Also files for sh (Serbian cyrillic) must be uploaded. Nothing is on Pootle for 3.4.1, but we have an SDF file. Using the same issue since the same volunteers will work on the two variants.

1. Who working now on AOO 4.0? How I can contact them?
2. Who must uploaded to Pootle, or create on Pootle service AOO 4 for Serbian?
Also, I find AOO 3.4.1 on Pootle, but, it is unclear for me to recognise is it Serbian sr (latin) or Serbian sh (cyrillic) version. 
3. What I can do now about translation?

This page is an internal issue for coordination between the OpenOffice Pootle administrators.

After this issue is closed, which will happen as soon as the Pootle administrators make the strings for OpenOffice 4 ready in Pootle at the (now empty) following URLs
https://translate.apache.org/sr/aoo40/ (Latin)
https://translate.apache.org/sh/aoo40/ (Cyrillic)
translators will be able to start. This should happen quite soon.

@sevalav: to contact other volunteers, help translating and get a Pootle account the right place is the l10n mailing list. See http://openoffice.apache.org/translate.html for more.

Serbian (sr) is now available on Pootle

https://translate.apache.org/sr/aoo40/ -> 6403 words needs translation
https://translate.apache.org/sr/aoo40help/ -> 100% missing

Related po files can be also found under 

http://people.apache.org/~jsc/translation/aoo40_ui_sr_po-files.zip
http://people.apache.org/~jsc/translation/aoo40_help_sr_po-files.zip

"sh" will be provided asap but it needs more work

sr as it was seems to include "cyrillic" but Pootle shows some problems that is probably related to a wrong encoding (not correct utf-8) 

And for "sh" Pootle does not know it anywas and we have to add a new language with new rules ...

I need some clarification

"sr" is handled as cyrillic and "sh"seems to include latin. What is correct? I found 

sr    Serbian
sh    Serbo-Croatian

but is this still correct? I don't know and need some input

I opened po file OOo_34-sr.zip and there is Serbian (cyrillic). So, "sh" will be Serbian (latin). 
Maybe, "sh" mean "Srpsko-Hrvatski" (in English: Serbo-Croatian). That was from period when two languages have the unique (the same) orthography. Serbia and Croatia are now two separated country, the best thing will be if you call Serbian latin version as "srl", wich means "sr" = Serbian and "l" = latin.
Serbian (cyrillic) can stay with only "sr"
That is only my suggestion, as someone who is from Serbian country. If you have good will and a little time more (about changing "sh" in something more apropriate), I can ask Gorana, who done this translations work in previous version of OOo, for suggestions about that.

well it should be a valid language code

OK. I will ask Goran about that for suggestion.

Just to ensure we don't have confusion here: when creating this issue, I retrieved "sr" and "sh" from older (3.3.0) pages like http://www.openoffice.org/download/legacy/other.html (sh -> Serbian Cyrillic, sr -> Serbian Latin, but this seems wrong with respect to the PO files we have).

A sample of what the PO/SDF files contain is:

sr:

$ tail sr/localize.sdf | cut -f 11
Појединачне речи
Места табулатора
Са сенком
Фонт
Позиција азијских фонтова
Азијски фонт
Увлачење
Величина фонта
Размаци у тексту
Треперење

sh:

$ tail sh/localize.sdf | cut -f 11
Boja fonta
Veličina fonta
Pojedinačne reči
Odvajanje
Razmaci u tekstu
Jezik pisanih slova
Stil stranice
Kontura
Zabranjeni znaci
Prozori

According to ISO 639, the standard we follow for codes, we have:
http://en.wikipedia.org/wiki/ISO_639:sr -> Serbian
http://en.wikipedia.org/wiki/ISO_639:sh -> Serbo-Croatian (deprecated, ISO suggests "srp" for Serbian, which is equivalent to "sr", see http://www-01.sil.org/iso639-3/documentation.asp?id=srp ).

So, from what I understand (but native speakers need to explain this better), "sr" is the code for both Serbian Cyrillic and Latin. But I don't know if OpenOffice supports modifiers to denote whether it's Cyrillic or Latin.

Which standard you want to see in AOO, when we talking about acronyms of languages? For me, that is the right question. AOO is in process to bring back all localisatons when this brand was OOo. In this point, for me, it will be smart to integrate some standard in that field.
ISO 639 do not helping much, because, for Sebian language, there is only  "srp" (639-3). The older equivalent 639-1 is "sr". "Srp" or "sr", it representing Serbian language. 

Goran told me for different standard: RFC 5646 / BCP 47. 
http://en.wikipedia.org/wiki/IETF_language_tag

With this standard everything is clear.  Authors showed several examples and one of them refers to the Serbian language: 
 "Language subtag plus Script subtag:
      sr-Cyrl (Serbian written using the Cyrillic script)
      sr-Latn (Serbian written using the Latin script) "
(Example: http://tools.ietf.org/html/rfc5646#appendix-A )

So, if we want to go further, we must have clear rules about that.

(In reply to sevalav from comment #11)
> Which standard you want to see in AOO

language codes have to follow what you can already find in the source code:
http://svn.apache.org/viewvc/openoffice/trunk/main/i18npool/inc/i18npool/lang.h?revision=1413471&view=markup#l304
http://svn.apache.org/viewvc/openoffice/trunk/main/i18npool/source/isolang/isolang.cxx?revision=1413471&view=markup#l236

because

> Goran told me for different standard: RFC 5646 / BCP 47. 
> http://en.wikipedia.org/wiki/IETF_language_tag

there is no support currently in the code for language tags, it should be written from zero (which would mean also adapting tons of existing code). Eike Rathke added support for it in other project, but this is unlikely to be contributed here.

(In reply to Ariel Constenla-Haile from comment #12)
> (In reply to sevalav from comment #11)

> there is no support currently in the code for language tags, it should be
> written from zero (which would mean also adapting tons of existing code). 
> Eike Rathke added support for it in other project

Good news. 

 Until that solution, the "sh" will still be used for Serbian (Latin), that si my proposal.
 Nobody wants some incompatibility problems.

(In reply to Ariel Constenla-Haile from comment #12)
> (In reply to sevalav from comment #11)

> there is no support currently in the code for language tags, it should be
> written from zero (which would mean also adapting tons of existing code).

My strong recommendation is  that AOO tim in near future implement that standard.  
Then, instead "sh" will be "sr-Latn" or "sr-Latn-RS". ("RS" = Republic of Serbia), and there will be no more situation like this.
If we have not better solution for last decade (for Serbian localisation), we can wait a little more.

(In reply to sevalav from comment #14)
> My strong recommendation is  that AOO tim in near future implement that
> standard.  

The request is tracked in Bug 109846

does anybody know the correct plural rules for Serbian Latin (tracked by us with "sh") until we have support for language tags

(In reply to jsc from comment #16)
> does anybody know the correct plural rules for Serbian Latin (tracked by us
> with "sh") until we have support for language tags

Maybe stupid question: plural rules in general, or plural rules only for words "Serbian Latin"? Plural rules in English language, I suppose...

plural rules that I can use in Pootle to add the language, see for example http://docs.translatehouse.org/projects/localization-guide/en/latest/l10n/pluralforms.html?id=l10n/pluralforms

sr = nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2);

(In reply to jsc from comment #18)

Serbian language has only one set of plural rules, one grammar, it is one language. But we can writting our words in two way: using cyrillic script or latin script. The same plural rules I must apply for both cyrillic and latin script.

I am not an programmer, but I thing that my previous post can help you.

(In reply to sevalav from comment #20)
> I am not an programmer, but I thing that my previous post can help you.

... I think ...

I see that "sr" is virtually completed in Pootle (when done, please file an issue if you haven't already done so: see http://markmail.org/message/4u45225tjp6lxqop ); can "sh" (meaning, with the conventions in use, "Serbian Latin") be obtained by simple transliteration? In that case volunteers might be able to have "sh" ready by the 6 September deadline too, if Pootle is updated with "sh".

This issue remains open for sh only.

@jsc: when creating the Pootle project, it would be good to import https://issues.apache.org/ooo/show_bug.cgi?id=117936 that was contributed for OpenOffice 3.4.1 (if still available).

Reset the assignee to the default "issues@openoffice.apache.org".

update for Serbian (sr) from Pootle into code(trunk)

If there are any now files for sh I can commit them to the code. We can also open this languae in Pootle for translation.

Otherwise

sr is commited to code (trunk)

For sh please open a separate issue to track it.

so this is fixed