The UK's HM Revenue and Customs has broken the Customs Declaration Service forms that importers and exporters use.

Around the world, programmers who have coopted characters used in "real life" have found that when real people (i.e. people who have real lives) use them, it confuses their programs. Instead of fixing their own mess, they make it so that those coopted characters are rejected if a user uses them. Hence why credit cards and passports with a hyphen are rejected by a plethora of booking systems that all say "enter the name exactly as it is on the.....".

But programmers are an isolated and bizarrely self-centred lot and they don't see that they caused the problem. They think it's caused by people who are behaving in the way that they behave in the real world.

So, HMRC has come a cropper. Someone has modified a system in such a way that the "validation" of entries fails if a "special character" is used. The quotation marks are to show that these are programmers' terms, not what we in real life use. But as if that's not bad enough, the system has been programmed to reject forms containing diacritic characters. That's not a programmer's term: it means, at its simplest, characters with accents but also includes apostrophes. Programmers hate it when real people use apostrophes because they are commonly used in many programming and scripting languages. So when a user uses an apostrophe, the program needs to know if it's part of the input or part of the program and that is really quite complicated to achieve.

Programmers also don't like to talk about problems. They call products "solutions" when they aren't and they call problems "issues". So when HMRC says there is an issue, we know they mean this: "our programmers really f***ed up."

This is what HMRC has said in the past half-hour or so:

The customs declaration service (CDS) is currently unable to accept special and diacritic characters.

Special characters include, but are not limited to:

%
_
{ }
[ ]
#

Diacritic characters include, but are not limited to:

é – accent acute
è – accent grave
ê – circumflex
ë – umlaut or diaerisis
ç – cedilla
ñ – tilde
ø – streg
ð – eth (capital form Ð)
å – bolle
æ – ligature
œ – ligature
ē – macron
č – háček
ŭ – crescent

We would ask that traders do not use special or diacritic characters in their Department for Business and Trade (DBT) ICMS 2, SPIRE and LITE licence applications until the issue has been resolved.

This issue is planned to be fixed as part of the CDS Release in Spring 2025.

So, "really f***ed up and we can't fix it but maybe we will some time in the next few months".
And that "including but not limited to...." means "we don't know how bad the problem is."

A conspiracy theorist might say that the programmers don't want to fix it soon because, if real people can be trained to avoid these real-world parts of language, then programmers' lives will be simpler.

Butchers' blocks, Doll's Houses, Goats' cheese might be left on the dockside for months or turned back because the documentation didn't work. Or maybe not: that's pure supposition but it shows how silly this problem is and why it needs to be fixed sooner rather than later - 'cos right now, it's a dog's dinner.