Gersona

@nijel need some guidance on one point: if the user uploads a `cs.po` file containing the English source with Czech translation, should this file be used as template for new...

@nijel for the sake of clarity, this issue only concerns Cyrillic to Latin transliteration right ? not in both ways

for languages that support both Cyrillic and Latin, is the type of script always informed in the code (like `"sr@cyrillic" or "bs_Cyrl"`)? if so, how to determine current script of...

about the `MachineTranslation.max_score`, since it's just transcribing from a script to another, is it safe to give it a `100` score ?

That's correct, `social_core.backends.linkedin.LinkedinOAuth` is no longer valid and should be removed from `AUTHENTICATION_BACKENDS` - Same for `social_core.backends.yahoo.YahooOpenId` as of https://github.com/python-social-auth/social-core/pull/523

for the sake of clarity, you want to have a field where we can provide context vectors or glossary IDs that will be included in ALL translation requests ?

One additional suggestion is to add a `file_format_params` field to the Project as well, it can be used to store default values for `file_format_params` when creating a new component

@nijel What additional formats should be supported for memory imports ? It makes sense to only allow bilingual format files , but should they all be included ?

Looks good. Can you resolve the file conflict and get the tests running?

How do we emit a warning when `move_language` is executed on `Language.objects.setup()`? It makes sense to log warning to console when executing an admin command, but what's the common way...