Multi-language support (localization)

[brlima94]

Dear Intel team,

It's nice to see this project getting back to some traction, however I have a few concerns about all the issues which were closed without too much of a criteria.

When ACAT was first released, we built a few language packs to expand this great project's accessibility to a wide range of users from non-English speaking countries. Unfortunately, our effort was lost with the many times this repo had its code entirely deleted and re-added in GitHub.

The latest source code only has EN support and part of the project (ConvAssist) is closed-source. The latest mention of localization/translation I found was in the v1.50 release notes.

I believe this is still a required feature because:

1. This topic was raised in multiple issues over the years, e.g.:

• https://github.com/intel/acat/issues/7
• https://github.com/intel/acat/issues/11
• https://github.com/intel/acat/issues/29
• https://github.com/intel/acat/issues/30
• https://github.com/intel/acat/issues/45
• https://github.com/intel/acat/issues/64
• https://github.com/intel/acat/issues/68

2. One user recently contacted me asking if we could update the language pack created for v0,98, which he was still using (you can check them here if you wish

So I would like to propose if we could give some priority to this topic.

I can open a few PRs with incremental work for each language, but before starting, I would like to request:

1. some commitment that this will not be another wasted effort (someone will review the PRs and provide feedback promptly, merging the code when it's all good)
2. that you do not delete and recreate the code in future releases. Instead, use git merges from your private repo to GitHub if needed. Feel free to ask when localization needs some adjustments, but do not delete everything without notice
3. for ConvAssist to be published or alternate instructions on how to build+debug the project locally
4. instructions on how to create new language packs, even better if I can reuse the n-gram DBs already built with Presage

If you don't mind, I want to start by adding a command-line switch to set the UI language. It was a huge pain point having to install multiple languages in Windows (or use a VM) just to run ACAT.

Please let me know your thoughts and if you can assist me with this - @saiprasadb01 , @michaelbeale-IL , @rdower or anyone else available.

[michaelbeale-IL]

@brlima94 - Thanks for the feedback, and apologies for the delayed response to you on this. A few thoughts on this topic...

1. We hear you on the instability of the code base, and we're hoping things will stabilize now. We've had some internal changes in processes that allows us to "develop more in the open" now, which should add some transparency to what is going on. We're also hoping to also put some version of a feature roadmap together to help give a sense of what work is in our pipeline.

2. Localization the ACAT UI lives! We've been able to prioritize this internally and are on track to have something available soon. You can follow along with our progress on this branch localization. We were able to pull in the localization work that had been done previously in 1.5 and apply it to the new code base. One complicating factor is...

3. ConvAssist - the predictor engine we are using also needs to have localization support added in parallel. Our team is also working in parallel in improving the documentation on what needs to be done in ConvAssist to support localization. Stay tuned for some updates on that repo as well.

Thank you again for your patience with us on this, we're looking forward to collaborating with you more on this.