aleman
Club Regular
Posts: 266 
Primary Vehicle: Mini Countryman
Year: 68 - 2018
Model Spec/Trim: S E ALL4 PHEV
Engine Capacity: 1499
Fuel Type: Hybrid
Transmission: Automatic
Drive Type: 4WD/AWD
|
Post by aleman on Feb 19, 2019 10:31:04 GMT
Ah, yes, and no ... the year 2000 issue was real and over the years prior to the start of the new millennium billions were spent on testing and modernising IT systems precisely so that nothing terribly untoward happened when the clocks ticked over. I guess that the issue was "over-hyped" but if it hadn't been the corporate execs probably wouldn't have been scared enough to prepare ... ... but just imagine the panic if we had social media back then! A lot of wasted money and thank god there wasn’t any social media. Can just imagine cults being set up on Facebook called Armageddon and then you would’ve had the nut jobs causing even more havoc Not so much a waste of money, it was much needed in a lot of computer systems ... Unfortunately not enough thought was given to the process ... Microsoft OS is still going to be affected by the 2032 (  IIRC) bug, and pretty much all software running on it ... SQL Server date fields have an issue, and there is a bit of a kludge around it allowing you to define just when the 'rollover' happens (is 24/10/50 1950 or 2050???). So although a lot of money was spent it was just , as usual, spent fixing the immediate issue, not necessarily the problem. I was thinking about buying a new Canyon Endurace 9.0 Or Cube Agree C:62 Race Disc from Germany, but it's not looking likely now  |
|
|
|
Post by Ravasher on Feb 19, 2019 10:33:08 GMT
A lot of wasted money and thank god there wasn’t any social media. Can just imagine cults being set up on Facebook called Armageddon and then you would’ve had the nut jobs causing even more havoc Not so much a waste of money, it was much needed in a lot of computer systems ... Unfortunately not enough thought was given to the process ... Microsoft OS is still going to be affected by the 2032 ( IIRC) bug, and pretty much all software running on it ... SQL Server date fields have an issue, and there is a bit of a kludge around it allowing you to define just when the 'rollover' happens (is 24/10/50 1950 or 2050???). So although a lot of money was spent it was just , as usual, spent fixing the immediate issue, not necessarily the problem. I was thinking about buying a new Canyon Endurace 9.0 Or Cube Agree C:62 Race Disc from Germany, but it's not looking likely now What’s the 2032 bug? |
|
aleman
Club Regular
Posts: 266
Primary Vehicle: Mini Countryman
Year: 68 - 2018
Model Spec/Trim: S E ALL4 PHEV
Engine Capacity: 1499
Fuel Type: Hybrid
Transmission: Automatic
Drive Type: 4WD/AWD
|
Post by aleman on Feb 19, 2019 10:47:31 GMT
Not so much a waste of money, it was much needed in a lot of computer systems ... Unfortunately not enough thought was given to the process ... Microsoft OS is still going to be affected by the 2032 (  IIRC) bug, and pretty much all software running on it ... SQL Server date fields have an issue, and there is a bit of a kludge around it allowing you to define just when the 'rollover' happens (is 24/10/50 1950 or 2050???). So although a lot of money was spent it was just , as usual, spent fixing the immediate issue, not necessarily the problem. I was thinking about buying a new Canyon Endurace 9.0 Or Cube Agree C:62 Race Disc from Germany, but it's not looking likely now  What’s the 2032 bug? May not be 2032, I think I have all the right number but not necessarily in the right order  Another date issue caused by storing / using a date in an 'inappropriate' data type (In this case int4) As computers only work with bits, those bits are grouped together in bytes to allow us to do 'sums' and represent data (characters) ... The Y2K issue was caused primarily by mainframes only using 6 characters to store a date 241050, as most mainframes didn't understand the concept of a date field (Which is usually some type of integer value anyway). The only problem with integers is they can represent positive and negative numbers -32767 to 32676 for example (int 2 bytes) larger integers (Int4 - 4 Bytes and BIGINT - 8 Bytes) cover a correspondingly bigger range, INt 4 I think is -21 Million odd to +21 Million odd If you start using say 1 to represent 01/01/1900 (Ala Microsoft) then eventually you will run out of positive whole numbers ... Not an issue you say? Well it is because integers roll over so you go from, just as an example, +32767 to -32767. That means you go from 01/01/1900 + 32767 days to 01/01/1900 MINUS 32767 days overnight, and computers won't handle that gracefully. (If you are running Linux I think you'll be ok, as most distros have been written to handle dates properly since the problem came to light)
Sorry, I'm a geek, I got paid to know this sort of stuff, but luckily since Y2K I haven't had to think bout it ... And I'm going to be retired when the next doomsday date scenario comes around ... although I could probably be enticed back to contracting for an obscenely large amount of money , as a lot of these younger developers are to isolated from the hardware to really grasp what is going to happen. ... and they have't been through it all before |
|
|
|
Post by Ravasher on Feb 19, 2019 11:01:45 GMT
May not be 2032, I think I have all the right number but not necessarily in the right order Another date issue caused by storing / using a date in an 'inappropriate' data type (In this case int4) As computers only work with bits, those bits are grouped together in bytes to allow us to do 'sums' and represent data (characters) ... The Y2K issue was caused primarily by mainframes only using 6 characters to store a date 241050, as most mainframes didn't understand the concept of a date field (Which is usually some type of integer value anyway). The only problem with integers is they can represent positive and negative numbers -32767 to 32676 for example (int 2 bytes) larger integers (Int4 - 4 Bytes and BIGINT - 8 Bytes) cover a correspondingly bigger range, INt 4 I think is -21 Million odd to +21 Million odd If you start using say 1 to represent 01/01/1900 (Ala Microsoft) then eventually you will run out of positive whole numbers ... Not an issue you say? Well it is because integers roll over so you go from, just as an example, +32767 to -32767. That means you go from 01/01/1900 + 32767 days to 01/01/1900 MINUS 32767 days overnight, and computers won't handle that gracefully. (If you are running Linux I think you'll be ok, as most distros have been written to handle dates properly since the problem came to light)
Sorry, I'm a geek, I got paid to know this sort of stuff, but luckily since Y2K I haven't had to think bout it ... And I'm going to be retired when the next doomsday date scenario comes around ... although I could probably be enticed back to contracting for an obscenely large amount of money , as a lot of these younger developers are to isolated from the hardware to really grasp what is going to happen. ... and they have't been through it all before I think I just about understood that🤔 thanks for the explanation 👍🏼 |
|
|
|
Post by philip42h on Feb 19, 2019 11:25:58 GMT
I think Aleman means the Year 2038 problemAnd as he says both the Y2K issue and its successor relate to the cost of storage and the number of bits dedicated to hold a date. Storage is much cheaper than it was and if in the next round of fixes folk move up to using 64 bit integers the problem will be solved for quite a long time. But that means a significant change (more storage used and different database models). Last time there was quite a lot of 'temporary fix up in place' going on where they just used the available storage differently. |
|
aleman
Club Regular
Posts: 266
Primary Vehicle: Mini Countryman
Year: 68 - 2018
Model Spec/Trim: S E ALL4 PHEV
Engine Capacity: 1499
Fuel Type: Hybrid
Transmission: Automatic
Drive Type: 4WD/AWD
|
Post by aleman on Feb 19, 2019 13:01:57 GMT
I think Aleman means the Year 2038 problemAnd as he says both the Y2K issue and its successor relate to the cost of storage and the number of bits dedicated to hold a date. Storage is much cheaper than it was and if in the next round of fixes folk move up to using 64 bit integers the problem will be solved for quite a long time. But that means a significant change (more storage used and different database models). Last time there was quite a lot of 'temporary fix up in place' going on where they just used the available storage differently. Thats the one, a good 15 years after I'll have ceased to worry about it  |
|
|
|
Post by widge on Feb 19, 2019 23:15:33 GMT
Well; that's it then init! 10% surcharge due to Brexit has made me cancel my order for the new Porsche. Anyone else been similarly affected?? To be fair if I understood correctly porche aren't increasing their prices, the are just warning any customers who have ordered since the middle of January that there maybe an additional cost due to import duty of upto 10%, they are asking customers to sign something to say that they understand this, presumably they have the option to cancel if they are unwilling to do that. Dave |
|
IIRC) bug, and pretty much all software running on it ... SQL Server date fields have an issue, and there is a bit of a kludge around it allowing you to define just when the 'rollover' happens (is 24/10/50 1950 or 2050???). So although a lot of money was spent it was just , as usual, spent fixing the immediate issue, not necessarily the problem.
