Hacker News Reader

Comment by liotier | Hacker News Reader

In the early 90's me and a couple of other French guys had taken to writing French unaccented, so that we wouldn't suffer the daily pain of character set problems - we really bashed our own heads to fit into lower ASCII !

View on HN

Same here in Poland. I got used to skipping the diacritics, and quickly learned to mentally decode text that was rendered with the wrong code page.

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.

I guess Polish people had it worse than Romanians and Hungarians. All our accents are simple to strip away and they still kinda sound like the base letter.

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

The funny thing to me is how these inofficial transcription rules differ from country to country. Seems most people with ö or ü from a Turkish name are happy to drop it to o and u (cf Mesut Özil) but Germans are absolutely not. (Not saying this is a rule, just what I observed.)

In certain situations like cross word puzzle it wasnt usual to use "umlaut" but instead to write oe, ue and ae. In Swiss German people dont use ß. So transcription was always a thing to know.

For German these rules are offical. They're used in the machine readable part of ID cards and passports. If you start at a german company and they setup an e-mail address for you they'll use these rules too. The origin of the letters ä, ö and ü in German are ae, oe and ue, people just put the e on top of the vowel and they slowly transformed into what we use today. You can still see this in some names like Goethe.

I know that, but it's not what I meant, but probably didn't make it clear enough.

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.

German has an official and fully information preserving transcription to basic latin (which is really what all this talk about "anglo letters" is, just the basic latin letters in common use in the early modern era, with no diacritics at all), which can be used in official documents, too. Other languages, like latinized Turkish, obviously copied the diacritics, but seem to have left out the transcription rules, most probably because they borrowed the letters long after their history was relevant.

> I always stick to US English language when using software

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 !

Metric measurements, English-language UI, UTF-8 and ISO 8601 dates could make a nice international default standard locale.

Describes my custom setup exactly. It would be nice to have it as a default.

If it weren't for Honduras we could have en_HN.UTF-8.

I really, really hate how Amazon tries to force US customary units (inches, etc.) on me in automatic translation just because I set the language to English.

Same for me. I am Italian but I always used computers with English localization because translation was often bad, especially IBM translations.

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.

> 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.

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.

In early 2000s Estonia, I remember people using numbers to denote certain accented characters when texting:

õ - 6

ä - 2

ö - 8

ü - y

Arabic speakers have a comparable system:

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

The trouble comes up when an Arab needs to communicate with an Estonian verbally. How do they spell their names to each other?

I think it could lead to an accidental diffie helman key exchange.

in German there is an official convention for the non us-ascii letters, AFIK it predates computers and is rooted in germanic dialects developing differently and later stuff like non-german specific type writers, printing press machines, etc.

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 letters ß and ss are pronounced exactly the same, a hard 's' sound. Their effect on the preceding sounds is slightly different, though -- 'ss' makes the vowel short, 'ß' keeps it long. In all, a rather minute difference.

After learning that ß is the ligature of literally "sz", I am astounded that "ss" is used instead to replace ß.

Reply on news.ycombinator.com