FEB
16
2008

Marbl10n

This Wednesday Marble got fully translated to Czech (thanks to Vít
Pelčák) - including some fixes to the localized city names in the Czech Republic (thanks to Karel Volný). Recently I've looked closer at the KDE Localization Statistics which seem to have improved a lot during the last few months. What has become more and more apparent to me is the fact that Marble might soon be the virtual globe that has the best and most translated GUI on this marbleous planet. Ok, this shouldn't exactly be the primary concern of a tool like this (Yes, I hear you asking for more and better maps instead and we are working on that as well).




Caption: Marble's UI translated to Chinese (Credits for the Screenshot: John Smith).

But it's still interesting: Taking a peek into Google Earth's ~/google-earth/lang directory I can spot 14 files in the familiar .qm format. So I guess that Google Earth's user interface is available in 14 languages. With NASA's WorldWind it's a bit harder to say. But from what I was told there is no real translation framework available for the current stable version (However this is about to change with WWJ), so translations either need to get patched in or need a special plugin. As a result there are up to my knowledge no other languages officially supported apart from english (although the NASA WorldWind Wiki lists some translations).

Looking at our Localisation statistics I can see that Marble is close to 100% fully translated in Arabic, Catalan, Chinese (Simple & Traditional), Czech, Danish, Dutch, Estonian, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Latvian, Low Saxon, Norwegian (Bokmål and Nynorsk), Polish, Portuguese, Punjabi, Romanian, Slovenian, Spanish, Swedish and Ukrainian. So according to The Count the stable version of Marble has GUI translation support for up to 26 27 languages ("Haha haha ha ...!"). With KDE 4.0.2 being near most notably absent are Krakhozian, Danish (thanks to Martin Schlander!), Farsi, Korean, Russian (thanks Nick! ;-) ), Turkish, and lots of other beautiful languages.




Caption: Marble's UI translated to Slovenian (Credits: Jure Repinc).

I understand that translation is a much more difficult task than most people might think -- especially for an application like Marble: There are lots of technical terms which don't get recognized as such easily -- even by translators:
For example the label of the "Relief" checkbox in Marble's legend reads "Relief" in English. This (or "shading") is the correct technical term in English and it's easy to understand for the mainstream audience. In German however the "official" cartographic technical term for this feature would be Schummerung. In my opinion this correct technical term is a pretty unknown word in German which has a strange sound to it and I'm pretty convinced that most people wouldn't immediately understand what it's about.

As the brilliant physicist and teacher Richard P. Feynman had pointed out on several occasions there is a big difference between knowing something and knowing the name of something. Of course it's a lot more important that people get a deep understanding of the nature of things than knowing the correct name. Still it's necessary to be easily understood and hence it is important to stick to wording standards. However these standards might vary depending on the audience you are talking to.

"Relief" is easily understood in German and gets the point across immediately (and it's not even an anglicism). Therefore I consider it a much better choice given Marble's target audience.

Now counting the supported languages of the GUI might be a nice thing to do for arithmomanians. But the real deal is translating the map:

The primary problem with places data is that at least the packaged database needs to be compliant with the Debian Free Software Guidelines (DFSG). So our current DFSG-compliant solution is using older data from Stefan Helder's World Gazetteer which includes about 20,000 city names (and some english country names from the CIA World Factbook which is public domain). This data consists mostly of city names which are provided in the latinized-localized official language of the respective country. Of course this is only a temporary solution. Ideally I'd like to let Marble display the name of the cities in the current local language and get the name additionally displayed in the original official language of the city (that would include the original glyphs of the language). To get this solution implemented there are two solutions which both have their merits and which I'd like to see implemented in parallel:

  1. compiling a rather "small" database of about 20,000 to 40,000 city names which would include translation. This would again be based on the current updated data of the World Gazetteer (which has improved quite a lot over what we have right now). This would also bear a chance to be a database that could get used by other applications on the Free Desktop.
  2. Once Marble gets started and is connected to the internet it could automatically download a localized index file that contains all the geo-referenced places in Wikipedia. This index file could get generated based on Wikipedia's API and get stored on the server.

There are several advantages of this approach: The Wikipedia data is very much translated already. It contains cross references to several languages and has some data which describes the type of content. The latter is important to get the data categorized and to annotate it with the correct symbols. Naturally each entry would automatically come with a Wikipedia Url already.

So these two approaches are something that I'd like to see implemented for Marble in KDE 4.1. If you feel like helping out with it then we'd appreciate any patches of course. So far we didn't have the chance to really start work on it as we are currently heavily working at a different Marble construction site. So you can't do much wrong if you just start off. I could also imagine that our translation teams would be happy about any qualified and dedicated help with translations!

Comments

I can't make 'Kyyiv' or 'Moskva' appear in Russian, so i get bad results:
http://ru.wikipedia.org/wiki/Kyyiv instead of http://ru.wikipedia.org/wiki/Киев

ideally, marble and kgeography should share such translations (for example via libkdeedu, or --even better-- kdelibs/geography.po)


By shaforostoff at Sat, 02/16/2008 - 23:39

Yes, that's about what I had pointed out: for the future database we'd like to support cyrillic letters as well as arabic/CJK glyphs for the names and for the wikipedia Urls. Also libmarble is meant to get into a more central place (likely kdesupport) so it can provide maps as well as a places database for all kinds of applications.

That's our plan.


By Torsten Rahn at Sun, 02/17/2008 - 02:58

Apropos counting: http://www.youtube.com/watch?v=3ZQI0Xm29To

«Our chief weapon is surprise...surprise and fear...fear and surprise.... Our two weapons are fear and surprise...and ruthless efficiency.... Our *three* weapons are fear, surprise, and ruthless efficiency...and an almost fanatical devotion to the Pope.... Our *four*...no... *Amongst* our weapons.... Amongst our weaponry...are such elements as fear, surprise.... I'll come in again.»


By robinst at Sun, 02/17/2008 - 00:28

From the World Gazetteer FAQ:
"If you would like to republish the data presented here, please do not change the data"
Doesn't that conflict with both the DFSG and the use Marble is actually making of that data (updating it, integrating it with e.g. the CIA World Factbook data, ...)?


By Kevin Kofler at Sun, 02/17/2008 - 06:07

Well, that was indeed a concern when I evaluated it. However the wording already indicates that the author only had this as a wish (and not as a requirement to make use of the data according to the license terms - just like we would of course wish that nobody would use KDE to kill people).
So if you look at the same text in the other languages then you'll see that this kind request doesn't appear in the other languages at all:

french:
http://gazetteer.de/wg.php?x=&lng=de&des=wg&srt=npan&col=abcdefghinoq&msz=1500&men=ipge&lng=fr#copyright
german:
http://bevoelkerungsstatistik.de/wg.php?x=&men=ipge&lng=de&des=wg&srt=npan&col=abcdefghinoq&msz=1500#copyright

"Dieses Projekt soll als frei verfügbarer Datenlieferant dienen. Einige Anfragen veranlassen mich dazu, die Informationen zum Copyright zu präzisieren. Wer die Daten oder Grafiken auf seiner eigenen Website oder in einem Buch benutzen will, sollte nach Möglichkeit folgenden Hinweis unter den Daten oder im Quellennachweis verwenden: © by Stefan Helders www.world-gazetteer.com" (please use babelfish for a proper translation)

I had contacted the author directly via e-mail and he confirmed that "not changing" parts of the data is in fact not a requirement for the terms of usage and that he is open to relicense it to a free software license if we'd come up with something that would fit (the problem back then was however that there is no fitting license for these kinds of data and as the statement on the german and french website already did clearly fulfill the requirements of the DFSG there wasn't a pressing need to do something about it).

See also: http://websvn.kde.org/trunk/KDE/kdeedu/marble/data/placemarks/COPYRIGHT?revision=731999&view=markup


By Torsten Rahn at Sun, 02/17/2008 - 09:58

I guess the copyright versions on the World Gazetteer site which are most clear about the license are the french one and the spanish one which explicitely refer to usage in software and make a clear request to mention the author and the website as the single requirement to make freely use of the data:

Informations sur le copyright
toutes les données © 2005 by Stefan Helders
Ce projet était conçu pour être un fournisseur libre pour des statistiques de population. Quelques demandes me font préciser les informations du copyright. Ceux qui veulent utiliser les données our graphiques sur leur site, programme ou livre doivent utiliser l'indication suivante sous les données:
© by Stefan Helders www.world-gazetteer.com

Informaciones sobre el copyright
todos los datos © 2005 by Stefan Helders
Este proyecto está considerado como proveedor de datos libres de población. Algunas preguntas me dejan precisar la información del copyright. Para usar los datos o imagenes en su sitio web, programma o libro hay que mencionar el siguiente informe debajo de los datos o en el índice de las fuentes: © by Stefan Helders www.world-gazetteer.com


By Torsten Rahn at Sun, 02/17/2008 - 10:27

Committed Danish translation for Marble a couple of hours ago. Had to play my part in putting Google (Earth) in it's place :-)

And thanks for paying attention to our work.


By cb400f at Wed, 02/27/2008 - 00:39

Thanks a lot for your work! I was born in Flensburg which is very close to Denmark. So I feel a bit embarassed that I don't speak Danish. But I'll definetely have a look at your work and I'll probably think back to the numerous summer holidays that I enjoyed in Denmark in my childhood.


By Torsten Rahn at Fri, 02/29/2008 - 10:44