As per my knowledge, Serbia, Bosnia & Herzegovina and Montenegro needs both Cyrillic and Latin scripts and Croatia needs only the Latin script. Hence, we need: Serbia-Latin, Serbia-Cyrillic (default for Serbia), B&H-Latin (default for B&H except Republika Srpska of B&H), B&H-Cyrillic (default for Republika Srpska of B&H), Montenegro-Latin (default for Montenegro), Montenegro-Cyrillic, Croatia-Latin (default for Croatia), and actually the proper default for Montenegro is not very certain as per my knowledge. Then, we may use something like localization of region-specific words in Chinese Wikipedia to deal with differences better.

Maybe this can be done with templates, e.g. '''Džordž Vašington''' ({{lang-en|George Washington}})

wich renders in cyrillic as: Џорџ Вашингтон (енглески: George Washington).

Perhaps a template with two parameters would be helpful, e.g. {{foreign|George Washington|Džordž Vašington}} to aid the conversion process.

@Federhalter :

T39617: Do not convert text marked as being in another language with a lang attribute ?

Some people above might misunderstand something. Here, for example, "George Washington" is just Croatian spelling (the Croatian variant does not require spelling of loanwords to be adapted to Croatian orthography), while "Džordž Vašington" is Serbian spelling.

@John_Smith_Ri :

So you would like to have

• sh-Latn-HR : George Washington
• sh-Latn-RS : Džordž Vašington
• sh-Cyrl-? : Џорџ Вашингтон

@Winston_Sung That is exactly what I found in Serbian and Croatian versions of Wikipedia. BTW, although Bosnian version uses "George Washington", I heard that Bosnian speakers also use Cyrillic alphabet, so I am not sure about the case in Bosnian (and Montenegrin). If an improved converter is applied, Cyrillic alphabets can be directly used in Serbo-Croatian Wikipedia, etc. In addition, the Serbo-Crotian Wikipedia is called "Wikipedija" at present, but the Cyrillic spelling is actually "Vikipedija" (there is no Cyrillic letter corresponding to "W"). Maybe this problem can be solved in recent time.

Hi @John_Smith_Ri,

Thank you for your interest in contributing to this language space.

I do want to let you know that, although the language converter still has some small kinks that need to be ironed out, the consensus seems to be that there is currently nothing substantial preventing a merge. The communities are not yet ready/fond of the idea and so this is where we currently need the most support towards this goal.

Do let me know if you have any questions and feel free to reach out directly if there is something that I can help with.

Kind regards,
Denis

The current converter works letter by letter (and 3 two-letter combinations).

An improved version needs to work on a word or phrase level.

It requires the implementation of the rules as described in Orthographic transcription.

This will be probably include maintaining a long list of phrases, some of which could be combined with redirects.

This needs to cover all grammatical cases (e.g. Johna F. Kennedy[j]a).

Templates who allow the editor to aid the transcription are required where the source language is not clear (homographs), or if there is an exception from the rule.

We are currently working on this on-wiki by creating the infrastructure for word conversion. This will include using appropriate templates, such as {{translit}} (for Serbo-Croatian words), {{translitN}} (for titles), and {{foreign}} (for foreign words).