Customer success storyHow Bigpoint achieved smooth continuous game localization using Git version control system and XTM Cloud
Video game developer
Bigpoint develops online games which get continuously updated. In some cases, these updates can continue more than a decade after their launch. As such, Bigpoint wanted to be able to follow the continuous flow of translations of their most lasting products as localization is a key step in a game’s continuous development. Bigpoint needs to ensure that the process is smooth, quick, and error-free in order to deliver the best product to customers worldwide.
However, the traditional localization process being used meant this wasn’t possible. Bigpoint’s workflow included many manual steps, leading to lots of back-and-forth and slow turnaround times. The localization process for strings looked like this:
The games team sent new content from SVN, their version control system, to a proprietary translation tool.
The content was exported and sent to suppliers in XTM Cloud for translation.
After being translated, the strings then had to be imported back to the game repository.
The whole process could end up taking days, with the export time alone equating to between two and four hours every time. This system also struggled to cope with the updates and regular changes which are commonplace in game-development projects, and continue even after the project release. The result was that many strings were outdated before their translation was even completed, making version control more complex.
The original process also led to a number of quality issues, such as:
Untranslated strings due to extraction issues
All these problems had to be solved reactively, rather than being spotted and resolved earlier in the process, which took up yet more time. Long delays and quality issues meant it was difficult for Bigpoint to keep up with their games’ continuous updates.
Bigpoint’s key challenges:
Too many manual steps leading to slow turnaround times
Outdated content being sent for translation due to version control challenges
Quality issues occurring because of the inability to get in-context previews
Bigpoint recognized that they needed to move from a manual system to a Continuous Localization (CL) workflow for a quicker, more predictable system. In 2020, Bigpoint began using XTM Cloud and Github to manage their localization process. Thanks to XTM Cloud’s GitHub integration, Bigpoint was now able to send projects for translation in one easy step from their authoring tool.
Projects are created in XTM Cloud and content is sent automatically from Git. Bigpoint’s development teams make changes in their source text directly in their git repository and push their new or updated content to the master branch. That is where the magic happens! As soon as new or updated source text is pushed to master, the git connector recognizes it and automatically creates a new translation task containing only that selected text. It extracts corresponding strings from the source files and sends only those specific strings to XTM Cloud, making Bigpoint’s localization project managers’ work much easier, as well as the translators’. Once translations are finished, they’re transferred back to Git for publishing, and any changes are updated in real-time in both XTM Cloud and Git.
Key features Bigpoint used to achieve a CL workflow:
Unique string IDs which make ID-based matching possible, improving leverage and reuse, saving time and money.
Real-time updates for a single, continuous localization process enabled by the fact that XTM is a cloud-based solution.
XTM Connect for Git – QA search can be used to search inside XTM’s Connect’s user interface to view and display assets to edit.
XTM Connect for Git – QA task creation can collate individual strings from multiple assets/files held in Git into a single QA project in the XTM Connect user interface. Issues are resolved as soon as they appear.
By implementing a continuous localization workflow, Bigpoint was able to design a smoother and quicker process, avoiding duplication and redundancy. This made the dream of continuous localization a reality, helping Bigpoint fulfil their mission of making great games accessible to everyone — in the right languages and at the highest quality.
Moving to a CL process has saved Bigpoint time, money, and stress. Previously, Bigpoint needed two-plus hours to create QA tasks. Now they do it in two minutes. The XTM Cloud Connect for Git means QA issues can be spotted earlier and resolved quickly, no longer requiring hours of manual work.
Overlapping text, untranslated strings and inaccurate translations have been hugely reduced by the use of the In-tool QA feature, resulting in higher quality translations and a better game experience for players. Using this feature also prevents duplicating work and translating redundant strings as the game evolves during the development process.
Using a CL workflow simplifies the process. Everything localization managers need is in one place: they can search the Git assets using word search or one or multiple string IDs without leaving the XTM Cloud platform. Changes are made in real-time and updated in both XTM Cloud and Git so there’s no need to painstakingly compare versions during QA.
The end result for Bigpoint is a process which:
Has fewer manual steps
Has the capacity to spot and resolve errors earlier
Is more streamlined and saves time
Consistently delivers accurate, high-quality translation
Using the Git version control system and XTM Cloud meant that Bigpoint were able to minimize errors, save on their overall localization costs, get their products to market faster and have peace of mind that every aspect of localization was covered.
Start facing the challange of going global and embrace your success now
Ready to see this technology in action? Leave us your name and email address below and let’s speak soon.