The Grenadier Forum

Register a free account today to become a member! Once signed in, you'll be able to contribute to the community by adding your own topics, posts, and connect with other members through your own private inbox! INEOS Agents, Dealers or Commercial vendors please use the contact us link at the bottom of the page.

New Year Service Overdue Bug

You would not be the first I had to fix up. I will check after adding the new numbers.

EDIT: The master list is on page 1, and it looks correct. Sometimes other lists are not updated, except the very last one in the thread. Or did I get the number wrong?
Sorry … I was looking at p.11 🙂
 
View attachment 7918106

1. The "Leap Year" & Unix Bug Hypothesis​

The date 1/1/2026 is significant. It is exactly one year after the end of 2024 (a leap year).

  • The Pattern: Many of these numbers are roughly divisible by 365.
  • Hypothesis: The vehicle's Internal Real-Time Clock (RTC) encountered a "Year Overflow." On Jan 1, 2026, the system likely tried to reconcile its "Service Due" counter against a hardcoded "Manufactured Date." If the software doesn't recognize the year 2026 properly, it defaults to a "Total Days Since Epoch" calculation, causing every car to suddenly realize it hasn't been "serviced" since its software was first written.

2. The "Integer Overflow" Logic​

Notice how many values are hovering around the 2,048 mark (e.g., acky at 2,048, Commodore at 2,057).

  • The Pattern: In computing, $2^{11} = 2048$.
  • Hypothesis: This is a memory bit error. The "Service Counter" in these vehicles may be stored in an 11-bit or 12-bit integer. Once the clock struck 2026, the calculation for "Days since last service" exceeded the allocated memory space for that variable, causing the numbers to "wrap around" or pull random data from the next memory stack.

View attachment 7918111
View attachment 7918118
Analyzing the patterns will be an all nighter...

View attachment 7918107
Interesting plots. That histogram definitely has has a positive skew. A log transformation brings it closer to normality but not quite. A normal population on a probability plot would be a flat line. Values above 2300 and below 1800 are true outliers to the main population so culling them will bring the distribution much closer to normalcy. Whether that means anything in analysing the data FIIK

1767337000991.png


How to interpret the Tukey Box and Whisker plot
1767337196541.png
 
I noticed that there aww a few of us with the same number of days showing. I wonder if we can figure out the communalities and then the root cause from that data.
My number (1995 days or 5 years 5 1/2 months) when added to my actual last service date takes me to exactly 6 years hence i.e. 2nd Jan 2032.
 
TALLY (Updated)
ResponsiblyReckless 1625 days
Chook 1840 days
anti 1841 days
Itsdchz 1841 days, first US car with the pox
BrandonADV 1852 days
TahoeGren 1852 days
Znarfgh 1864 days
PB60 1865 days
Pat-Ard 1868 days
Rok Dr 1869 days
Sonoranrover 1875 days
Michael H 1889 days
Robt 1890 days
Chem409 1890 days
Duncan McDonald 1892 days
joejet 1894 days
Korg 1899 days
GrenADV 1901 days
Grumpy 1906 days
HeavyFoot 1915 days
pedrogb1915 days
Tom Ireland 1916 days
Woody 1919 days
Force Pro 1924 days
OGrid 1927 days
singlefin 1927 days
WhiteBear 1931 days, first car outside ANZ
ecohen2 1932 days
Happyoldgit 1933 days
OkGo 1937 days
prezioson 1940 days
NixGrenadier 1940 days
Peterr's neighour 1944 days
ddv 1945 days (Aug 2023)
Tergosa 1946 days
Johnb 1950 days
Fred 1950 days
Camilondo 1953 days
Peterr 1961 days
Jeremy966 1967 days
Ex-defender driver 1968 days
TheDocAUS 1968 days (Feb 2024)
Denis M 1975 days
Mohs9 1989 days
MattG 1989 days
Green Gren 1903 days
TJD 1985 days
Phred 1994 days
douggie 1995 days
Glen 2007 days
Badger4x4 2007 days
Adam 2007 days
Baron van T 2009 days
gratefuldiver 2021 days
Silverjay 2022 days
Chris Hilton 2024 days
Tony H 2025 days
SaudiGR 2026 days
Landmannn 2044 days
acky 2048 days
Commodore 2057 days
NeoGTi 2057 days
HelgeL 2059 days
Greasemonkey 2059 days
Lollo050968 2074 days
Wile E Coyote 2095 days
Dschubba 2097 days
Stu Barnes 2097 days
Manuel4x4 2106 days
Mazeman 2107 days
Tazzieman 2137 days
Phobiac 2141 days
pipo 2149 days
Logsplitter 2150 days
Skymark 2156 days
slukell 2175 days
Emil 2190 days
bigleonski 2250 days
Ron50 2253 days
Clark Kent 2367 days
C-Mack 2458 days
mirwantandyo 2668 days

NET ZERO
Steve B, baders and Jean Mercier have no alert.

INEOS TALKS
INEOS has acknowledged the issue here.

SUMMARY
82 members listed so far, from lowest to highest number of days. Plus 3 members with no alert.
 
Last edited:
View attachment 7918106

1. The "Leap Year" & Unix Bug Hypothesis​

The date 1/1/2026 is significant. It is exactly one year after the end of 2024 (a leap year).

  • The Pattern: Many of these numbers are roughly divisible by 365.
  • Hypothesis: The vehicle's Internal Real-Time Clock (RTC) encountered a "Year Overflow." On Jan 1, 2026, the system likely tried to reconcile its "Service Due" counter against a hardcoded "Manufactured Date." If the software doesn't recognize the year 2026 properly, it defaults to a "Total Days Since Epoch" calculation, causing every car to suddenly realize it hasn't been "serviced" since its software was first written.

2. The "Integer Overflow" Logic​

Notice how many values are hovering around the 2,048 mark (e.g., acky at 2,048, Commodore at 2,057).

  • The Pattern: In computing, $2^{11} = 2048$.
  • Hypothesis: This is a memory bit error. The "Service Counter" in these vehicles may be stored in an 11-bit or 12-bit integer. Once the clock struck 2026, the calculation for "Days since last service" exceeded the allocated memory space for that variable, causing the numbers to "wrap around" or pull random data from the next memory stack.

View attachment 7918111
View attachment 7918118
Analyzing the patterns will be an all nighter...

View attachment 7918107
The actual answer is much simpler - the internal clock leapt ahead to 1 Jan 2032 rather than 1 Jan 2026 (2032 is the next occasion that 1 Jan is a Thursday). Number of days overdue was determined by previous due date to 1/1/2032. Even once the date is reset, the alert remains active until reset by dealer.
 
The actual answer is much simpler - the internal clock leapt ahead to 1 Jan 2032 rather than 1 Jan 2026 (2032 is the next occasion that 1 Jan is a Thursday). Number of days overdue was determined by previous due date to 1/1/2032. Even once the date is reset, the alert remains active until reset by dealer.
Spot on. My due date was/is 19.8.2026 and that is 1961 days to 1.1.2032.

For some reason HVAC vent and temperature knobs went unresponsive the same day that the service reminder went haywire. Fan power knob works but the vent and temp knobs do nothing.. Lovely to have only cold air blowing to my face and nothing to windscreen (winter and -12 celsius outside so can't see = can't drive). Of course i'm waiting for my HT coolant reservoir to be replaced because it's leaking but I'm constantly topping it up so that is not the problem now. Mute button reset doesn't help.
 

Attachments

  • 26_bug.jpg
    26_bug.jpg
    2.3 MB · Views: 3
I'm on 1980 days overdue, before the service reminder popped up my Android auto gave a security warning, checked the date and my Ineos my 23 had gone to the future... Aligns with an overflow error of some form....
6368.jpg
 
See post #250 above. 😉

🤣

Exactly the issue I had yesterday. I couldn’t get the bleeding procedure to work correctly.

I think it may be a controls thing, I could get good heat from the dash top vents by using the front screen demist - proper hot air but any other setting have nothing but cold air - no heating at all.
 
Back
Top Bottom