This FSF form assumes everyone's full legal name can be written in ASCII (2011) | Hacker News Reader

This FSF form assumes everyone's full legal name can be written in ASCII (2011)

orgmode.org

20 pointsomaranto44 comments

This is a hugely misleading, if not outright clickbait, headline. The author of one random plain text form surely does not represent the views of "The FSF", even if they work for them.

I don't know if the change I just made to the title makes it better or not, but I think it at least captures my intent better.

Just file an issue. This doesn't need to be a news story.

I thought it was funny to have such an obviously impossible instruction in the FSF form and never once thought it would bother anyone. If it bothered you that I submitted this, I do apologize.

Completely clickbait.

Clearly FSF can deal with their maintainers having non-ASCII names. Have you talked to a human being regarding this?

https://github.com/gcc-mirror/gcc/blob/master/MAINTAINERS#L4...

I haven't. I gave them my legal name and an ASCII approximation of it (which just changes an "í" to an "i"), and I'll let them decide whether what to use. I doubt it will ever matter anyway.

> curl https://orgmode.org/request-assign-future.txt -si|grep last last-modified: Wed, 12 Jan 2011 10:19:50 GMT

Random text document from at least 10 years ago with a misleading headline, at best. How is this not flagged?

Spawning some interesting discussion, no?

It feels like this may be a legacy document. The actual root website has a Japanese version with definitely-not-ASCII characters. Seems like a simple oversight of not updating a form.

Now, if the administrators are still 100% behind this being the only way, then maybe there's reason to get upset.

It certainly has a huge 'legacy' look about it.

Using '-------' as a horizontal line. Using less than even 80 characters on a line. Definitely something that has its roots way before the 2011 it's marked as.

I'm amused, not upset. I tried to follow this impossible instruction as best I could and decided they probably cared more about ASCII than about getting my legal name, so I changed and "í" to an "i".

And even if this is not the official FSF standpoint they are right because passports totally do this.

Here's the transliteration of names for the passport of the Russian Federation http://transliteration.ru/mvd_1047/ for example. More recommended transliteration rules for names in machine readable passports by the ICAO in https://www.icao.int/publications/Documents/9303_p3_cons_en....

Is it correct to refer to those transliterations on passport as "full legal name"?

I would suspect that varies from jurisdiction to jurisdiction. It surely is a representation of it, to quote the ICAO PDF:

> Prefixes and suffixes, including titles, professional and academic qualifications, honours, awards, and hereditary status (such as Dr., Sir, Jr., Sr., II and III) shall not be included in the MRZ except where the issuing State considers these to be legally part of the name

Hungarian law for example says the passport contains an MRZ (compiled from various data) which allows identifying the citizen and the passport. Is that the legal name then? Is this a meaningful discussion to be had?

I can't find Canadian law but https://www.canada.ca/en/immigration-refugees-citizenship/se... says "The last two lines at the bottom of the page are the machine-readable zone, which repeats the passport holder’s personal information "

Workaround: Use the unicode description for every non-ASCII character in your name, submit a change request to change this text file to UTF-8.

While this might be an unpopular sentiment, I think ASCII name isn’t completely unreasonable.

In an effort to make it easier for people to communicate, we have designated English as the unofficial language for most open-source projects.

Most names can be romanized, and many people would have trouble typing 陳 or 石川 if they weren’t acquainted with Chinese or Japanese, and I think ASCII is a reasonable request.

I don't think ASCII is unreasonable at all here! But the instructions could be worded in a way that it is possible to comply with them, instead of this formulation which is technically impossible for many people. It could say: something like "Please use your full legal name as the subject line of the message if it consists of ASCII characters, otherwise use an ASCII transliteration."

Agreed

For the record, my name only contains one non-ASCII character, an "í". So what I did was change it to an "i" for the email subject line, and in the body of the email point out I could not follow their instructions literally, and also told them what my legal name is.

Yeah but it's kind of a whole thing to respect people's autonomy to control their own names. Plus, you can just copy-paste.

Edit: Not to mention if you are asking for their "legal name" for legal purposes then you would have to accommodate any valid legal name in any country.

Having not studied Korean or Arabic, I would have a real hard time with copying and pasting, and knowing whether the name is a surname, or first name, or whether those cultures employ a first name or last name.

(For example, many Asian cultures with Chinese influence start their names with LASTNAME FIRSTNAME, so one might have a name of 蔡英文, where 蔡 is the surname, and 英文 is the given name. Japanese also has the same custom, 武金城, where 武 is the surname, and 金城 is the given name.)

That’s why I think a harmonized approach—even if it has a bit of English hegemony—is acceptable, just as we have done so in scientific journals and publications.

But romanizing doesn't solve any of those problems (except maybe having to copy/paste, but with uncommon/foreign names you'll have to re-look at it thrice either way). You're still gonna wonder if 'Sung' is the first, second or third name, because your mistake was assuming that it works like this in the first place.

If you read scientific journals, they have dealt with ASCII names, sometimes with last names in all caps, with the given name.[0]

For example, KANESHIRO, Takeshi, with the names of the authors of a paper.

[0] https://reader.elsevier.com/reader/sd/pii/S0022247X9895987X?...

A better approach is just using whatever name they provide in full. If you need a short name (which is not the case in the original .txt file), ask for one. Just treat names like opaque blobs.

There's always a tradeoff between autonomy and practicality. Do we need to implement drawing for all "text" inputs (and outputs) so that people like Prince won't have any issues expressing their autonomy with something like this? https://en.wikipedia.org/wiki/Prince_(musician)#/media/File:...

(I'm aware that this was a stage name, but it is a legal name in the US)

Well, if you're fine with being occasionally called/written Yay-souwn (ah the mistakes that can crop in when translating names into languages with different phonetic constraints), I have no argument against that.

My parents have adopted romanized names. My last name is 洪 which in Pinyin is “Hong,” but in Taiwan, they’ve been using a mix of romanization schemes. In America, my name has been romanized to “Hung.” And I’m fine with that. And I don’t expect anyone to know how to write 洪.

Are there names that cannot be transliterated? Somehow all the banks and passport issuing governments in the world seem to manage this, if it's good enough for the banks and tax authorities it's probably be good enough for the FSF.

Every single passport has a transliterated version of the holders name in the MRZ.

I think you will find that non-U.S. banks don’t use ASCII for this exact reason.

A transliteration is arguably not someone’s legal name, and furthermore it’s kinda rude to demand that we mangle our names to fit someone’s 1970s technology system.

Practically all CRS tax information sharing happens using ASCII-7

Yes, since that’s basically been dictated by the U.S. Government. That does not mean the banks over here are using ASCII. Very systems generally use whatever national encoding was used before Unicode (ISO-8859-1 aka Latin 1 in my case) and more modern systems just use UTF-8.

Trust me, I have non-ASCII characters in both my surname and my address. The only times I have to transliterate them are if travelling, shipping or banking internationally.

Many organizations handle this very poorly.

US immigration and customs forms do not have sufficient space or characters for many common Hispanic names.

Famously, people with the last name Null are often at the mercy of SQL database input validation.

I'm having trouble finding the article (haha) but there's a person named either A or Aa (a not-uncommon Southeast Asian name) who has problems with flight ticketing systems that use that string as a special code.

There are multiple competing romanizations of most Asian languages (Mumbai vs Bombay, Peking vs Bejing, Sato vs Satou) which can result in mismatched name records.

The most inclusive format is a text field that accepts arbitrary length unicode strings. At the same time, you probably don't want people to put their names in as the Eggplant Emoji or USA flag. It's not a trivial problem.

> US immigration and customs forms do not have sufficient space or characters for many common Hispanic names.

Also the other way around. I can't officially order from IKEA here in Spain because their online order entry demands a second surname (segundo apellido) which I don't have :D This is a typical Spanish thing. I do however have a middle name which tends to trip them up a lot (my middle name is more recognisable as a first name so it often confuses them)

I just put '-' or 'NA' or something but technically the delivery guys could make issues about it considering that is not on my passport.

However the rules are not taken very strictly in Spain which is one of the nice things about living here.

> It's not a trivial problem.

It seems pretty trivial. You just follow existing international standards and use the transliterated version of the name found on the passports MRZ if accepting unicode strings isn't easy for your particular use case.

What if there are multiple standards? (Hint: there are.)

What if the passport has the same issue as US immigration and doesn't actually fit the name?

What about LGBT users whose passports may or may not reflect their names, depending on varying government policy?

What about people who have never had a passport? (Common in the US, as well as displaced/vulnerable persons worldwide)

These are not hypothetical edge cases, these are millions of real people. My own name doesn't fit in a lot of older forms and gets incorrectly parsed a surprising amount.

All names can be transliterated if you can agree on the transliteration system with the FSF. :P Even if you stick to just "standard" transliteration probably most names can be transliterated. The problem is that the transliteration is unlikely to be the person's legal name! In my case it wouldn't be. My legal name has an "í". I wrote to the FSF with an "i" to satisfy the ASCII requirement, but the thing with an "i" is not my legal name. I doubt in my case there will any trouble, but I thought it was funny the very first instruction in the copyright assignment process was literally impossible to carry out.

I bet your passport has a transliterated version of your name on it. In fact, you'll have a really hard time finding anyone who does not.

> In fact, you'll have a really hard time finding anyone who does not.

This part is a little false. It's actually pretty easy to find people without a transliterated version of the their name on a passport, because it's pretty easy to find people without a passport! But I agree that passports usually have a transliterated version of the name, at least every single passport I've ever seen for someone with a non-ASCII name does have a transliteration.

I do happen to have a passport and it does have a version of my name in only ASCII characters, in all caps. I'm not sure whether what is on my passport is "my full legal name" or not. Indeed I do not know if that term has a precise definition. I can tell you my name is written slightly differently on my birth certificate.

I didn't use the passport version of my name for this FSF email because it felt weird to use all caps. Instead I just changed the non-ASCII "í" to an "i".

Cannot? Yes, loads. English isn’t the entire phonetic alphabet.

But more than that, why should they be transliterated? Unicode exists and is well supported, not using someone’s actual name when it’s entirely feasible is just rude.

Edit:

Fun example: “Yahweh” and “Jehova” are both transliterations of «יהוה»

> Unicode exists and is well supported, not using someone’s actual name when it’s entirely feasible is just rude.

It's really not, and I say this as someone whose "actual name" has to be transliterated for an ASCII representation. It hardly ever makes sense to put in the work to convert old systems to accept unicode names, nobody actually cares.

Every single government transliterates their citizens names on their passports.

That's some pretty poor wording but they probably either mean UTF-8 and/or transliterated into latin characters.

I think they don't mean UTF-8, since they said ASCII. But I agree that they probably did mean something like "if your legal name cannot be written in ASCII and there is some standard way for it to be transliterated to ASCII use the transliteration instead of your legal name", or at least that's what I assumed they meant and filled out the subject line accordingly.

Reply on news.ycombinator.com