tracking transcriptions/translations

[willkg]

We've enabled Amara for all the videos on pyvideo. That's great.

However, I want to be able to generate a report for a given conference that tells me the following:

• which videos were transcribed? (or alternatively how much of which videos are transcribed?)
• for the videos transcribed, which languages are they translated into?

Why? Because transcription and translation can only be done in one of three ways:

1. I do it myself. Due to severe chronological constraints, this isn't feasible in any meaningful way.
2. I pay someone to do it. I've tossed around trying to raise money to transcribe videos at conferences. I'm not sure we'd do all the videos, but rather some subset of videos that are "interesting" for some method of determining that.
3. We all work on transcriptions/translations. If we go this route, then one of the key roles is to keep track of where we are and proactively reach out to get help for the things that aren't done, yet.

In order to do number 3, I really need to be able to see where things are at.

I think there's an Amara API for determining this. There are a couple of things I don't know.

1. what's the unique key for each video that Amara uses to determine what we're asking about? should we be storing this key in our db?
2. how do we ask Amara about videos? do we need an account? an API key/token?

This issue covers figuring all that out and implementing at least a Django management command that generates a report for a specified category.

[willkg]

A better idea might be to have the command run and toss the data in a couple of models so we can run a cron job once a week, update the data, and then show the report on the site.

If we maintained historical data, then we can see improvement over time which would be great. Something like:

Transcriptions:

            4/1    4/8    4/15   4/21
PyCon 2014  0%     0%     8%     15%

Then maybe you click on PyCon 2014 and get the video-by-video summary of transcribed and translated.

This has the interesting effect that conferences can compare themselves against one another and previous years. Does this comparison urge people proud of their conferences to do better?

I'm game for making that a future iteration.