Table of Contents

Background Information

In the past, those who wanted bilingual Japanese and English setups of JIRA on the same instance to rely on a Japanese specialized version of JIRA which is offered by imaHima, this version is usually a few increments behind the official Atlassian releases (currently up to version 3.12.2 from Atlassian Japan) an example of the imaHima Japanese-English JIRA implementation includes the Adobebugs public JIRA.

Ever since the official JIRA 3.13 release, Japanese has been added as a supported language pack that comes with the base product, many will find this to be
more appealing than the imaHima release as it contains more features and is fully supported by Atlassian, as such, some organizations have decided to implement a Japanese and English instance of JIRA 3.13. As this feature is quite new to the official JIRA product, I have conducted some investigations to learn the full Japanese capabilities made available in JIRA 3.13.

Description

I have recently completed a bilingual (Japanese and English) implementation of JIRA 3.13 for a client, and there were several key discoveries during my investigations to the internationalization capabilities and limitations of JIRA 3.13 in regards to this, I will break up my findings into 3 sections, the first will focus on the capabilities of the system to handle the actual bilingual data (mainly storage and display), the next one will focus on the capabilities of the functionality of the system (indexing, searching, etc.) and finally the presentation (basically what the user will see).

Bilingual Data Handling

The actual ability to handle the characters being entered into the system seemed to be flawless (as long as the underlying database could handle it), being able to store and retrieve anything in both Japanese and English with no apparant problems, both for field values, configuration items and even usernames and passwords, I have done some testing just to see what can be handled digitally in regards to the Japanese input types and compiled the following table:

Input Type	Field values	Login Details	Configuration Items
English
Japanese - Alphanumeric
Japanese - Hiragana
Japanese - Full-Width Katakana
Japanese - Half-Width Katakana
Japanese - Kanji

The only limitation that the system has is that the keys for the project cannot contain Japanese characters, but this should be understandable as the project key is used in the URL link for the project as well as any issues within the project and does not pose as a problem from the user's perspective.

System Functionality

The key note in this section is the searching module of JIRA (based on Lucene's indexing mechanisms), although this functionality is functioning as expected for a large portion of the data within the system, there are some cases where they are unable to come back with the proper results, as the searching system is completely depending on the indexing and not the database, the underlying database encoding does not affect the results of this investigation (we have tried this using multiple database encodings with the same results). The following table details the various input formats and what results they produce:

For the following tables, the first column specifies what input format was used to search the system, and each subsequent columns shows if the equivalent in each format is being returned as a result or not.

Indexing Language - Chinese/Japanese/Korean

Input Format	English	Alphanumeric	Hiragana	Full-Width Katakana	Half-Width Katakana	Kanji
English
Japanese - Alphanumeric
Japanese - Hiragana
Japanese - Full-Width Katakana
Japanese - Half-Width Katakana
Japanese - Kanji

Indexing Language - English

Input Format	English	Alphanumeric	Hiragana	Full-Width Katakana	Half-Width Katakana	Kanji
English
Japanese - Alphanumeric
Japanese - Hiragana
Japanese - Full-Width Katakana
Japanese - Half-Width Katakana
Japanese - Kanji

Indexing Language - Other

Input Format	English	Alphanumeric	Hiragana	Full-Width Katakana	Half-Width Katakana	Kanji
English
Japanese - Alphanumeric
Japanese - Hiragana
Japanese - Full-Width Katakana
Japanese - Half-Width Katakana
Japanese - Kanji

System Functionality - Summary

From the results it seems like there is always 1 case where the search could not produce the desired results, this is a bit of a problem, so it is best to warn the users to avoid the problematic one when possible when using one of the cases.

Presentation

JIRA uses language packs to apply various languages to the default JIRA system, with a special language pack for the Japanese language, in most of the system, the translations are quite decent, however there are several cases where there were problems, the most noticeable ones included:

1. "with open issues due to be fixed per version" - which did not get translated in the language pack (left as English)
2. "A change log of the recent versions for this project " - which was caused by an error in the i18 parameter name within the com\atlassian\jira\plugins\projectpanels\changelog_ja_JP.properties file, the properties should be projectpanels.changelog.description and projectpanels.changelog.name as opposed to componentpanels.changelog.description and componentpanels.changelog.name

A JIRA issue has been raised in regards to the language pack, and since then, I have made the requested changes (with the help of a Japanese client) to the language pack and have submitted the changes back to Atlassian (attached to the issue), hopefully this will get incorporated back into the base product in the 3.13.2 release of JIRA.

Overall Summary

Although there are some issues in regards to the bilingual setup of JIRA 3.13, most of the issues are either cosmetic (in regards to the presentation section of the system) or minor with workarounds, so as such, JIRA is very capable of dealing with dual Japanese-English setups, and with the future improvements of the system (in versions 3.13.2), I expect the overall quality and capability of JIRA to increase beyond where it is now.