2018-08-12

EPA-Translation as Community-Driven Localisation for Web Apps



Just redefined description for epa-translation:

the community driven translations of web application will extend the application audience beyond the initially targeted countries and lingual dialects.


Why I have started a new project? Ideas are cheap, "every day a dozen" appear and are back logged forever. But this time it became another project to worry about. Worthy of the effort, ApiFusion.com and EPA-WG are actively in progress. 

First, a bit of history. When a new more or less valuable idea appears, it is usually written out and sometimes shared with friends who could have potential interest. Usually that is where its life ends. But this time, Hogne Titlestad from FriendUp invited me to discuss it in their IM group. Apparently the work on translation is in progress there. 

To make it visual for discussion, I thought 3 hours would be sufficient for proof of concept. That is when the project started. And as usual the appetite came during the meal, making me feed-frenzy towards the unusual challenges.

The challenge factor is crucial for hobby style projects. Otherwise, why bother, right?  However, it is not just the challenge swung me to work on proj. The EPA-WG needs a few clear microapplication samples, better in web 3.0 context( cross- service vendors UI and integration ). So far epa-translation seems to be complex enough in functionality and simple enough to understand. 

There is final reason to get it done: epa-translation fits into concept of community-driven content and knowledge/culture evolution which drives modern civilization to progress, where only collective intelligence could counter-balance AI. ( I do recognize other options such as AI and human fusion, but that is a bit in the future. Rather, community initiatives are acting now.)

epa-translation is a part of EPA-WG (Embeddable Progressive Application Working Group) initiative. #OpenSource hosted on GitHub.


PS. This project's challenges came from various sides:
* how to make the translation process seamless?
* how to embed microapplication into host app context?
* How to split several flows ( view, edit, sync to cloud) into independent microapplications?
* What radically new pattern microapplication gives and does it give significant advantages?
Some in HTML5 
* How to identify the text chunks in page, select and substitute?
* What local storage to use - localStorage vs IndexedDB? There is large matrix of pros and cons on each side.

Cloud layer
* Should it be server-less based?
* Should the platform be open?
* Could distributed DB be used? BlockChain vs CDN.

I am pretty sure the list of challenges will grow.