I am unsure if this falls into Studio, Web, or Engine - as to my understanding it technically touches all three. The use case is for development testing though, so I’ll keep it as Studio for now.
Use Case
As a developer, I cannot easily play through my game and test which strings have been properly scraped for localization.
Sometimes the text is not translated yet, sometimes the text is not scraped properly, and sometimes the text is not even scraped at all. I only understand English - which makes this process even harder if I use automatic scraping/translation. I cannot easily see improperly formatted scraped strings.
For example, I just went through 2600+ strings in my game’s localization table. This took me over 6 hours. Without a doubt there are still mistakes within the table. There is currently no way for me to easily figure this out through playtesting.
Proposed Solution
Add a locale meant solely for testing, a completely fake “language” or locale - a literal Lorem Ipsum.
By default, this locale should autopopulate with random filler text in order to show what has been scraped and how it would be translated. It should ideally be easy to see where the text starts and stops without being overly repetitive.
Conclusion
Having a “Lorem Ipsum” locale would make it extremely easy for developers to test localization in their games. It would help demonstrate how text is displayed and if it is even scraped.
What I’ve seen game devs on other platforms do is to just play with a locale with an encoding that is very foreign to them (e.g. English devs looking at Chinese/Japanese/Korean characters, or just “wingdings” version of the text) to make it very obvious what is translated and what is not. Not sure if having “lorem ipsum” (e.g. English words) is easy enough since that is still English so it might be hard to parse at a glance mentally what is translated and what isn’t.
(Should also be noted that verifying grammar and substitution of localization parameters in English strings is not necessarily indicative of it being correct in other languages, so you likely still want to have some sort of LQA step done by people that speak that language / have a community feedback channel so they can report issues with translations.)
A big point of my use case is the ability to tell two separate strings apart. Your suggestion covers seeing if something is scraped/translated - but it does not help when it comes to identifying unique strings.
Additionally, it still requires manual work to implement - and that assumes neither the developer nor user uses the locales of Chinese, Korean, and Japanese. For a game that already has live translations out there, it will not work.
