to double down on it it's not even about a weird name per-se but lot of very normal EU names

if a lot of early IT wouldn't have been dominated by a US "works for us must work for everyone" approach I think we never would have ended up with such limitations common in legacy systems (there still would be limitations, pre-unicode the solution was custom code pages and similar, which all supported some subset of non us-ascii but only a subset)

luckily today unicode is the standard (through for some cultural and historic aspects it's sometimes not enough)

I'm all for bashing the US, but I think this is a global ghing, not unique to the US.

This was an accident of history, not some deliberate plan to get the world to bow to the English speakers. And English was already well established as a major language in trade (due to it being superficially simple to learn), next to German, French and Spanish. China was pretty isolated for a long time culturally as well as geographically and the complexity of its script is another barrier to it being accepted as a common language by the rest of the world.

One of the more interesting things along this line in recent history is that with Brexit the EU no longer has an England/Wales/Scotland and a chunk of Ireland in it, but another chunk of Ireland remains. This led the French to immediately propose that French become the official language of the EU parliament but the rest of the countries wouldn't have it, and rightly so.

https://www.independent.co.uk/news/world/europe/brexit-franc...

I, however, don't think most of the people who started using one of the named languages instead of their mother tongue ever really selected English using that specific criterion.

Didn't happen, they just said they'll use French during their council presidency (not the parliament, it's not even mentioned in your article), that's all, there are no rules against that. They would've done it regardless of Brexit.

Obviously this means the EU should run things in Venetian.

> lingua franca, (Italian: “Frankish language”) language used as a means of communication between populations speaking vernaculars that are not mutually intelligible. The term was first used during the Middle Ages to describe a French- and Italian-based jargon, or pidgin, that was developed by Crusaders and traders in the eastern Mediterranean and characterized by the invariant forms of its nouns, verbs, and adjectives. These changes have been interpreted as simplifications of the Romance languages.

- if you move to the US and have a name made up of non-ASCII chars you are more likely to either drop them/substitute them with ascii chars, or use the Anglicized version of your name if it exists, or adopt an English name. And then it's kinda easy to legally change or your name. Or screw it, it's kinda easy to just show up and tell them you're Johnny Awesome and then you're Johnny Awesome.

- if you move to Germany, you can't legally change your name at all without good reason, every document ever, no matter how informal (especially at school) will probably have your full name, maybe hopefully just "First Last" and not all 7 of them, everyone of authority will refuse to call you Johnny Awesome if your name is actually Johnathan Jean-Pierre Awesome-Livingston, and so on, oh and they will also fail to not butcher your name if it's not so easy a 4y old can learn it.

We can't be the only ones leaning more towards #2. And no, I'm not making this up, my go-to example is that I've seen cases where things like officially not calling "Bill Gates" "William Gates" have met resistance. Your name is your name, and I'm still not sure how people in the spotlight are able to be called Dick, I'm not joking.

here is one character made up of 11 others: https://en.wikipedia.org/wiki/Biangbiang_noodles#Chinese_cha...

this shows that new characters can be created not by inventing new strokes, but by simply combining existing characters to convey a new meaning, much like we occasionally do create new words in english by combining existing ones, even though that process in english is not productive, unlike eg. german, where it is quite normal. the difference is that these new words only have one syllable.

with the digitalization the creation of new characters essentially ends. the creation of the simplified chinese character system also pushes against creating new, more complicated characters.

it is going to be interesting to see how that will affect language development. new "words" can still be created by using a sequence of characters, but that means that each character keeps their syllable sound. whereas new compound characters would have a single syllable. so if a new meaning emerges for a syllable, a new character can't be created for it. will this prevent new single-syllable words? or will it lead to multiple characters being pronounced with a single syllable?

But a major caveat is that pronunciation can be wildly different when spoken with other dialects. Mandarin and Cantonese reading of the same text, even with same meaning, sound entirely different.

I have encountered even worse issues digital forms that only accept kanji (Chinese characters) or hiragana (syllabary designed for native Japanese words), the latter of which usually does not support certain voices that katakana supports. Ashley Tisdale, for example, is normally rendered as アシュレイ・ティスデイル (ashurei tisudeiru) - ティ is actually te with a small -i modifier, which does not usually exist with hiragana. Forcibly converted to hiragana, it turns into あしゅれい・てぃでいる - but ぃ is not accepted by the form, even if it exists in UTF-8. Your options are either converting the ティ into ち (chi) or て (te), neither of which are ideal, and may cause mismatches to other systems that properly support the katakana version.

The problem extends further into physical paper forms, where often they provide a very limited amount of boxes for characters, because native Japanese and Chinese names can easily fit within 8 characters. Combine this with the digital systems above and you're bound to have several versions of your name floating around on official documents all mismatching each other.

Some systems that need to print onto physical cards (e.g. getting a 1/3/6 month route pass on your SUICA or PASMO contactless smart cards) are even worse and turn dakuten (diacritics for hiragana/katakana) into their own character. As an example, the character ほ (ho) can be turned into ぼ (bo) using a dakuten, or ぽ (po) using a handakuten. The system will instead render those as two separate characters: ほ゛ and ほ゜ respectively, which cuts down on the number of available characters for the already limited textbox space you're dealing with.

The world is full of presumptions about names even today.

This happens in Europe quite often, even though many people have longer names.

Any idea if this is why, in Japanese-dubbed anime, the voice actors seriously mangle some English words/names? E.g., they often add a vowel sound to the ends of English words that should end with a percussive syllable.

I.e., do you think it comes from those words/names being written in katakana or hiragana in the dialog scripts, and those systems just can't express the correct pronunciation of such English words/names?

I don't know anything about anime, and little about Japanese, but I think Japanese (and Chinese) have a fairly strict consonant-vowel form for all their syllables. That makes foreign words that have runs of consonants or do not end it a vowel hard to pronounce, so speakers of those languages have a tendency to insert extra vowels to make pronunciation easier for themselves.

It's kind of like how English speakers will usually change the Pinyin "X" (as in Xi Jinping) into an English S or SH sound when they try to speak it, because the actual sound doesn't exist in English.

It's not really that the written language makes the names hard for them to pronounce, it's that the spoken language doesn't make it easy, and there's probably not enough care to try to pronounce them. Where the written language does make it hard, it's usually when people try to localize Japanese media into foreign languages, and the intended references in names are lost because of the mangling process of transcription into katakana.

Every person has a certain limited set of consonants, vowels, diphtongs, triphtongs, tones, and even syllables that they are able to recognize and reproduce. This is something you can train to recognize more, but you will probably never be able to pronounce or even distinguish the totality of all those used in all languages, even just the living languages on Earth.

Even if you did, there is an added complication that some languages actually used multiple sounds interchangeably, and explicitly distinguishing them may actually confuse you. For example, most European languages recognize various consonants as the same "R" sound, even though they are vastly different (French R is a back of the throat trill, Italian R is a trill near the palate, and English R is articulated next to the palate without any trill). If you come from a language where these are distinct sounds, you may have trouble understanding that two people who use different R sounds are pronouncing the same word.

https://www.youtube.com/watch?v=J_HLY0Rss-g

Second, in Japanese dubs these words are not usually actual English words, but Japanese words originated as borrowings from English language, so voice actors don't actually mangle them, the same way as English speaking people don't mangle the word "coffee" as they usually pronounce it, despite it being different from how Italians pronounce "caffè".

I mean, I don't expect you to bow to me.

But at the same time, the software I produce at work is usually entirely consumed by americans who speak english (ok, well, there's one canadian customer that I'm aware of). Because that's who pays for it, and none of those customers is particularly looking to pay for translation.

And the software I produce during my off hours is generally meant for me and my friends to consume. I'll put that on github/gitlab/source hut and you can use it if you want, but I definitely don't have the budget for translation either.

but it's especially stereotypical for mainly the US, China and I think Japan.

Which are all countries where to due to various reasons (size, culture/nationalism) there are a lot of people doing technical decisions which: 1st only speak the countries language, 2nd have little interaction with very different cultures

(in the US it's complicated, they have a lot of mixing other cultures into them, but do so in a very very US specific way with a lot of unaware cultural appropriation (and I don't mean this in the "bad/evil" way it's often used today but the cultural normal way) and the US is so large that there is little reason to make a trip to a country which is very different, and even if they do so, it's often in a form which is very touristy. This leads to situation where e.g. US citizens claim they are Spanish because of some ancestors and claim to practice Spanish culture but they are 0% clueless about actual Spain even after having traveled there twice or so. Contrast this with the EU where e.g. spending on study semester in another country which does have a completely different language and culture isn't rare, and non touristy holiday trips to other countries are common too (I mean in some cases it's just a few hours by care) and it's very easy, to have people which just don't know better.)

And maybe this is a form of Stockholm syndrome, but to this day, I don't really mind sticking to lower ASCII - it just makes things easier (or at least until recently it did), and I don't really care about that '³' in my surname. Sorry, I meant 'ł'.

Possibly related to that:

- I always stick to US English language when using software, even if it offers Polish, because I don't trust translations. They're usually done by people who don't have enough context and knowledge to do it right. I've been burned too many times by this. Plus, localized error messages hinder searching for solutions.

- I wouldn't mind if everyone switched to English[0] as first language and called it a day; there would be tremendous economic and social benefits from that to everyone, far outweighing the loss of a little bit of cultural variety/noise.

- I'm strongly in favor of meeting machines half-way. LLMs aside, it's trivial for people to learn a small controlled vocabulary here and there (like e.g. "OR" and "AND" and quotes in search queries), allowing to make interfaces vastly more predictable, reliable and comprehensible.

--

[0] - Or French, or Chinese, or Swahili, neither of which I know - to stave off the usual replies of the "you only want it because you already know English" kind.

But I agree with you: I don't trust translations and I think the benefits from humans using a single language would be amazing.

Germans know these official rules and maybe linguists, but if you present the typical English speaker with Möller vs Moeller it's confusing. Look at the media (who could, maybe, do some research?) who write Jurgen and not Jürgen. That's my point, the official rules don't help if everyone ignores them, for whatever reason.

Of course, the computer's native language is English - anything else would be silly. This may be a generational meme: young people have French language environments - even most of the computing professionals... But I keep my habit of mixed locales with metric measurements, English-language UI, mixed ISO 8601 and French dates. God bless UTF-8 though !

So I got used to using US keyboards also, and never using accented letters even though they are frequently used in Italian, e.g. instead of writing "Mario è alto" (Mario is tall) I would write "Mario e' alto". It helped a lot that I worked for almost two decades for companies having only non Italian clients and all communications internal and external were in English.

Now that I am working for an Italian company with Italian clients, I am slowly getting used to Italian keyboards and accented letters.

Yes and no. French sure, but in my experience of trying to learn Japanese, the difficulty is insane, much higher compared to the "western" languages.

õ - 6

ä - 2

ö - 8

ü - y

https://en.wikipedia.org/wiki/Arabic_chat_alphabet#Compariso...

ä => ae

ö => oe

ü => ue

ß => ss

but one gotcha is that it's a fuzzy one way trip, some words, especially city names, can have a ae,oe,ue in their correct native spelling. Worse ss is a normal language building block in German which is pronounced differently then ß so writing it as ss is quite confusing for anyone which doesn't happen to know that it's correct spelling is with ß. To top that of some cases of ß spelling have officially changed to ss spelling over time due to people anyway pronouncing it more like a ss and getting it wrong all the time. And to some degree ß is semi-official abandoned by now.

The reality is that there is a limited char set that is actually understandable at an international level, and it's not that different from ASCII. Even with paper systems, if you go to Rome to sign into a hotel and you give your name as 依诺, they will not be able to even write it down, nevermind pronounce it. And even if you tell them it's Yī Nuò, they will likely ignore the accents since they won't know what those mean. Similarly, if you go to China and say your name is Sângeorz-Băi, they will not know what the diacritics mean and will not be easily able to write them down.

In all times in history, when multiple cultures interact, they have to find a common subset of their languages to communicate in, and that includes names. The situation in writing is actually much much better, even if limitted to ASCII, than it is in actual spoken language. Maybe you can write my name down perfectly (Simionescu), but I would bet you won't use the proper pronunciation unless you happen to know Romanian - you will likely use different vowels and consonants.

But none or at least very few of the symbols missing from ASCII are actually broadly understood by people from more than a handful of countries (which can still mean a billion plus people in the case of Chinese, Arabic, and Indian scripts, of course).

Probably Arabic is the biggest counterpoint to my claim, as there is quite a large array of countries across two continents that recognize it. However, even there, there are far more people in countries which use Arabic writing that also recognize Latin letters than the other way around.

The many diacritics used by various European languages are definitely NOT something that has any wide adoption or meaning. Perhaps only the umlaut sign and the accent are even used by more than one or two European languages.

So again, my claim is that any system of writing that is intended for global international communication will have to restrict all names to the A-Z characters in ASCII with spaces as separators (and perhaps 0-9 and a few other characters that would anyway get ignored). Nothing else will work if people around the world are supposed to recognize the name in some meaningful sense. And relying entirely on automated OCR is a no-go for many use cases.

And just like people who interact with those outside their cultures have to accept that their name will be pronounced in a myriad of ways, they have no reason not to accept that it will be written in different ways as well.

fun fact: some of the symbols included in ASCII were intended to be used as (non-spacing) diacritical marks, specifically the tilde/caret/backquote characters...

[too lazy to dig up a proper source at the moment but the Wikipedia ASCII article covers some of this]

https://en.wikipedia.org/wiki/Machine-readable_passport#Name...

its the only requirement because

1. it's only meant for OCR

2. it's clear that it won't be used by only OCR but at least also human interacting with the system, potentially phone calls passing this information by voice, and anyone who can't pronounce the original spelling. For fairness if you e.g. didn't sing at all in your life are somewhat tone deaf and now are expected to pronounce a asian you probably have to spend days or more until you can do so (just as a extreme example).

Take a look at the spec if you are interested: https://www.icao.int/publications/documents/9303_p3_cons_en.... The transliteration tables start on page 24 and are scattered throughout the document.

Newer versions of the transliteration encourage stripping diacretics, so that would be HAMALAINEN. Much more readable to native speakers, but obviously loses information.

I wouldn't recommend it as there are official tranformations from the countries which uses diacretics and they most times are not to strip them. It's kinda another case of people forcing stuff onto other cultures. And if you do bussiness in some of the countries in some industries you might even get into legal trouble if you apply that.

> Section 6 of the 9303 part 3 document specifies transliteration of letters outside the A–Z range. It recommends that diacritical marks on Latin letters A-Z are simply omitted (ç → C, ð → D, ê → E, ñ → N etc.), but it allows the following transliterations: [...]

You said

> you might even get into legal trouble if you apply that.

We're talking about passports, this seems not relevant. For passport-related use such as travel, you use the form of the name written on the passport, exactly as-is.

And software in those countries always has the "English name" field.

Oh… where do we start… We do not have to go as far as finding an intersection of multiple languages. Consider English as an example. I have written up a fictitious but a reasonably real dialogue between:

1. A layperson from a lower socio class, who has not attained high educational levels and speaks English using vernacular and predominantly Germanic vocabulary of the English language.

2. A state citizen of the upper-class descent who speaks English almost exclusively with Latin/French/Greek-derived vocabulary.

This layperson complains about not receiving a welfare payment from the state.

Layperson (L): Oi mate, I ain't got me geld from the state yet. That's daft, ain't it? Every man's got a right to his share, right?

Educated citizen of the upper class descent (U): Pardon my incredulity, but are you referencing the monetary allocation designated by the government for individuals of a particular socio-economic standing?

L: Eh? Oh, you mean the dole? Yeah, that. They owe me, but there's no dosh in me pocket yet.

U: If I interpret your sentiment correctly, you are perturbed due to the delayed disbursement of your financial entitlement. Have you endeavoured to communicate with the pertinent authorities?

L: Talk to who now? Oh, you mean the blokes at the town hall? Aye, but they keep spieling some rubbish. Can't make head or tail of it.

U: My advice would be to liaise with the relevant office, elucidate your predicament, and seek resolution. It is paramount to ensure you have met all requisite criteria for the stipend.

L: Right, so you're saying I should have a natter with 'em and make sure everything's shipshape? Just want what's owed to me, y'know.

U: Precisely. Engage in a dialogue with them, ascertain the cause of the discrepancy, and ensure you have fulfilled the necessary prerequisites for the allocation. You deserve your due compensation.

L: Cheers for that. It's a bit of a muddle, all this, but I reckon I'll give it another whirl.

U: I wish you fortitude in your pursuits. If there is an inherent right to such financial assistance, it is imperative you receive it posthaste.

Even though the layperson does understand responses in the fictitious dialogue, that would not be the case in real life. Both speak the same language, yet the responses are generally incomprehensible to the layperson.

…which I’m sure actually happens in real life.

Which is also true.

It was a thought experiment to highlight the fact that the lack of comprehension could also arise within the boundaries of a single language. In linguistics, the term for this specific phenomena is «the social register», and there are plenty of active and thriving language that employ the social register in the daily speech. Korean, for instance, is renowned for having a highly complex system of the social registers (effectively, parallel vocabularies) embedded in the spoken language. There are other languages as well.

> If the "educated" citizen is truly educated (and not just upper class), they ought to not only understand the lay person, but chose lay words in return.

And that is also true. Social registers have largely disappeared from mainland European languages, yet an English accent and the choice of the words of an English speaker can reveal sufficient details about their socio-economic background.

Maybe there’s simply the COBOL equivalent of the following somewhere:

  if len(first) + len(second) + len(last) <= MAX_LEN_ON_TICKET:
    first_name_on_ticket = first + second

I asked a checkin agent to fix it but they said it will start rejecting my ID if they change it at all.

I don't remember the precise details, but some airline website's password had restrictions at one point that made it super-obvious that they were internally converting alphanumeric passwords to digits based on the US telephone key mapping.

I remember thinking at the time it might ultimately have been due to SABRE (because I believe that's literally one of the oldest computer system still in use), and screen-scraping some telephone menu system depressingly seems like something someone would do for expediency.

I wouldn't be surprised if a system like that also mangles names.

> Usernames and passwords containing letters need to be translated to numbers to enter them in a Fidelity phone system (like FAST, or if you call a representative). Use your telephone keypad to convert the letters to numbers. There is no case sensitivity. Substitute an asterisk (*) for all special characters. https://www.fidelity.com/customer-service/need-help-logging-...

This is completely normal for official documents in Germany and it makes sense for us.

Technically we do not have first, second or any other numbered names. Our given names form a set in the mathematical sense and any one is equally valuable. This comes from the tradition of given names being given by godmothers and godfathers and we wouldn't want to get into the issue to ever have to value one of them over another. At least this has been the case in some parts of Germany and has influenced the official regulations for names.

Of course the names have to be put into an order on your ID and to keep things simple banks, schools, authorities, etc. ask you to use that order on their documents.

Traditionally, official documents just used the surname with "Herr" or "Frau" but nowadays they often use just the given name in first position on your ID.

If never heard of a "First Second" case with one exception:

Given names can be connected with a dash. In this case the order is fixed and the whole unit is treated like a single name. While in principle arbitrary names can be combined there are certain very common combinations, like "Hans-Peter", "Karl-Heinz" or "Franz-Xaver". If you happen to be named "Hans Peter" (without dash) it's likely that they assume the dash and will call you "Hans Peter" or "Hans-Peter" all the time.

There is a very mild version of that in the US -- lower likelihood of blind assumption, but still present: when a set of two given names starts with "Mary" or ends with "Ann/Anne". Examples include "Mary Jane", "Mary Kate", "Jo Ann", and of course "Mary Ann". Some have simply merged into single names like "Maryanne" and "Joanne" more recently. There are probably others.

Of course, and I mildly apologize for my case of Whataboutism because I actually described the reverse. They're taking the rules too literally and are using the thing they need for official documents everywhere (their marketing/status emails).

I'm just just kinda puzzled why they'd think it's a good user experience, especially for people who are not just not used to reading their government id name but actually uncomfortable (i.e. pending name change).

I think you mean “built for us, and meets our needs”. It’s not the US’s problem that other countries don’t necessarily take innovation risks, but instead buy our old stuff.

ASCII on the other hand derives much more from communications equipment (telegraphy) than IT gear.

¹ https://en.wikipedia.org/wiki/Punched_card#IBM_80-column_for...

But it was also an "8-bit" thing and a "extremely limited computing resources" thing. EBCDIC was designed in 1963/1964.

I mean, when you've got 8 bits to represent a character, and there are more than 256 possible characters... what do you do?

A truly robust solution like Unicode would not have been feasible with the resources of the day, and even a "simple" 16-bit scheme would barely be able to contain all 50,000 Chinese characters.

The blame here lies with the Dutch bank who willingly chose an EBCDIC solution in 1995, although I'm sure they were dealing with various constraints and pressures as well.

You are wrong. It was rather a poor decision on the buyers' side: "It does NOT work for us, but who cares?".

Most of the early work was in English and the people buying these systems all understood English. Nobody back then had a problem with a lingua franca for aspects of tech because there were still people around who had to learn German to study science.

Get used to it.

Once the government realizes it can shift the burden of enforcing a law onto the private sector, it never un-discovers that. It's a ratchet.

I worked in a similar national organization. I fought for having a newly built systems in 2015 created in anything different than cp1252 and ISO8859-1. I lost. The architects flat out forbade UTF-8 and required 1252, in the name of consistency.

The only place where the difference between 1 letter and 2 letters still is visible is in names that start with a long ij, because of the capitalization. E.g. the name of the city IJmuiden is written like that and not as Ijmuiden.

In handwriting it's written as one character, but I think that's just a ligature: https://nl.wikipedia.org/wiki/IJ_(digraaf)#/media/Bestand:IJ...

Also, when spelling words people use 'ij' as a single letter for instance a Dutch person would spell 'mijn' as 'm', 'ij', 'n' (and they would likely not say 'lange ij' because there is no word 'mein' in Dutch.

It also causes no end of trouble with storing and searching because a Dutch person might expect IJ to be sorted after 'X' but instead it appears sorted as the combination I J . The fact that different reference works use different methods doesn't help either.

Not anyone under the age of fifty, unless they want to deal with the follow up question too. ("Is that a long or short ij/ei?")

Here's a screenshot from just one random Dutch children's spelling quiz online:

https://postimg.cc/47wFHy0v

Separate letters for 'ij', despite it formally being a single letter.

Because of that, my grand father and his sister happened to have different family name in their ID cards: when they got registered by the priest at the beginning of 1900, on a paper book of course, the priest who registered the sister added an "I" at the end of the family name as he thought that that should be the right spelling in proper Italian. At that time ID cards did not even exist so nobody bothered.

e.g Papadopoulou vs Papadopoulos.

Then there are the cases where a Greek wife takes the male surname form to co-exist more easily in other countries where it is expected for a husband/wife surname to match (or if different, be more considerably different).

Czech has this too, sort of, where "ch" is considered one letter but it is composed of two separate characters.

Here is the full alphabet

A B C Č Ć D Dž Đ E F G H I J K L Lj M N Nj O P R S Š T U V Z Ž

The language is unfamiliar to me, and quite intriguing. I find the pronunciation of Zuid very counter-intuitive.

More than you ever wanted to know about 'IJ':

https://en.wikipedia.org/wiki/IJ_(digraph)

I still have a typewriter that has it as a single letter.

Yes, this can lead to bugs in software, but so can anything related to names.

It is deprecated in Dutch to use a single codepoint for the IJ. That was never really an option in any of the character encodings in popular use.

The fact that it is one letter is relevant in cases like (vertical) lettering (which most designers nowadays fuck up), in typography (the number of fonts which make ij look awkward and unaligned is huge), and in collation and sorting using a Dutch locale. I will defend its proper use and treatment where possible, but representing it as a single codepoint is not a sensible goal, and never was.

I do not believe this to be correct. Wikipedia says it's a digraph of two letters. It does say that the codepoint is deprecated, but it's only defined as "compat", not as deprecated in the unicode data.

If you dig deeper on the Unicode website you'll find that the reason those codepoints are included is compatibility with 'certain very rare legacy (non-Unicode) character encodings'. They are not 'deprecated' as compatibility characters for those old legacy encodings, but 'deprecated' as suitable for rendering Dutch text unencumbered by those early code pages.

If you are claiming that 0x0132 is a codepoint in common use or required for correctly spelled Dutch, you are mistaken.

I made no comment in support or opposition of that. I cannot talk to that as I’m not familiar with either Dutch or that letter. However there are many compat characters in Unicode and there are incredibly few deprecated ones so I was addressing the deprecation claim (and what I believe is a misuse of the term letter). You’re interpreting things into my replies that just aren’t there.

Compat characters are very useful and even if they are not stored all the time, they often show up in text processing in memory for better glyph selection.

https://unicode.org/faq/casemap_charprop.html#6a

> The Unicode Standard encodes these two compatibility characters [0x0132 IJ and 0x0133 ij] to provide support for roundtrip conversion of the Dutch letter 'ij' in certain very rare legacy (non-Unicode) character encodings. It is strongly preferred (and far more common) to use the two character ASCII sequence 'ij' to represent this letter instead.

You can dig in the Unicode mailing lists for discussions on this from over twenty years ago. The bottom line is that you shouldn't use 0x0132 and 0x0133 in modern text. By now this is a resolved issue.

The 'IJ' is a letter, culturally speaking, in the sense that it is capitalized as one, and that it is often rendered as a single unit (which you can see in vertical lettering, if done right), and sorted as if it was a single letter. In terms of character encoding however, it is a 'i' followed by a 'j'.

Cf.:

* https://onzetaal.nl/taalloket/ij-plaats-in-alfabet (first sentence!)

* https://taaladvies.net/ij-alfabetisering/

See my other posts for 0x0132, which is a compatibility character for rare obsolete character encodings which did encode it on a single codepoint.

Swiss story 2: I was outvoted having an application properly internationalized, they dumped instead all resources in extra database columns because "this country will only ever need three languages". Fast forward, it took one year to get it in resources because somebody high up decided English and Rumantsch wold be needed too and to rewrite all those language-related SQLs turned out overkill.

Yeah, some people think about their heritage as not compatible with the brave new world, so get rid of it. I know people giving their child dedicatedly ascii-only name common in the USA to make sure their children won't have a "weird name" when going to the USA.

It is a debatable practice, yet I think this is somewhat protecting the children from silly decisions of parents (some example from Hungarian tabloid media where parents were outraged: Fradi / Fradika (from the slang name Fradi of the FTC football club, without and with diminutive appendix), Traktorka (tractor with diminutive appendix), various names from soap operas, WoW characters, Satan.

So think of the children! :) But really, this is the other end of the spectrum. And this (giving especially weird names to children) is a fine recipe for making the child subject to bullying in my opinion.

Australia seems to have a good policy where any name is allowed, as long as it isn't offensive or too silly.

Musk's children's name would be fine, for example, while calling a kid Hitler would not be.

Also we have the idea of "nameday" which is minor occasion for celebration and giving small gifts every year, based on the person's given name, and the calendar contains the mapping between names and dates, one without an official name would be in trouble getting that free a bottle of wine :D

IMO only the names that would be damaging beyond a reasonable doubt should be prohibited.

Also, the nameday idea sounds cool :)

Even the Icelandic singer, Björk sometimes had her name pronounced more correctly by DJs. I guess the diacritic was a big clue that the vowel is not pronounced like the ‘o’ in “fork” (the most common way her name was pronounced) and pretty much everyone knew that the ‘j’ sounds like a ‘y’ in English.

And? That's what I did - a simple name that he won't have to repeat, spell out for English speakers, not have to spell out for many non-english-speakers, or have trouble non-latin scripts.

Anything you can do to make your child's life easier trumps any value you think they might get from maintaining cultural or traditional links with a mostly dead past.

Even better, when I was a kid and I got my first official document, the national healthcare card, the software in use did not allow a space character in the first name. The operator then decided to add a hyphen to enter the two parts of my name as separate words.

Fast forward many years, and I my first name shows up with the hyphen on the ID card and healthcare card, while it does not contain the hyphen on the driving license, which of course I got much later when software had improved.

Generally this is a non issue, nobody ever said that it's not me because there is a hyphen or a space in my name, however when I signed the mortgage for my house, the layer asked me to sign with the hyphen, and in the document he wrote both variants of my name with an A.K.A. clause.

On virtually every airline ticket for which I provided my first and middle name it was just concatenated.

For example: I would provide

Frank John Sample as my name it appears as

FRANKJOHN SAMPLE on the ticket

Old reservation and ticketing systems, indeed.

Lufthansa for example does this in a particularly annoying way, if I give my name as LAST NAME, it will automatically smush it to LASTNAME. However, if I then want to retrieve my boarding pass, it will only find my booking when I enter my name as LASTNAME, because the look-up does not smush things automatically.

Four names isn't really a lot; some German aristocrats (or their descendants) have 6 or 7 names, and it's quite customary for Arabs to list a shedload of their ancestors in their name (e.g. Ahmed bin this bin the-other bin whoever).

> the layer asked me to sign with the hyphen

s/layer/lawyer/

That's nuts. If a signature is anything, it's the way you customarily write your name. Fortunately my signature is unreadable, and nobody could tell whether I'd written a hyphen or not. I suggest acquiring worse handwriting (good handwriting is almost useless these days).

The lawyer even insisted that the hyphen had to be clearly visible in the signature. I think that in those types of documents in Italy you are legally required to use a readable signature, because I bought/sold house a number of times as family grew, and every time different lawyers always insisted on this aspect.

My last name is spelled with a double-s by one side of the family, but with a single s by the other. I've been refused to pick up packages at the mail-parcel centre only because the spelling on the parcel did not match my ID card, despite having a notice slip with delivery number that had been delivered correctly to my physical mail address.

One thing is if the database can handle Unicode, but what about employees? They now need to be able to differentiate Chinese characters as well as Sanskrit, Hangul and Thai script?

Of course not.

Airlines have pretty paved the way for everyone having an ascii encodable version of the name and this is so standardized that it’s even in your passport.

I find it laughable that airlines are given as a shining example of operational excellence.

Not even an English/US thing, this is the Latin script from Roman times.

Edit: what do you suggest as a universal alternative for bit-perfect verbal communication... Spelling out Base64 Unicode over the phone?

The Latin alphabet does not have a single pronunciation. And the vowels are ambiguous across the most common languages so you can't even make an argument that they are intelligible.

You've really lost the plot on this one.

The clerk answers in and speaks Classical Latin.

Not everyone in countries which use the Latin alphabet can read or write it (called analphabets or illiterate). And there isn’t a single way to speak it, either, pronunciation depends on the language. So I’m not sure I buy there aren’t people who can read and write only in a non-Latin alphabet.

I have no clue what you mean when you say everyone can hear the Latin alphabet. Listening to sounds has no relation to the language spoken. I can hear Korean just fine, doesn’t mean I understand the meaning of the words or understand their alphabet.

> Not even an English/US thing

Even languages who use the Latin alphabet have diacritics and characters not present in ASCII.

And even if they know how to type in all 52 letters, it's absolutely not a given that they are able to transliterate from their native script to the English alphabet.

"Everyone, everywhere" does not have a "Latin" alphabet as the basis.

It's not even standard for English: there's a UK keyboard that is organised differently and has different characters.

- There is no such thing as a "standard keyboard"

- Not everyone can type Latin alphabet

- Almost nobody can read, speak, hear or write latin, and it makes no sense to "hear latin alphabet". How do you pronounce "Bordeaux" using only your knowledge of the Latin alphabet? How do you pronounce "Queue" using only your knowledge of the Latin alphabet? You realise native speakers of various languages will pronounce words (for example "pain") and even letters (for example "w", "j", "y") very differently? Etc.

- Not every country is related to the Roman culture.

Your statement isn't even remotely true.

It is not a technical issue but a social one.

Also, let's not forget the context here, we're talking about a bank not a restaurant

Because sometimes other people have to read it. Do you actually expect your bank has someone who can read Νίκος Καζαντζάκης and 刘慈欣 and ᠲᠠᠲᠠᠲᠤᠩᠭ ᠠ and משה ברבי מימון הספרדי and ᐱᔭᐃ ᐊᕿᐊᕈᖅ? (Or even notice which one Chrome doesn't render correctly?)

That's even more absurd. So now, just because I don't speak Thai, I should reject all Thai customers, since I can't read their names?

I really don't see what's absurd about that. This website is so English language biased it's hilarious.

Btw we're talking about banks not Starbucks or mom and pop shops

As is most international websites. And systems. And content, in general.

It's a sad fact of life that, as we sail into a globally-connected future, the world is going to consolidate on a small number of languages, and the majority of languages are going to be left behind, discarded, and eventually die out.

People want content. They will learn whatever language gets them the most content. Right now almost all content is produced in a small handful of languages, with (in the west) English being dominant.

At this point it looks like English[1] is going to be in that small set of surviving languages.

It's inevitable. Railing against it is a pointless waste of energy.

[1] My home country has 11 languages, all official. Until widespread internet arrived it was common to find locals who could not speak English. Now, I'd be hard-pressed to find non-english speakers, even in the very outlying areas.

Sure, as long as you are prepared to be sued for discrimination.

The point of identification is to identify people to some level of trust. In the vast majority of cases, that means that I should know that you’re the same person that I did business with yesterday.

Airlines need to tie you to an official document, which is much more complex. They do a pretty decent job at it considering all of the stakeholders who make that happen.

The hang-wringing about American hegemony is a projection of some other nationalist feeling. The constraints of Hollerith cards made it difficult to accommodate different character sets and the accommodations are codified in international treaties and business process. It will improve over time, probably first in the more cosmetic CRM side.

Accepting customers and accepting character sets or glyphs are two very different things. Accommodation is a two way street - if your name on the Starbucks cup is in Arabic or Greek, the barista in the US isn't going to be able to call it out. That not because the barista is some ignorant rube, they just don’t speak Greek.

The magic is we have a global system where many people have the ability to step in a plane and go almost anywhere and immediately conduct their business or pleasure with minimal friction. One of the friction points are issues like character sets, or poor accommodation of long names, etc.

I think parent makes a point about employees. You can argue all you want that it is reasonable to accept and store any unicode character, but it is in no way reasonable to expect that the person on the other side of the glass to know every single script in the world, and every single unicode character.

Your position is reasonable for those people entering their own name, on their own device, which is configured to their own language - a system should not fall over on that - but that is not the use-case presented by the parent.

> I find it laughable that airlines are given as a shining example of operational excellence.

They're certainly doing it better than anyone else, IME.

The use-case is "employee has to enter the name, on their device, with their keyboard". What's your better alternative, the one that makes you use words like "absurd" and "laughable"?

After all, this use-case is not going away anytime soon.

Whether we like it or not, having a name with uncommon characters is going to make your life difficult in those cases where your name has to be entered on a device and keyboard that is non-native to you.

The airlines get their data feeds from the travel reservation systems – the ones we (or the travel agent) interface with via the airline website or a dedicated web portal. There are two major global ones, Amadeus and SABRE (there are other ones as well). I do not know how any of them interconnect, though.

There have been numerous attempts to modernise both (including rewrites from scratch), and all the attempts have failed so far due to the complexity of the logic: calculating the most optimal travel time for connecting flights in a multi-leg trip is, like, very hard, apparently + applying a correct airfare prices based on the selected itinerary and a myriad of other variables. Subsequently, there has been the decades' worth of growth of the said logic – there was an epic story on here some years back about one such endeavour (I think it was an Amadeus rewrite).

From what I remember, the character encoding was not even considered to be a problem in that endeavour – as in «acknowledged, is trivial to solve, now let's move on onto the actual problems».

Naturally, maybe the goal is just to get this train moving somehow.

No. But it's one thing to not support Japanese the other not to support apostrophes or accents in text written in latin characters

But guess what was also incompatible with this? ASCII. Hence the é things in HTML

You can fix this if you encode it

So I suspect that the ruling really is that the bank should use a system that allows to correctly write the country's official languages, which seems quite reasonable.

Now under EU laws this would probably extend to EU languages, with the likely caveat that, I think, different alphabets are expected to be transliterated into the local one. E.g. Greek names are expected to be transliterated into the latin alphabet in Western Europe.

I am pretty sure that there is no expectation that people can use their names written in, say, Chinese characters as that is not reasonably legible.

You can just put a note on the account: "Actual name contains accent acute over second e of first name: our system is unable to render this. User is quite sensitive about this so apologize again for our inability."

Not every organization would require this, it depends on the processing. A bank would since it sends you a bunch of correspondence. A restaurant taking a reservation wouldn't if they wouldn't save and sell the information or use it for marketing.

When it comes to the amount of ass-pain this can cause, that's just the tip of the iceberg.

Sometimes it's funny though. At hospitals, there's an automated system that announces the next person by name (GDPR? Privacy? What's that?). For non-Chinese names, it reads the name letter by letter. So now everyone in earshot knows there's a foreigner around somewhere.

But the security problems are real and very annoying, even with a "normal" name. For some silly historical reason, my middle name has two different spellings depending on which official document you're looking at, which causes no end of problems. Conversely, I know people with very simple, common names, the kind where two people born in the same city on the same day have the same name, that suffer a load of different but equally frustrating issues.

These problems, however, aren't just limited to computer systems [0]. Border officers, bank clerks and government officials and other general bureaucrat types are all just as bad as a strict strcmp implementation.

The issue is that many people operating in an official capacity seem to work under the assumption that a name is a globally unique identifier that has exactly one representation, but that's not how names work in any culture.

On the other hand, if not names, then what? Combined with birth dates, it's the closest thing we have to a globally recognised unique immutable identifier, and any better alternative is going to feel invasive and face a lot of opposition on privacy grounds.

[0] https://www.youtube.com/watch?v=nq-dchJPXGA

Ask anyone with western name who moved to Japan (or a myriad of other countries, but I am in process of doing the same right now so it hits close to home) whether the state has a single source of truth for their name.

It may be the source of truth, but it's not a very good one.

The USA doesn’t really have a civil registry. https://en.wikipedia.org/wiki/Civil_registration#United_Stat...:

“In the United States, vital records such as birth certificates, death certificates, and frequently marriage certificates are maintained by the Office of Vital Statistics or Office of Vital Records in each individual state. Other documents such as deeds, mortgage documents, name change documents, and divorce records, as well as marriage certificates for those states not centralizing these records, are maintained by the clerk of court of each individual county. However, the term 'civil registry' is not used.”

I think that means your marriage may be documented in county C1, your subsequent divorce in county C2, and your name change in county C3, none of which need be in your state of birth.

In Germany, it is intended to be hard to change your name; in the USA, the situation is different. Also in Germany, if you change your name voluntarily (without having a very good reason (CLARIFICATION: of course marriage is a good reason)), this is considered to be a strong sign that you deeply hate your parents (and did the name change because of that). Indeed, the people who I know who changed their first name to their middle name exactly did it because if this.

This is in fact one of the iconic places where lack of diversity in software teams cause problems. Men may have a blind spot around name changes that women don’t, because of their lived experience.

Huh? I’m German and I have never heard of this.

I am also German. The only people (in Germany) who I know and did a (forename) name change did this because they hated their parents, since changing the name is an open signal that you deeply hate what the parents did to you your whole life (including giving you your old name).

But they're the closest thing we have that would be acceptable to most people. Better suggestion?

When asked "So what is the correct spelling?" he snarkily shot back "You tell me."

It's not a standard western name. It wasn't originally written I'm English. Attempts to translate it to English -- well, they are guesses on how to spell what someone thinks they are hearing from a language that the Roman alphabet isn't designed to capture.

Demanding that your name is written down in a specific way is just stupid.

I always found it elitist when people insist on a specific spelling. Typical examples are "Philip/Phillip/Philipp", "Stefan/Stephan", "Harald/Harold", "Erik/Eric/Erich", "Michael/Michel/Mikail" and so on... all these variations refer to the same name. Different dialect, different language, the spelling varies. So what? It's still the same name!

IMHO, the cited Article 16 doesn't necessarily demand a rectification of the spelling. This is about meaning, not syntax. But that's surely up to interpretation.

Will a bank in a country that doesn't use the Cyrillic alphabet agree to write my name in Cyrillic? I'm 99% sure no. So, organizations fully supporting the writing system of the country in which they operate is a reasonable expectation. But I never expect anyone abroad to agree to write my name in an alphabet they can't even read.

Tangentially, Turkey renamed itself in English into "Türkiye". This suffers from the same issue — there is no letter ü in English. No one knows how to spell that. So it's no surprise the new name didn't stick.

Then there is the issue where women’s surnames are conjugated into a feminine ending, so women have similar but different last names than their husbands. The US didn’t like this and didn’t understand this practice for a long time. I know a few women that have dual citizenship and maintain a passport for each country with a different spelling of their last name. Updating documentation in one country to the other causes all sorts of issues with this setup (since everything requires a translation via an apostille). My friends recently got married in Greece and then separately again in the US at a courthouse to avoid all of the paperwork and translation pains from the woman’s Greek maiden name differing from her American maiden name.

Many of these issues are easier to deal with now, but they have a long history of confusion and extra effort going back more than 100 years. Safe to say, I’m relying on documentation from my other Greek grandfather where things are more straightforward for pursuing my citizenship.

U.S. concept of automation is that a cheap employee does something in the back office..

Every Turk, Russian, and Serb enters the chat.

Transliteration without deterministic, reversable fidelity sucks.

PS: I have a "normal" given name people spell 4 different ways and surname people spell 3 ways. ]: I always spell them out. Just be glad your last name doesn't have 4 "y"s as their legal American name like a former coworker.

This problem is too well-known to everyone who have their original docs written in non-English alphabet (read - everyone with diacritics or cyrillic script or hebrew or heaps of other non-Latin-based writing systems).

To add insult to injury - this is not even about some common transliteration tables from non-Latin to Latin script. Sometimes? rules of transliteration are changed based on local govt whims and if you got caught in such bureaucracy - all hell let loose. You can't obtain a document with previous "latin spelling", because it's changed in a system. This way you may have a ticket with one spelling and passport with another. Or, like in my case, my brother and I have different surname transliterations to English. And my first name would be spelled with 4 different letters than it's in all my other documents if I have to issue a document from scratch. Good luck explaining all this nonsense to anyone, who is trying to "character-match" someone's hard-to-pronounce surname from official document. Even worse when they would try to re-type any of UTF-8 characters without knowing a correct character codes or not having corresponding layout installed (most of the time in international airports and customs).

So, I'm personally fine with something is incompatible with GDPR, as long as certain systems would stay ASCII as long as they can. We'll be opening a whole new can of worms if internationally used documents would be in UTF-8, imo.