The ironic thing about this language is that it is an attempt to be "inclusive" by introducing weird shibboleths that exclude most people.
I was once informed by someone at a big tech that the reason for these "x[n]y" codes is because these words can be hard to spell for non-native speakers of English. Also, they added "a11y" (accessibility) to the list.
I think the real explanation was that these were introduced because "internationalization" is a long word and file path names used to be limited in the 1980s.
This covers strings, but what do you do about right-to-left languages, cultural differences in (say) information density, date pickers, color usage, UI preferences, emoji use, whatever? Did you make a conscious decision to only localize strings?
What do you mean with showing you the diff BEFORE you merge? Do you run a script comparing your current repository translation files with the ones on the google spreadsheet?
Our project is Logto, an open-source solution that helps you build the sign-in, auth, and user identity within minutes.
The first public beta launched two months ago, and it is incredible that many people love our project and are willing to contribute even before our “Bounty Hunter” series. The community contributed translations of Français, Türkçe, 한국어, and Português, along with connectors for Azure AD, Kakao, Naver, and Discord.
We cannot be here without the community’s help, and we’d like to hear more from you. Feel free to let us know if Logto can help you or what you expect from Logto. Cheers!
Hi, this is super cool, congrats! I went through your documentation and it looks like your contributors have to manually edit files. My friend and I are currently working on an open source git-based translation editor and dev tools to ease contributing and managing translations. Check out our repo and let me know what you think! https://github.com/inlang/inlang
Highly recommend integrating with CrowdIn, it's free for OSS and makes managing the translations _much_ easier, you can have their git integration push approved strings directly to your codebase.
The community has been fantastic for translating Outline too, we've dealt with some small bits of abuse and misunderstandings but the result is currently 21 languages supported, most with over 80% of the application covered. It's taken about 2 years to get this far…
It certainly makes you realize that translation is as much an art as a science. For example we have a core concept of "Collections", many languages this could be translated in multiple different equivalents and we want to make sure that the closest is chosen. Misunderstanding might be someone that translated to Collection -> Drawer which isn't really correct.
Abuse has been minimal, mainly things like adding smiley faces into translation strings or just what you'd consider to be purposefully poor translation here and there. We use CrowdIn to manage the community and it provides tools to ban contributors in these cases.
Readit News
We put all the translations in a spreadsheet, https://docs.google.com/spreadsheets/d/1FoYdyEraEQuWofzbYCDP.... We didn't like the idea of storing in any more complex than a csv file.
All changes are shown to us in a diff before we merge, the csv is then downloaded and the i18n files created automatically after we merge the branch.
We've never had a single instance of abuse in over 10 years.
Mozilla has a good article explaining the distinction: https://blog.mozilla.org/l10n/2011/12/14/i18n-vs-l10n-whats-...
a10g means "abbreviating" by the w1y.
I promise it will not burn up too many more calories of soylent to just type it out, nor will it take much longer.
I was once informed by someone at a big tech that the reason for these "x[n]y" codes is because these words can be hard to spell for non-native speakers of English. Also, they added "a11y" (accessibility) to the list.
I think the real explanation was that these were introduced because "internationalization" is a long word and file path names used to be limited in the 1980s.
Our project is Logto, an open-source solution that helps you build the sign-in, auth, and user identity within minutes.
The first public beta launched two months ago, and it is incredible that many people love our project and are willing to contribute even before our “Bounty Hunter” series. The community contributed translations of Français, Türkçe, 한국어, and Português, along with connectors for Azure AD, Kakao, Naver, and Discord.
We cannot be here without the community’s help, and we’d like to hear more from you. Feel free to let us know if Logto can help you or what you expect from Logto. Cheers!
https://github.com/tolgee/tolgee-platform
https://github.com/outline/outline/tree/main/shared/i18n/loc...
Abuse has been minimal, mainly things like adding smiley faces into translation strings or just what you'd consider to be purposefully poor translation here and there. We use CrowdIn to manage the community and it provides tools to ban contributors in these cases.
Thousands of people wasting minutes trying to find out what that is.
Deleted Comment
Deleted Comment
Heads up, the animations on your website are pretty laggy on Firefox (Firefox 104, Windows 10 1909).