Making Lingo exporting only actually UPDATED files

[r_gigante]

Hi community, please bear with me if the question sounds noob or malposed.

I'm currently investigating how to switch from a home-made help authoring software to Flare+Lingo.
We have our user documentation composed by more than 11K Html pages and we currently support 7 languages.

Our help authoring workflow looks like: author the master in German, translate from German to English and, in the end, translate from English to the remaining languages. All the translations done across the years never went into a documentation translation memory of ours.
Because of this, we've understood we can't make use of any alignment process to build out of the current translated contents a usable Lingo Project.

The idea is to use Lingo-based workflow only for new contents which would turn in the process described by the following steps:

• Create, for each language of our user documentation, a corresponding Flare project (e.g. Flare_EN, Flare_DE, Flare_IT, Flare_FR, ...)
• Create a German-to-English fake Lingo project (Lingo_DE-EN)
• Mark all the entries to be translated in Lingo_DE-EN as "Translated" - even though they haven't been touched at all -
• Export the Lingo Project to generate a faked English project (Flare_DE-EN) with content still being German
• Overwrite the contents in Flare_DE-EN with the contents previously imported in Flare_EN
• Update the Flare_DE with new contents
• Update the Lingo_DE-EN project using the "Update Project" functionality in Lingo to track the changes in the source Project
• Execute a real translation in Lingo for those new items only
• Re-export the project from Lingo to the Flare_DE-EN

Everything would look nice, if not for the fact that I can't make Lingo only export in the last step the newest file coming from the changes I've done in Flare. For a reason unknown to me Lingo re-export everything which is indeed SLOW (11k files) and also doesn't make me implement the flow above.

So my question is: how can I make sure Lingo export only the files that have seen changed by Lingo after running "Update Project" and leaves already exported contents untouched?

Last but not least if you see pitfalls or you could point me to a better approach on managing our documentation beast with Flare+Lingo feel free to drop me a note.

Thanks in advance for reading and, hopefully, for contributing.

Riccardo

[L10N Coordinator]

Riccardo,

having reviewed your post here I have a couple of questions for you.

Can you list all languages needed, because I am having a difficult time seeing why you need a pivot language?

If you have the English and the relevant, why cant you not align? If the documentation matches up (version-wise) you should be able to align and therefore built the translation memory from there.

As for only exporting the updated/changed files that's by design, because when you have a new version for example you want to generate the full version. If for example only some links changed you might miss this in the update and you would generate faulty results if you only plug in the changed files for example.

[r_gigante]

Hi L10n, thanks for reviewing the post and following up here.

Our documentation is authored in German, then it's internally translated in English (1st-level translation) and from this Italian, Spanish, French, Czech, Korean, Japanese, Russian and Polish come from external translator. Maybe it's me being a newbie on aligning already existing localizations using MadCap Align but if I parse a file per time I can get an reasonable (not perfect) segments localization, but If i parse the whole content folder, then the alignment is easily screwed up with even filenames not being aligned.

Any further comment or hint will be more than appreciated.

[L10N Coordinator]

First of all, now that you have listed the languages, using a pivot languages does make sense. In regards to your alignment you are correct thats a big challenge especially with these languages involved. My suggestion would be to check and see if you can convince your superiors to approve the one time expense of alignment through an agency. In this manner you get a translation memory created quickly. While there is a cost involved, the benefits would make this (IMHO) easy to justify.



Uwe