re: GPS week rollover

"When GPS was being specified a 1K field was more than enough given the memory limitations of the time"

That's 1kweeks, not 1kbytes or 1kbits. In other words, it is ten bits. Memory was expensive in those days, but so were rocket launches and they paid for a lot of them. Had they sprung for 16 bits for the week counter, they could have kept all the rest of their design and only had a rollover every 1265 years. I think that, even in 1973 when the project was started, they could have afforded less than a byte extra in memory per satellite. A single game console of the period had more memory than they'd have to add throughout the whole constellation.

[SkippyBing]

Re: GPS week rollover

Given how long legacy code hangs around for I think a roll-over every 20 years is probably better than once every 1265. At least it's often enough people know how to deal with it!

[doublelayer]

Re: GPS week rollover

I think the chance we're still using the GPS satellites in 3238 is quite low, especially as we already have three other options that work everywhere and don't have such a restricted version. By then, those satellites won't work if they're still in orbit at all. Even if we didn't ever redesign them, one problem that affects people in 3238 beats problems for people in 1999, 2019, apparently also offsets of those years, and almost certainly in 2038. The 1265-year option is only if they wanted to keep their design--they also had the option to report dates in a more configurable way.

[Eclectic Man]

Re: GPS week rollover

I was going to suggest a radio-controlled clock, one that uses the NPL time signal (other time signals are available overseas, I believe), but then I bought one, and it annoyingly used to re-set itself several times a week to an incorrect time, often 15 or 45 minutes wrong. One friend of mine was invigilating an exam when their timing clock re-set itself, causing some consternation amongst the staff and students.

I am tempted to buy a new mechanical wristwatch, but that's probably because I am easily swayed by promotional advertising and shiny sparkly things.

[ThatOne]

Re: GPS week rollover

> a radio-controlled clock

That's a problem with the time signal, not the watch: A a radio-controlled clock/watch is a bog-standard quartz watch, which at some time during the night listens for a radio signal, and uses it to set itself to what should be the correct time.

If it doesn't receive a radio signal during the 10-15 minutes/day it listens for it, it will just keep running as an ordinary quartz watch, slowly drifting over time like an ordinary quartz watch.

Only if the radio signal gives it incorrect information will it set itself incorrectly. This is what seems to have happened in your case.

[Steve Foster]

Re: GPS week rollover

The only flaw in this otherwise wonderful (and informative) thought is that the GPS week rollover occurred in April 2019. Not at the end of 2021.

So, if Honda had screwed up dealing with the rollover, you'd have expected the reports to have begun back in 2019, not now.

[Robert Carnegie]

Re: GPS week rollover

I think the patch for GPS week rollover was, roughly, to add 1024 to the week number if it is less than X. So if X is 512, then weeks 512 to 1023 are used as is, followed by weeks 0 to 511 which are treated as week 1024 to 1535. As noted, if this is put in place just after week 512, then it will go wrong 19½ years later, but you should update the software before then using a different value of X, such as 768 coming up about 5 years later.

Or to put it another way, you can design the Week1K failure to happen on any future week date of your choice, within 19½ years.

[Sykowasp]

Re: GPS week rollover

Maybe the core code in this 2007 GPS module was written in 2002.

It includes a line like this to cope with the week rollover, perhaps some of the test cases were pre-1999:

if (year < 2002) then year = year + 20

because they knew at coding time that the year would never be under 2002, so this delayed the week rollover problem by 3 years (i.e., the cars have been thinking it's 1999-2002 for three years now).

but this wouldn't fix itself in August on its own. So there must be some other code that also munges the GPS date and week and that will return a higher year number prior to this code, come August.

Who cares about the year in a car anyway? But the time is essential, and day/month can come in handy.

[Hero Protagonist]

The two types of software hardest to get right

Cryptography, and date/time/calendar calculations

[VRocker]

I was waiting for this news article to appear (i was thinking of prodding you guys about it myself but i'm lazy...)

So i have some more info on this as it's affected my CR-Z and i've been looking into it. The date reported by the GPS is May 2002, exactly 1024 weeks ago so it is the rollover issue. It's not due to GPSd though as these units run Windows CE for Automotive. The reason it's going back to 1am every time the car turns on is due to some error handling in the code itself. If the date is before a certain date, it resets to 1st Jan 2002 at midnight. The 1am comes from the auto summer time kicking in (which goes on the date reported by the GPS, not the 'corrected' date) so being May 2002 means it's summer time. Timezones also get applied to the reset time which is why Americans are seeing 4/5am.

The suspected reason why they say it'll fix itself in August is that the date valid check could well be 2003. On 17th August, the unit will think it's 2003 and may say the correct time, although auto summer time won't work for obvious reasons.

I have my doubts that it'll correct itself this August though, as everything i'm seeing in the code seems to point to 2004 being the valid date check... but we'll see.

I currently have Ghidra open with the firmware for my own nav trying to find this check to nop out... we'll see how that goes!

[AndrueC]

One thing that bugs me about the systems on my recent cars (Honda Jazz and Toyota Corolla) is that despite both cars asking me what timezone I'm in neither of them automatically enable/disable DST. I can understand that they might not want to stick their neck out and support DST because it's a bit of a movable feast but in that case why do the cars need to know the timezone?

[ThatOne]

> why do the cars need to know the timezone?

I guess they work internally in UTC, so they need to know the offset to apply depending on where you live: Else your clock would be 8 hours off in St Francisco for instance (and let's not even mention Australia...).

I can understand they don't want to be responsible for DST though, as DST rules change regularly and always change over very short distances: Your car would need to calculate where your car exactly is at any moment legally, so it can apply the correct and current DST legislation for that specific place. Only to be told you don't really want your clock to change automatically on every timezone you drive through (or come near to), you want it to remain on your own home time... It's too much of a hassle.

[AndrueC]

I guess they work internally in UTC, so they need to know the offset to apply depending on where you live: Else your clock would be 8 hours off in St Francisco for instance (and let's not even mention Australia...).

Both vehicles need the time setting manually as well though. I could run my Corolla on EST if I wanted to even though I'm in the UK.

[ThatOne]

I see. In this case I guess it's probably software reuse from a configuration using networked/GPS time?

My 2007 car's clock is a simple "enter hour and minutes" job. (I deliberately bought the version without the inbuilt GPS, because I knew I would keep it way beyond the useful life of inbuilt car GPSes)

[Daniel 18]

Not setting DST automatically is a wise choice.

Even in the same time zone, there are cases where some regions use DST, and some do not. Given discussions I have seen, some regions are considering going to permanent DST, or may have done so. These regions could be countries, provinces, states, or cities - I believe there are examples at each of these levels.

Any automatic setting of DST based on time zones would leave some users with clocks that persistently reset themselves to the wrong time for about half the year.

The most logical answer would be to ditch DST altogether.

The other good alternative would be to go to UTC for the entire planet, simplifying a number of things.

[AndrueC]

The problem with ditching DST (from my perspective) is that I'd want to shift my working hours forward and backward so that I could continue to 'move' an hour of daylight from the morning to the evening where I can make better use of it.

If the UK didn't have DST the sun would be rising at 4am in June. That means at least 3 hours of wasted daylight. I've said before that I'd support double DST for May/June/July in the UK and when I'm retired will likely wake up an hour earlier in those months.

[Daniel 18]

I'm not sure that's a problem.

You just start work at 0900 in part of the year and 1000 in the rest of the year.

The next time zone over could do 1000 and 1100, and so on.

I've never understood why some people thing DST is necessary in order to be working the same time as companies in an adjacent DST region. Just change your nominal start time, and you are done, no fiddling with clocks and other devices required.

Also, logs kept in local time wouldn't have discontinuities.

[AndrueC]

That depends where you work. I think a lot of employers would be irritated if half the staff started work an hour earlier than the rest. And suppose all the staff want to work an hour early - will an employer be happy to have the business running out of sync with its competitors and customers? And what if you need to rely on public transport?

The advantage of the current system is that by and large everyone is working to the same timetable.

[Daniel 18]

DST is bad and double DST is worse if you are trying to do astronomy and astrophotography.

It is already an annoyingly long wait in the summertime for it to become completely dark.

[AndrueC]

Well I'm not an astronomer so I don't see how that is relevant to my post.

But if you want to open the debate up to a wider audience then I'm fairly sure that the number of people who like to be outdoors in the dark is dwarfed by the number of people who like to be outside in daylight. Both groups are (sadly) dwarfed by the number who prefer to be indoors but they have no reason to care either way - unless possibly they save on their lighting bill.

And the number of people who would enjoy daylight before 7am is equally small, at least outside of Scotland perhaps. And the number anywhere in the UK that would appreciate daylight before 6am must be vanishingly small. So moving that hour to the evening benefits a lot more people than it annoys I reckon.

You can't please all the people all the time but the current DST system seems to work well enough for most people. It's hardly difficult to deal with and as I've posted previously most clocks can be replaced with ones that perform the switch automatically these days. Even watches can do it.

And I'll re-iterate that my comment about double DST was a personal preference ;)

[Dante Alighieri]

Streetlights

Would be better if there were a few hours when the streetlights turned off

The physical/nominal time is irrelevant

[jtaylor]

My 2019 Honda Fit (Jazz) has the lower trim "color display" option and requires me to manually set the time and also manually change it for DST. I feel less deprived now. :)

[Dizzy Dwarf]

No leap second this year

Nor in 2002, thankfully.

Can you imagine the chaos?

[Paul Hovnanian]

Woo hoo!

The warranty on my old crap-mobile is back in effect again.

[The Oncoming Scorn]

Re: Woo hoo!

Hopefully they will stop ringing you about it now.

[brainwrong]

useless

Easy solution =>

1) stop being lazy fucknuts and learn to read a map, there's plenty available in print and online. satnav not needed.

2) set your clocks manually using the pips on the hour on radio. use proper FM radio.

I despair at humans being so lazy that they become dependant on technology that they're too lazy to make work correctly.

[ShortLegs]

Actually, were I to own a Honda, what would annoy me is the usual non-apology

"American Honda is aware of a /potential concern/ related to the clock display...,"

A tweeted complaint c/w picture is only a 'potential concern' my arse.

Its like the apology for "any distress that might have been caused" after incidents where people have been physically injured, or suffered personal loss.

Corporate communications employees... should be up against the same wall as marketroids and salesnakes.

[jtaylor]

"is aware of a /potential concern/ related to the clock display...,"

That phrase really gets on my wick. "Potential" indicates that it's not realized. Hint: you're replying to an expressed concern. It's more than a concern, too: it's actually happening.

"any distress that might have been caused"

Double-ugh! "People told us about the problem, but we don't believe them."

[ThatOne]

The first thing any company lawyer will tell you is to not admit anything, at most say you've heard rumors. Admitting there is a real problem is opening yourself up to lawsuits, so companies are willing to pay huge sums to avoid admitting actual fault.

[Jilara]

Backdated Banking

Dates have been hinky in some other places lately, as well. Just before Christmas, I logged into my credit union, and found it was June 1, 2015. I had been planning to transfer some funds, but decided that wasn't the best idea at that time. Fortunately, the date was normal when I logged in a few days later.

Where does money go when it travels into the Past? Does it become quantum currency? Can you fix an account that was overdrawn in the past?

[TRT]

Re: Backdated Banking

Deposit a penny in a savings account in your own era (at BoE base rate) and by the time of the end of the universe the action of compound interest will mean that you can probably afford a small bowl of chips at Milliways.

[ThatOne]

Re: Backdated Banking

Nah, you forgot to factor in inflation.

Most smaller black holes out there actually started as saving accounts, their value getting smaller and smaller over time until they collapsed into a singularity.

[vincent himpe]

how hard can it be

5 bit for hours (24 format), 6 bit for minutes , 6 bit for seconds , 5 bits for day (0 to 31 where zero is invalid), 4 bits for month , 3 bits for day of week . gives 29 bits. the remaining 3 can be a checksum , an am/pm flag (even though we have 24 hr format) and a leapyear flag.

This gives one unsigned 32 bit number. if you don't want to muck with signs : ditch the am/pm flag and leave msb always at zero.

The second 32 bit number is the year. (starting at 0). The sun will be halfway to going supernova by the time we run out of bits ...

Come on , get it done.

[Lennart Sorensen]

Re: how hard can it be

GPS is trying to fit a lot in a small message. So they went with 10 bits for week number and 19 bits for seconds of the week. So 29 bits handles the entire date + time for a 20 year period. Newer GPS signal formats did extend the week to 13 bits making the entire message 32 bit and handling 157 years with second level precision. A rather efficient format really. Unfortunately only the newest GPS satellites use the 13 bit week format, most are still 10 bit, and almost no receivers are taking advantage of the 13 bit option yet, probably since most of the signals they receive wouldn't contain it yet.

Keeping crap like day of week, length of each month in days and such out of the format simplifies things a lot. Converting to human format can be done in software from the raw signal afterwards.

So GPS managed 20 year cycles with 29 bits (less than half of your proposal) while the newest GPS does 157 years in 32 bits (exactly half your proposal). I think the designers were rather clever.

[vincent himpe]

Re: how hard can it be

Who says the clock has to rely on gps ? there is plenty of other systems out there.

And time is not a problem only in gps. we had a year 200 problem and the unix year rollover problem is also looming... many programs have trouble with time/date stamps. since all computers ( well, most) these days are 32 bits .. two words can give you to-the-second time.

[Lennart Sorensen]

Re: how hard can it be

Using GPS makes sense when you have a GPS navigation system already and you want accurate time without the user having to keep fixing it. Of course Honda has managed to show that they are not good enough to actually use GPS correctly for this. My Prius which does have a GPS system does not use if for the clock. It has a couple of buttons for me to set it. Maybe the Toyota engineers were on to something when they kept it simple, even though I thought it might have made sense to let it at least have an option to set it by GPS. Or perhaps they thought since not every trim level had a GPS, the clock might as well work the same in every trim level.

And at least in the case of linux, they have already been working on the unix epoch issue. 64 bit systems no longer have any problem, and 32 bit ones are close to fixed, so at least they seem to be working on it with about 16 years to spare, which is a lot better than the Y2K work that was rather last minute.

[Death Boffin]

DST Fix?

My 2006 Odyssey clock is running exactly 1 hour fast. It will be interesting to see if it fixes itself when Daylight Savings Time rolls around. Or will it become off by some other amount.

[TRT]

Unseen...

Y2K was obvious really. What happens when the shortcut way we wrote dates isn't cutting it no more? Obvious problem because humans can see it and read it.

But less obvious what happens when machines can't read it? Well... we don't notice those so much so we?!

[FirstTangoInParis]

Unsurprisingly…..

My SOs current and past Jazz suffers from an impenetrable UI for the in car entertainment (or really most of the car). To call it a mess is to equate it to a teenagers bedroom. It’s far worse. When you cannot work out how to listen to the radio, and need to resort to YouTube to discover how to pair a phone or reset the tyre pressure alarm, But then there is the car manual, same issues as any other car …. Why do I need to consult page 500-odd then page 400 and something to simply pump the tyres up and reset the alarm?????