sycamorex is the coordinator for Polish (Poland). Thanks!

I'll submit it soon to SBo, but if you need to use the transifex command line python client, a testing slackbuild is here

http://ponce.cc/slackware/testing/transifex-client/

I'm happy to confirm that Alien BOB is our coordinator for Dutch translation. We could hardly find a better one :cool:

Thanks, Didier.

Have I missed something or there's a bug in the script. During the initialisation and when I run slint.sh to create an iso it spits out the following errors:

That's a bit weird as it works here. Did you check the integrity of the archive with md5sum?

There *should* be a "setpkg" file in ...slint-20130427/data/32-13.37/source/ as it is present in the archive (I just checked downloading it from http://slint.fr/pub).

Weird, indeed. I must have done something to it when I was cloning it between machines with git as it wasn't there. I have no idea. I've downloaded it again and everything is fine. Thanks

Thank you for your kind words, Chess. This can't but encourage us try hard doing something useful for Slackware community.

Glad to see translation began, folks. And that's only what has been already uploaded ;)

I will be nearly unavailable next two days. Next slint version will include modified /isolinux/{message.txt, f4.txt} and 'setup' to allow testing for work in progress languages.

Also, translators please post here or send me a message including in your locale, encoded in UTF-8, following sentence: <Language> (<Territory>)" or "<Language> spoken in <Territory>", as e.g. "Français (France)" or "Français parlé en France".

Isolinux can't yet display texts encoded in UTF-8 (before booting), but I should be able to cook a custom font / code page including up to 128 characters of your choice, thanks to Anton Zinoviev for his bdf2psf script in Debian's console-font package :cool:

No, not really. es_UY would be the same as es_AR (mentioning it as that one is a lot more common), and 99.99% same as es_ES :) I have never seen a difference in localizations for es_ES and es_AR in projects that actually have both (like Firefox), and if I ever get to see one, I am sure it will be due to different teams doing the translations, and not because the difference actually exists. Now that I think about it, it is possible that for american countries, the locales select the "latam" keyboard layout instead of "es"

Most differences between the spanish variants are minor, mostly in style, with translations from Spain sometimes preferring a more formal tone when addressing the user. It is not like en_US vs en_GB, where there are differences in spelling of many words. I believe a single "neutral" spanish localization would be just right for all spanish speakers, and that is why I have registered in transifex for es_ES instead of es_UY.

The sentences you asked for (hoping LinuxQuestions is using UTF-8 ;):

Español (España|Argentina|México) (these are the most common locales for Spanish)
Español hablado en España

Well, I'm afraid I have to disagree with you there. ;)

There are significant differences between the Spanish language as spoken (and written) in Spain and Latin America.

Even in Latin America there are marked differences, specially addressing people. As you know, there are three different ways of addressing someone in Spanish (tuteo, voseo, ustedeo), and which way you use depends on the geographic location, formality or informality of speech and the country custom.

For example, in your country (Uruguay) is common to use the voseo in general and it's socially accepted. In other parts of Latin America, the voseo is an informal and intimate way of addressing someone. My mother, my brother, my sister and my best friend address me that way, but I would feel very uncomfortable if a complete stranger starts talking to me like that. I believe that with computer systems happens exactly the same. I don't want a cheeky program talking to me as if I'm his brother. Darn! Shut up! I'm not your buddy, I'm your master, you silly program! :D

Some grammar in Spanish from Spain is also different than the grammar in Latin America. In Latin America the Spanish forms vosotros/vosotras simply doesn't exist, and likewise the pronouns os/vuestro don't exist. I have seen Spanish translations of software (from Spain) written like this, and while most Latin American Spanish speakers would understand them, it's just not the way we talk or write, it simply feels foreign.

Certain major computer vocabulary is different as well, compare the use of the word "ordenador" (computer) by Spanish people. Most Latin Americans would use "computadora" or "computador" instead, the same happens with the word "fichero" (file) used in Spain, when in Latin America is commonly used "archivo" instead. And the list goes on and on, for common computer terms such as script, buffer, debugging, etc.

Another aspect where the Spanish from Spain and Latin America differ is in the use of anglicisms. The spanish are somewhat more reluctant to accept and use anglicisms in the language, while latin americans are not. For us latin americans is common (and we understand them) terms such as "script" or "frame buffer", while in Spain you'll find translations for those two terms ("guión" and "memoria intermedia").

I almost agree with you there. A "neutral" spanish translation would work fine for most spanish speakers, not all. I believe someone from Spain would prefer a Spanish translation where the computer terms he/she knows are used, someone from South America would prefer a voseo way of addressing and someone from México would prefer the tuteo addresing and wouldn't mind a few anglicisms.

Just ask yourself, why are there so many different translations/variants in Spanish? They exist for a reason.

By the way. I'm up to 60% with the Spanish translation, and I expect to finish it tomorrow. Simple and direct, using ustedeo addressing, accepting a few anglicisms, avoiding pronouns and colloquialisms like the plague. Care to join forces? Or would you prefer a translation tailored for your region?

Thanks fo your answer, AlvaroG.

I was embarrassed as I received two requests for Spanish in Transifex: one from you and one from Diantre for es_CR.

As I had no reason to favor one of these requests I accepted both.

But both locales differ very slightly (your post confirm that) and I don't want to duplicate efforts uselessly.

So please AlvaroG and Diantre consel together about that and tell me which locale I should keep and who will be the coordinator. In any case it's translation that will remain and all translators will be credited.

EDIT: I saw Diantre's answer to AlvaroG just after having posted this ;)

Also it is confirmed that gabrielmagno will do the translation in Portuguese (Brazil) and rmoki, aka dreamwalking, the translation in Greek, thanks.

Well, I was referring to software translations, which are simpler than human interaction. Specially if we refer to program output rather than documentation: most issues and differences can be avoided. I might have been thinking in a too-simplistic way, however, so I stand corrected.
I don't want software telling me "vos". Even when it is the way I address 95% of the people I interact with, from a computer it just feels wrong. But I do recognize that both "tu" and "usted" are valid.
To be honest I have not even downloaded the files, work has been complicated this last few days. What you describe is what I was planning on doing, I don't think it needs to be forked ;)

Just one little request: Please make sure that you don't force translated messages on me when I select a foreign keyboard layout or system locale in the installer (like other distributions do). In my humble opinion GNU gettext is just a great obfuscation tool. ;-) Their design doesn't support high quality localization well.

Sorry for the flood of updates in the Transifex project that happened now.

I'm translating the PO file locally using Lokalize, and submitting it to Transifex. When I tried to submit today's update to the site, I got a syntax error, without indication about where it was in the file. So I had to submit the translations strings in buckets, until I found where the error was.

Only the locale setting (through LANG's value) is considered in slint to choose installation language, not the keyboard layout.

If you would like to set the locale (other than en_US) in the installer but still have installer's message displayed in US English, I understand that you actually request ability to set during installation the default locale in the being installed system, even though that locale was not used during installation itself. Is that correct?

That wouldn't be difficult to implement. At present slint only asks the user whether he or she wants to keep in the being installed system the last locale setting used during installation, but instead it could as well offer to set another one. We would only need to store in the installer a (very small) table of available locales.

I would like to know if that feature would interest other folks as well though, before putting it in the TODO list.

Sorry, I just do not understand this statement. Gettext is only a tool to mark messages to be translated, then display a translation of these messages, if available, in the locale set by the user. This has nothing top do with the quality of translation. But maybe I did not understand what you meant by "obfuscation tool" or "high quality localization". In that case, please elaborate a bit.