On the name of Persian Gulf

General talk
31

Dear Matija_Nalis
I read your content. But one problem remains…
If you want to speak impartially, why does the main tag (name tag) remain in Arabic? In fact, the main problem is the same tag. If other languages want to see their desired phrase, then at least the main tag should be according to international documents. All international documents say that this is the Persian Gulf. But we see that injustice has happened here and they have easily discarded all the documents

Matija_Nalis عزیز
من مطالب شما را خواندم. اما یک مسئله باقی میماند…
اگر بی طرفانه بخواهید صحبت کنید پس چرا تگ اصلی (تگ نام) به زبان عربی باقی میماند. در واقع مشکل اصلی هم همین تگ است. اگر سایر زبان ها مایل اند عبارت دلخواه خود را ببینند پس حداقل تگ اصلی باید مطابق مدارک بین المللی باشد. تمام مدارک بین المللی میگویند که اینجا خلیج فارس است. اما میبینیم که در اینجا بی عدالتی رخ داده است و به راحتی تمام مدارک را کنار گذاشته اند

32

So 25% are Arabic (language), 25% are Persian & 50% are English (as the international maritime language).

33

Make no mistake, 25% belongs to 7 countries and 25% of these waters belong to Iran. The problem is in the main tag. We have to proceed based on the documents. Even if we assume that this sea is international waters, then it should be called according to its international name. The international documents mentioned above.

اشتباه نکنید، 25 % برای 7 کشور و 25 % از این آبها هم متعلق به ایران است. مشکل هم در تگ اصلی است. ما باید بر اساس اسناد پیش برویم . حتی اگر فرض کنیم این دریا آبهای بین المللی باشد پس باید بر اساس نام بین المللی آن نامیده شود. که در بالا به آن مدارک بین المللی اشاره شده.

1 Like
34

This is the problem that vector tiles won’t solve. The gulf has multiple names with valid English translations (Persian Gulf and Arabian Gulf), so which should be displayed if the user selects English as their language?

Wikipedia suggests we have it the wrong way round in OSM. With the primary English name being Persian Gulf and an alternative (newer) name being Arabian Gulf.

It’s clearly a politically contentious issue. A useful read which suggests just using “The Gulf” (but I doubt that’ll please anybody!).

1 Like
35

Given that the international language of the sea is English, perhaps you’d prefer that in the “name” tag, leaving all the language names as well? Or perhaps “English / Arabic / Persian” in odd years and “English / Persian / Arabic” in even ones**? :slight_smile:

** “at the right” being “first” in an RTL language.

36

But don’t you see - if all the visible solutions (like the one mentioned above) implement that the names are always displayed in user-preferred language, then the issue goes away, as name does not get used.
In fact, one might then suggest that name tag gets completely removed on all disputed areas as to not show preference for anything, and thus avoid all the political bickering and waste of time (which could be better spent doing actual mapping). That would be ideal, don’t you think?

It could potentially also avoid other problematic issues (like disputed territories, warzones etc) in the future.

In fact, the main problem is the same tag. If other languages want to see their desired phrase, then at least the main tag should be according to international documents. All international documents say that this is the Persian Gulf. But we see that injustice has happened here and they have easily discarded all the documents

I’m not sure who are “they”, and I have no idea about political situation, or international treaties there, etc. sorry. :man_shrugging:

For me (who can read neither Persian nor Arabic), the displayed name is even worse – no matter the issue of political preferences, I cannot read either AT ALL (nor more than half of the labels on the planet on default osm.org rendering, for that matter). Surely we can agree that having unreadable name is even worse than having one which is less popular/preferred/politically correct/whatever?

So, yes, I agree that current solution on map displayed on osm.org by default is obviously not a good solution, yet it is what we have at the moment at osm.org until vector tiles get implemented there, which give the possibility of better solution. Which is why I’m quite keen on seeing it.

But I looks like it is not international waters, but exclusive economic zone? I have no idea how are those governed in this specific case.

name:en ? It is not about translation of words of string one by one into some other language; it is about how it is actually called in English. E.g. if you are British citizen, how do you (/your country) officially calls it (for British English in this example).

There is int_name too, so some users might want to choose that instead.

1 Like
37

See the source I linked to. Both are used. That’s why it’s not just about translations and vector tiles. It’s the old disputed/multiple names issue again.

38

So? I don’t think such problem is solvable globally at all (or else we would not have wars, for example). So, I set lower standards that are actually achievable - I just want to solve/avoid the problem for OSM.

E.g. in the case of language uses of USA officials mentioned in the Guardian (if you are referring to that?).
Or, you may prefer to call it shifting the problem from OSM to USA/whatever government what they want to call official in which year, sure. But that too solves the issue in OSM (Governments at least have more resources to waste at the problem and that is what they do, and OSM can simply copy whatever powers-that-be decide with no wasted time/resources instead concentrating on what it does best).

It’s the old disputed/multiple names issue again.

Yes, those are solved with such solution too. If I choose in for example OsmAnd that I want labels in Croatian, I’ll get name:hr (when that’s available). If a Serbian visiting Croatia chooses to see labels in Serbian, they’ll get name:sr. If an Italian tourist chooses to see labels in Italian, they’ll see name:it. Quite good.

Is it perfect? Well, no, but nothing is ever perfect.

Better question is “would it be an improvement over current rendering on osm.org”? And the answer is resounding “Absolutely, vast improvement” IMHO. Don’t let the perfect be the enemy of the good, and all that, eh?

2 Likes
39

Yes, this would be a vast improvement. Naïve data consumers that only consider the name tag would show nothing, which would be an improvement over the current situation where the same impasses are brought up for the debate. It would also encourage the use of name:xx and xxx_name tags for data consumers that want a more complete display.

The same situation exists for features that have no obvious primary language such as Atlantic Ocean as noted above, and there too, the obvious solution is no name tag.

The fact that we’ve not come to this very obvious conclusion is a testament to the fact that (a) people care very much that a label is rendered on osm-carto for for features they expect to see and (b) we lack a fundamental ability to or interest in making that decision as a project. OSMF’s decision about where to put the RU/UA boundaries with respect to Crimea was far more controversial than would be a decision to omit the name as a general policy for disputed features that are at an unresolvable impasse.

1 Like
40

I think you conflated two completely different subjects. Was that a intentional (unmarked) sarcasm for humorous effect, or something else? I can’t tell with all the :popcorn:.

So, to take it at face value, if you did not intend sarcasm:

The “Naïve data consumers that only consider the name tag would show nothing” of your reply seems to be about this quote above instead, and actively disregards the (inseparable, obviously!) “if” part of the quote – which seems somewhat disingenuous.

“if all the visible solutions” (my quote) implemented that, that would by definition preclude existence of “Naïve data consumers” (your quote) that didn’t implement that. Yes, it is in the possible farther future, and not tomorrow by noon (and I did not claim any deadlines!)

Yet the message of mine that talks about “vast improvement” - it talks about one single data consumer only – osm.org default map. If that one supported languages better (as explained), than it would have made vast improvement, i.e.:

That issue is completely orthogonal to potential (eventual) removal of name tag on disputed territories in the farther future (which would obviously only happen after at least major data consumers support that. Which would take some time, I’m sure – but also we can see that instilled sense of urgency e.g. OAuth1 deprecation can make data consumers move at faster pace :open_mouth: ).
Only connection between those two ideas (from two separate posts) is that osm.org map improvement implementing better name handling is one of the essential prerequisites to even consider descending going down “removing name on disputed territories” route.

It would also encourage the use of name:xx and xxx_name tags for data consumers that want a more complete display.

Yes it would, to everyone’s benefit. When they see what what osm.org does, they might decide to follow suite. Thus I propose that change as a first step. Other steps might (or might) follow later, depending on how that one goes.

Were I not completely given up on suggesting improvements on Carto default map style, I’d suggest that on low zoom levels it renders int_name as priority to name (which could be done today)…

41

Then we would simply be arguing over which label to place in int_name. What the belligerents to this particular conflict care about is what is on the map on osm.org and to a lesser extent to any other map that renders the name label as the first class provider of a rendered label. Eliminating the name fixes that problem for all maps that render name while also accurately describing the situation where no one name or language is singularly appropriate for a large international feature.

1 Like
42

For that to work though, wouldn’t every POI need to be tagged with every language option?

44

This is not a problem of different languages in the first place, but a naming dispute: Names - OpenStreetMap Wiki
Therefore using ‘name:xx’ cannot solve this unfortunately.

You are mixing languages and countries. The example already shows: if ‘name:en’ was about how officials call it: is it the British or US American official name or maybe the Indian one or the Ghanaian or the Australian?
This is getting worse with ‘name’ obviously.

  1. unreadable depends on what language/scripts you are familiar with. For a lot of people in the world English is unreadable.
  2. do you more like a shameful to you, but readable to others ‘name’ for a place in Croatia? Or should it better be in the local language like described in Names - OpenStreetMap Wiki

Problem: there is no local language for international territories.
But assuming we solve this by defaulting to English, we still end up with a dispute about the English name “Arabian Gulf” vs “Persian Gulf”.
That is the topic of this thread.

Problem: there are no “people on the ground” in the middle of the sea.

Watching the Atlantic Ocean: repeated name removal debate the osm community has not found a solution for this yet.

It’s our turn:

  • to find a solution we should ask “how to resolve naming disputes when there are no people on the ground”.
  • for politically disputed areas we could also more strictly use A/B names with unpolitical ordering like A being the name that existed first or the first in the alphabet
1 Like
45

Thanks to all of you who spend time and energy on this issue…
We are in a dilemma to choose a name (not the Arabic or Persian language, but the name itself).

I have a question for all of you: Why don’t you use any documents to determine who is right?

The use of both names is for when there is no acceptable evidence for any of the two names. But there is strong evidence in this case.
If you try to remove the problem text (remove the “name” tag), the problem will not be solved. Even if you use the name in an international language, you can still refer to international documents to see what words should be written in the “name” label. These documents cannot be easily ignored. In addition, the documents available in Iran have not been used so that there is no room for doubt. The documents are in the UK.

ضمن تشکر از همه شما که برای این قضیه وقت و انرژی میگذارید…
ما در یک دو راهی برای انتخاب نام (مقصود زبان عربی یا فارسی نیست بلکه خود اسم است) هستیم.

سوالی از همه شما دارم: چرا برای اینکه مشخص شود حق با کیست از مدارک هیچ استفاده ای نمیکنید؟
استفاده از هر دو اسم برای موقعی است که هیچ مدرک قابل قبولی برای هیچ کدام از دو اسم نباشد. اما در این مورد مدارک محکمی وجود دارد.
اگر بخواهید متن مشکل را از بین ببرید که (حذف برچسب “نام”) مشکل حل نمی شود. اگر هم از نام به زبان بین المللی استفاده بکنید باز هم میتوانید با مراجعه به اسناد بین المللی ببینید که چه عبارتی باید در برچسب “نام” نوشته شود. این اسناد به راحتی قابل چشم پوشی نیست. در ضمن از اسناد موجود در ایران استفاده نشده تا جای هیچ تردیدی نباشد.
اسناد در بریتانیا هستند.

46

This is because there is no authoritative international government. OSM often sees this with international data - as an example, the boundary between India and Pakistan is disputed, and both states argue (sometimes violently) about what the land in both their claims should be called.

There is, of course, the United Nations - but their naming conventions just follow the political makeup of the organisation - see the references to the UN in here. Also, the UN is notorious for passing contradictory resolutions in difficult cases (e.g. Israel and Palestine - both sides can find UN resolutions supportive of their views).

4 Likes
47

For those interested in reading why there will not easily be a one-name solution: Persian Gulf naming dispute - Wikipedia

1 Like
48

It could be possible to introduce a “political perspective” as Overture is doing?

49

The current language used for the ‘name’ tag is in Arabic rather than Farsi, and there are likely several reasons for this:

  • Seven Arabic-speaking countries surround the Gulf, while there is only one Farsi-speaking country.
  • The combined coastline of the seven Arabic-speaking countries exceeds the total coastline of the Farsi-speaking country.[1]
  • The populations of the Arabic-speaking countries surpass that of Iran, the sole Farsi-speaking country in the Gulf.[2]
  • There is a higher count of Arabic-speaking cities and towns overlooking the bay compared to Farsi-speaking ones.
  • The populations of the Arabic-speaking cities overlooking the bay outnumber their Farsi-speaking counterparts.

It’s important to note that the Farsi name is correctly specified in ‘name:fa’, along with all other foreign language-specific tags.

It’s worth mentioning that this topic has been discussed multiple times in the OSM talk mailing list.

In consideration of the comments, it appears that the user is troubled by the Arabic name being in the ‘name’ tag and suggests either replacing it with Farsi, English, or removing the ‘name’ tag entirely. This prompts a question about the specificity of this concern, particularly when observing the Caspian Sea in northern Iran, where the ‘name’ tag has multiple values, each representing a different name in a different language, without similar objections.

Regards.

References

[1] The coastlines for Iraq (58km), Kuwait (500km), Saudi Arabia (560km), Bahrain (161km), Qatar (563km), UAE (1318km), and Oman (100km) sum up to 3,260km, whereas Iran has a 2250 km coastline along the Gulf.
[2] The populations of Iraq (43.53 million), Kuwait (4.25 million), Saudi Arabia (35.95 million), Bahrain (1.463 million), Qatar (2.688 million), UAE (9.365 million), and Oman (4.52 million) sum up to 101.763 million, while Iran has a population of 87.92 million.

50

Overture is not (yet) doing this, it is an rfc. Everybody can write a similar proposal for osm.
It’s interesting however.

The perspectives section still ends with the need to pick a default, which can be controversial as for “name” in osm.

The term “default” does not imply any value-judgments, but is typically the most commonly-accessed perspective.

With maintaining perspectives, it has the potential for more transparency and gives data consumers the ability to decide. Not so for the consumer of the default data - just like osm carto.

51

Interestingly, Natural Earth is doing this. This hasn’t prevented the same sort of issues from cropping up as in this thread. Arguably** they’ve had more “internet vigilantes” causing them problems than OSM has - the issue seems to be that one side in some debates does not believe that the other side even has a right to a view, so an attempt to provide an opposing “point of view” to match the complainant’s causes just as many complaints.

** a subjective judgement - I see NE github tickets, OSM DWG complaints and discussion in places such as this.

1 Like