Coming up in 2010

By Edouard on December 18, 2009

Web Translate It is now on feature freeze. What I am currently working on right now is too big to be finished before the Christmas break. So that’s it for 2009!

What’s coming up next?

In the near future (January 2010), I will focus on enhancing existing feature. The stats and the API will be updated.


The new stats will be able to count strings, words, and keep a daily history. You will also be able to generate nice graphs displaying the amount of strings and words per language over time.

To make this work smoothly, some work has to be done under the hood. Calculations have to be deferred instead of being made again and again for every page load. This will use once again the same technology than the file import/export workers that proved to work really well. About 30% of Web Translate It’s calculations currently take place in the background.

The advantage is that not only you will have more detailed stats, but the pages using stats will load faster.


The current API works really well, but doesn’t allow you to fetch a specific language file. This is a problem if you have several language files in your project.

The new API will allow you do just that. Although not advertised anywhere, it is actually already in production and used by some customers.

The API will also allow you to upload an update for a translation file. You could for example set up a periodic task on your server to update your translation files on Web Translate It nightly.

The Web Translate It plugin for Ruby on Rails 2.3 will also be updated to integrate these changes.

Crowdsourced translations

By the end of January, I will tackle crowdsourced translations.

I will chunk this large feature into several, simple iterations. For the first iteration I will focus on working a very simple, yet fully functional crowdsourcing interface.

  • Your own user community will be able to request for a translator account. I will implement login by OpenID to make it easier for everyone to register.
  • You will be able to accept or refuse these requests.
  • Your community will be able to suggest new translations or vote for existing translations.
  • You will be able to make the interface visible publicly or not.
  • You will be able to choose among 3 different translation strategies for each language. For example:
    • Traditional: let professional translators translate that language.
    • Half Crowdsourcing: Let your users translate and suggest translations, and hire professional translators to proofread their translations. You save a fair share of the translation cost, while having high quality translations.
    • Full Crowdsourcing: Let your users translate and vote to decide which translations will go live. It can work very well if you have a huge user community for that language. It usually doesn’t work so well for smaller user communities.

That’s pretty much everything I have planned for now.

I wish you a pleasant holiday, see you in 2010!