MAT v4.0 Beta update 1 is available
The Multilingual App Toolkit v4.0 Beta Update 1 (v4.0.1413.0) is available. This release contains customer reported fixes for issues with Windows Store projects as well as Translation State counters and better filters in the Editor and improved results from the Language Portal provider.
Windows Store project issue
Customers were reporting issues with some, but not all Windows Store Projects. Across the board most projects worked as expected so it was hard to track down this issue, A repro was provided by the author of ‘Paint 4 Kids’ that allowed us to see the root cause. The repro required the project’s name to end in a language culture code – in this case a MAT 4.0 test project ‘MatRU’ was causing issues with Russian as a target language.
This issue is manifesting as two separate bugs. One bug reports the error "A resource file referenced by 'filename.xlf' was not found in the project". The second silently fails to add the target resource file when adding a new language, but would surface during the build operation as the first issue. Both of these issues are fixed with this update.
Readiness of Target languages
MAT considers all resources are translated unless they have a state of ‘New’. This works great for added translations into a project, but can make is hard for a developer to known if their language is ready to ship. While we do track and display indicators and apply filters in the editor to let you see the different states, we felt that is data should be always available.
The Editor has been updated to so the counts of the different translation states are visible status bar. This allows you to glance to see the translation state of your resources. We also moved Translation State filters to the primary Ribbon as well as changed them to checkboxes, allowing filtering of multiple states at the same time.
Note: The status bar always indicated the state of the current view. Be sure to reset all filters to see the status of the entire file.
Improved Language Portal results
I’m sure the most will agree that nothing beats direct human translation. However this may not always be an option for every app or even every language in an app.
One of the challenges with providing automatic translation is to figure out which result is the best available option. In a recent review with the Language Portal team, we decided to tune the request to tighten the result set just a little. The result was an overall better match with translations from the Language Portal service.
One such example was the translation of ‘Help’ into ‘German.’ Based on our previous settings as well as post processing, the best result was ‘HELP-TASTE’. Even as a non-German speaker, this was not exactly what I would call the best available option. With the change using case sensitive matches the result is now ‘Hilfe’, which is a much better option. There are other similar examples, but overall this change improves the top result from the Language Portal Provider.
Thank you,
The Multilingual App Toolkit team
multilingual@microsoft.com
User voice site: https://aka.ms/matvoice
v4.0 Technical Preview: https://aka.ms/matpreview
Twitter: @CameronLerum