Missing "official name" entries.

Bug #44528 reported by Mysha
20
Affects Status Importance Assigned to Milestone
iso-codes (Ubuntu)
Confirmed
Wishlist
Unassigned

Bug Description

The template for translation for iso-3166 for some cases does not have seperate "offical name" entries, presumably because in English they are the same as the "name" fields

As this might not not be valid for other languages, all countries should probably have those fields, even where that means that in English the text is the same in both.

Eg.
BRN:
  en
    name: Brunei Darussalam
  fy
    name: Brûnei
    official name: Brunei Darussalam

Revision history for this message
Emmet Hikory (persia) wrote :

The difficulty is defining the "official name". As an example, one might be tempted to use the values from ISO 3166 (for Japan this is "Japan"), but some may argue it should be the name used by the reigning government of the country (for Japan "日本" ("nihon" in roman characters), which may not even be properly visible in some locales.

Revision history for this message
Mysha (mysha) wrote :

Though there may be difficulties in determining official names, if we use them then all countries should have entries for them. No doubt the Japanese translator would distinguish between Nihon and Nihon-koku, even if others might not.
(ISO 3166 names only exist for English and French, if I'm not mistaken.)

Revision history for this message
Mysha (mysha) wrote :

I'll slightly revise that statement :
Though there may be difficulties in determining official names, since we use them then all countries should have entries for them!

Revision history for this message
Simos Xenitellis  (simosx) wrote :

The idea is that for the English locale you put the official language name, and if individual countries want to be called something else, they can put that name is their localisation files.
That is, the localisation coordinator for the language can pick and choose the proper localised name they think is good.

Revision history for this message
Mysha (mysha) wrote :

The idea is that on Australia there is:
name for AUS: Australia
Current Frisian: Austraalje

while on Austra there is:
name for AUT: Austria
Current Frisian: Eastenryk

official_name for AUT: Republic of Austria
Current Frisian: Republyk Eastenryk

Now, where is the entry:
official_name for AUT: Commonwealth of Australia
Current Frisian: Steatebûn Austraalje
?
Both are exactly the same, namewise, they haven an official name and a colloquial name, so whatever reason is valid for the one must also be valid for the other.

Further above I used the less extreme, but poignant example of where apparantly in English the colloquial name is the same as the official one, and therefore the rest of humanity are denied the possibility of indicating a difference between the two.

So, no it's not a matter of "English" just putting in an official name, and countries putting in different names in "their" localisation files.
                                                                                                           Mysha

Daniel T Chen (crimsun)
Changed in iso-codes:
importance: Medium → Wishlist
Revision history for this message
Mysha (mysha) wrote :

How does one get this bug from Wishlist to Design flaw?

Revision history for this message
Mysha (mysha) wrote :

So, how does a bug blow out four candles?

Four years for something that should be less than an hours work in itself. As it doesn't seem to influence anything else, checking functionality should add much more than that.

Why is this not a simple implementation error, long fixed? And even if we assume that, since it isn't fixed, there must be a design element in it: How can this be such a complex design flaw that it can't be fixed by removing the code for handling the case of a country with just one name?

(And how can a bug remain a false duplicate for four years?!)

Revision history for this message
Mysha (mysha) wrote :

Read: "As it doesn't seem to influence anything else, checking functionality shouldn't add much more than that."

(When does this bug system get enough editing capabilities to remove typos?)

Changed in iso-codes (Ubuntu):
status: New → Confirmed
Revision history for this message
Christian Perrier (bubulle) wrote : Re: [Bug 44528] Re: Missing "official name" entries.

Quoting Thomas Hotz (<email address hidden>):
> ** Changed in: iso-codes (Ubuntu)
> Status: New => Confirmed
>
> --
> You received this bug notification because you are subscribed to Ubuntu.
> https://bugs.launchpad.net/bugs/44528
>
> Title:
> Missing "official name" entries.
>
> Status in “iso-codes” package in Ubuntu:
> Confirmed
>
> Bug description:
> The template for translation for iso-3166 for some cases does not have
> seperate "offical name" entries, presumably because in English they
> are the same as the "name" fields
>
> As this might not not be valid for other languages, all countries
> should probably have those fields, even where that means that in
> English the text is the same in both.
>
> Eg.
> BRN:
> en
> name: Brunei Darussalam
> fy
> name: Brûnei
> official name: Brunei Darussalam

Well:

1) I think that thinking that a given language should translate the
official name and the "short" name differently when the English name
is NOT different is plain wrong. The standard has "Brunei
Darussalam" for both entries and the translations should follow
that. If they don't, they're incorrect. FYI, the only translation that
has an official status in the standard is the French one and the
translation for this country is "Brunei Darussalam"

2) I don't see how that can be achieved with gettext

So, well, I think this bug is to be tagged "wontfix".

Revision history for this message
Mysha (mysha) wrote :

1 Ah, that seems like a new point of view. Let me see if I get this straight. Over time, some countries have received colloquial names in English. This is a completely arbitrary set, determined by historical events, personal tastes of news editors etc.
Are you saying that other languages should have conformed themselves to the arbitrary set of English colloquial names and never should never have developed such where English didn't?

2 Not sure where to fix it, but the simplest solution would seem to be to create double entries for English for all countries, even if they sometimes have the same values. That way other languages can differ where necessary like English can.

(And #44528 is not a duplicate of this. )

Revision history for this message
Christian Perrier (bubulle) wrote :

Quoting Mysha (<email address hidden>):
> 1 Ah, that seems like a new point of view. Let me see if I get this straight. Over time, some countries have received colloquial names in English. This is a completely arbitrary set, determined by historical events, personal tastes of news editors etc.
> Are you saying that other languages should have conformed themselves to the arbitrary set of English colloquial names and never should never have developed such where English didn't?

We are not talking about colloquial names. We are talking about an
international standard.

ISO-3166 standardizes country names in English (and French). I
therefore expect the translations to follow the spirit of the
standard: if a country's official name is the same the its short name
in English, it should be the same in all translations.

>
> 2 Not sure where to fix it, but the simplest solution would seem to be
> to create double entries for English for all countries, even if they
> sometimes have the same values. That way other languages can differ
> where necessary like English can.

Technically speaking, gettext makes it nearly impossible to achieve.
And non-technically speaking, that would defeat the standard.

Revision history for this message
Mysha (mysha) wrote :

Hi,

You're completely right (except for the bit about ISO 3166 standardising country names).
The header of the file very clearly states: "Translation of ISO 3166 (country names) to LANGUAGE".

Not that it solves the problem, of course, but it solves this bug report.

For the curious:

- Problem 1: From my first encounter with Launchpad translation, I felt it didn't provide the translator with enough information about the intention of the text. The fact that aforementioned header is not displayed is a demonstration of this.
- Problem 2: ISO 3166 claims to merely standardise the codes for the names, not the names themselves. Those names are taken from the UN documents. The UN apparently takes the position that they standardise names for use in UN documents only. Does this make the list of country names in ISO 3166 fit for any specific purpose in Ubuntu? (The case above, regarding the naming of Australia versus that of Austria, and the case of Macedonia can serve as demonstrations of the problem.)

But I've basically given up on bug databases, so I won't pursue either problem further. Anyone should, of course feel free ...

Bye
                                                                                                                               Mysha

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.