Why does the Arabic version of the arabic_arabian_peninsula sky culture have such a different description?

[10110111]

First, what made me wonder is why the section structure is completely different in ar vs en versions. Apparently, one of the reasons is that the Arabic text doesn't need the transliterations and translations present for the other langugages, which might justify the switch from a table to a bunch of <h4> sections.

But, the text itself also differs considerably. Compare, e.g. the text for constellation "النعايم":

English:

ID	Arabic	Transliteration	Translation	Comment
CON 200	النعايم	al-Naʿāyem	Al-Nayem	Meaning is not known for certain. It could mean the camels. composed of the seven stars of the Big Dipper.

Arabic:

النعايم - السبَّع - بنات نعش
سبعة نجوم شهيرة شمالية. ويرد اسم النعايم كثيرا في الشعر العامي

Translated from Arabic (by Google Translate):

Al-Na’aym - Al-Saba’ - Daughters of a Coffin
Seven famous northern stars. The name Al-Na'ayim appears frequently in colloquial poetry

Same mismatch exists for some (many?) others.

I've also seen some mismatching descriptions of the Chinese culture in translations like French, which are much shorter (and have fewer sections) than the English one.

I'm currently in the process of converting the current sky cultures to the format used in stellarium-skycultures repo (which would make it possible to avoid too much change at once when switching to this format), and this situation puzzles me regarding how I should proceed. Should I maybe just drop all these old cultures and simply import the ones from that repo, converting only the missing ones?

[alex-w]

Sorry for late answer.

The situation is existing by historical reasons. Previously all translations has been made by hands and not all translations were synced. After switching to Transifex the situation coming better, but again - not all translations for sky cultures were updated/added via Transifex.

I fear the question of priority of description for each sky culture should be reviewed separately, because new format are missing some features/data, were coming in desktop planetarium later introducing of new format of SC. Plus, the names of regions and licenses are different too.

It is possible that adding support new format of SC in desktop planetarium will cause some changes in all planetariums :(

@xalioth @gzotti @guillaumechereau @sushoff

[10110111]

new format are missing some features/data

What is missing? So far I haven't found any serious missing part (except minor things like classification).

Plus, the names of regions and licenses are different too.

Examples?.. What's the difference in licensing except the "Special pemission is granted to Stellarium Labs to include the licensed material." present in e.g. Arabic cultures and lacking in their desktop counterparts?

[alex-w]

What is missing? So far I haven't found any serious missing part (except minor things like classification).

Classification is a good example here. Plus description for Modern sky culture is really updated old Western sky culture.

Plus, the names of regions and licenses are different too.

Examples?.. What's the difference in licensing except the "Special pemission is granted to Stellarium Labs to include the licensed material." present in e.g. Arabic cultures and lacking in their desktop counterparts?

Desktop planetarium uses United Nations geoscheme (UN M49) for regions + specific region "World" for all modern sky cultures (IAU-approved 88 constellations with boundaries) - see an example the Anutan SC: Oceania region in SSC repo and Melanesia in desktop. Desktop planetarium also using standardized naming for licenses (include versions of license), and parser for human-readable displaying info for licenses restrictions.

[10110111]

Some of the issues I've solved in my skycultures-new-format branch. Particularly, the licenses are also parsed and displayed there, just as classification (which I added in my fork of SSC).

The differences you mention like the region are one of the reasons I was planning to first convert the existing cultures to the new format, and only later to replace them with the ones from SSC (maybe having altered them to conform to the requirements). This is why I came across this issue with the translations.

[sushoff]

ugh!

well, translations are always a snapshot in time! If somebody is kind enough to translate a description and then, later, the original author changes the description in the original language (but of course not in all languages), the translation(s) start to deviate. It will always happen because research goes on, we (researchers) update as many languages as we can (at least I do) but when I don't understand a language, I don't dare to change it. ... Our problem of software developers/administrators will always be that we never know that the original language of the researcher is (e.g.: I may possibly write in English / French directly or - on bad days when the language regionn of my brain doesn't work properly - in German ... so you as a developer cannot even guess)

Of course, as Alex said, this issue is for historical reasons.

What I currently see for many websites that I administrate: people/ institutions start to implement DEEPL (or: GoogleTranslate), so the author can write only one original (e.g. English) and the translation (to Spanish or whatever) will be done by the machine and on demand.

PS: Alex, I got the request only now as you mentioned me in the threat. ... I thought, we had already solved this issue.