3 ms·

Seems reasonable tbh. UTF-8 is 31 years old at this point. Unicode is even older, and various 8-bit code pages older still. At some point you have to get with t

Seems reasonable tbh. UTF-8 is 31 years old at this point. Unicode is even older, and various 8-bit code pages older still. At some point you have to get with the times.

corbezzoli 3y ago
> At some point you have to get with the times. Nobody wants to spend money and risk everything. There are so many industries where "getting with the times" is absolutely out of the question due to legal issues, as well as momentum. Aviation, healthcare, banking, …
- davedx 3y ago
  That portrays it as if all the risk is on changing the system. Banks probably have to deal with a constant level of customer support if they don't store names properly. Credit checks probably go wrong, sometimes in ways that incur real financial losses on them. There is very real risk in refusing to make your software match reality.
- jeroenhd 3y ago
  And, as this lawsuit shows, not getting with the times exposes the company to repercussions as well. I doubt a business would prefer their own schedule over one set by the court. ING Belgium found out the hard way, and I doubt this change to their systems was cheap. They were forced to alter a system that was due to be replaced (though that didn't happen in the expected timeline, surprise surprise) when they could've rolled out this change over a long period of time somewhere in the last two decades.
- iSnow 3y ago
  If someone brings this up, I am actually happy if legislation forces them to do it. More often than not, it is possible and companies end up with better systems.
rsynnott 3y ago
Even if they are married to EBCDIC, EBCDIC _has code pages_.
- extraduder_ire 3y ago
  I saw that mentioned in a comment under the blogpost. Where would info on which code page is meant to be used stored? I assume it's not inline with the text.
ezoe 3y ago
UTF-8 may be invented 31 years ago. But it was widely used until recently. I don't remember when the switch happened, but I think it was widely adopted sometime in middle of 2000s to late 2000s. GNU/Linux distro changed their default locale to UTF-8 and Web sites starting to use it.
- bawolff 3y ago
  This graph https://en.wikipedia.org/wiki/File:Unicode_Web_growth.svg https://en.wikipedia.org/wiki/File:Unicode_Web_growth.svg suggests 2008 is when UTF-8 ate the world. That's still 15 years ago. However it should be noted, that the encoding that UTF-8 was replacing was largely (not exclusively) iso-8859-1 which probably had the characters we are talking about.

Seems reasonable tbh. UTF-8 is 31 years old at this point. Unicode is even older, and various 8-bit code pages older still. At some point you have to get with the times.

corbezzoli 3y ago

> At some point you have to get with the times. Nobody wants to spend money and risk everything. There are so many industries where "getting with the times" is absolutely out of the question due to legal issues, as well as momentum. Aviation, healthcare, banking, …

davedx 3y ago
That portrays it as if all the risk is on changing the system. Banks probably have to deal with a constant level of customer support if they don't store names properly. Credit checks probably go wrong, sometimes in ways that incur real financial losses on them. There is very real risk in refusing to make your software match reality.
jeroenhd 3y ago
And, as this lawsuit shows, not getting with the times exposes the company to repercussions as well. I doubt a business would prefer their own schedule over one set by the court. ING Belgium found out the hard way, and I doubt this change to their systems was cheap. They were forced to alter a system that was due to be replaced (though that didn't happen in the expected timeline, surprise surprise) when they could've rolled out this change over a long period of time somewhere in the last two decades.
iSnow 3y ago
If someone brings this up, I am actually happy if legislation forces them to do it. More often than not, it is possible and companies end up with better systems.

rsynnott 3y ago

Even if they are married to EBCDIC, EBCDIC _has code pages_.

extraduder_ire 3y ago
I saw that mentioned in a comment under the blogpost. Where would info on which code page is meant to be used stored? I assume it's not inline with the text.

ezoe 3y ago

UTF-8 may be invented 31 years ago. But it was widely used until recently. I don't remember when the switch happened, but I think it was widely adopted sometime in middle of 2000s to late 2000s. GNU/Linux distro changed their default locale to UTF-8 and Web sites starting to use it.

bawolff 3y ago
This graph https://en.wikipedia.org/wiki/File:Unicode_Web_growth.svg https://en.wikipedia.org/wiki/File:Unicode_Web_growth.svg suggests 2008 is when UTF-8 ate the world. That's still 15 years ago. However it should be noted, that the encoding that UTF-8 was replacing was largely (not exclusively) iso-8859-1 which probably had the characters we are talking about.