Difference between revisions of "Viddy"

From Archiveteam
Jump to: navigation, search
m (Corrected user count.)
(Updated status on API scraping)
Line 37: Line 37:
 
==== API Scraping ====
 
==== API Scraping ====
  
Through scraping the public API (reference for dedicated API [http://developer.viddy.com/docs here] - this mostly applies to the public API: http://viddy.com/v1/... ), computerfreak compiled a list of ~535k user IDs and ~1.05 million media IDs. The user IDs are available here: [http://archiveteam.nickproductions.com/viddy/viddy_user_ids.csv.gz viddy_user_ids.csv.gz], and the media data (UserID, MediaID, comment_count, like_count, view_count) here:
+
Through scraping the public API (reference for dedicated API [http://developer.viddy.com/docs here] - this mostly applies to the public API: http://viddy.com/v1/... ), computerfreak is compiling a list of user IDs (currently ~4 million retrieved) and media IDs (currently ~500k retrieved). These are available [http://archiveteam.nickproductions.com/viddy/ here] (user IDs in the viddy_user* files, and the media data (UserID, MediaID, comment_count, like_count, view_count) in the viddy_medias* files).
* [http://archiveteam.nickproductions.com/viddy/viddy_medias.csv.gz viddy_medias.csv.gz] - original set of ~1.05 million
 
* [http://archiveteam.nickproductions.com/viddy/viddy_medias_run_2.csv.gz viddy_medias_run_2.csv.gz] - An additional 443k media ids that were missed in the original set (forgot to specify count= in the request URL to return all videos).
 
  
The user IDs were obtained by doing a BFS on user's followers (script stopped after about 2 days of running), and the medias were then obtained by getting the media listings for each user. Obviously this is not every single user or every single video, but it is certainly a good starting point.
+
The user IDs were obtained by doing a BFS on user's followers (script stopped after about 2 days of running), and the medias were then obtained by getting the media listings for each user. <s>Obviously this is not every single user or every single video, but it is certainly a good starting point.</s> As of Nov. 29, computerfreak is continuously running a script to get new user IDs and is pulling off the new IDs and fetching the associated media info.
  
 
== Download Your Data ==
 
== Download Your Data ==

Revision as of 04:09, 30 November 2014

Viddy
Viddy logo
Viddy - Capture Life in the Moment..png
URL http://www.viddy.com
Project status Closing
Archiving status Not saved yet
Project source viddy-grab, viddy-items
Project tracker viddy
IRC channel #viddiot (on EFnet)
Project lead Unknown

Viddy is a video hosting service like Instagram. The service has its own url shortener service viddy.it, just to make sure the impact of the closure is even worse.

Viddy announced to be shutting down December 15th 2014 after joining Fullscreen.

Shutdown

Message on viddy.com website:

"Viddy is powering down, but you’ll hear from us again soon…
We’ve been building something new and exciting since joining Fullscreen in January, and as we focus on the future, Viddy will be pulled from the App Store on Tuesday, November 4th, 2014, and will officially shut down Monday, December 15th, 2014."

Site Structure

  • www.viddy.com/JustinBieber/v/soundcheck-LDS3p9
  • www.viddy.com/media/af2745dd-d085-4b3a-9017-085f72f6334a
  • www.viddy.it/Syv46W

How can I help?

Content discovery

Viddy item discovery is being done through the URLTeam project, scraping the viddy.it shortener. If you want to take part, join that project, see details on its page. (You'll get tasks of other shorteners as well, they are distributed randomly. You should start multiple threads.)

API Scraping

Through scraping the public API (reference for dedicated API here - this mostly applies to the public API: http://viddy.com/v1/... ), computerfreak is compiling a list of user IDs (currently ~4 million retrieved) and media IDs (currently ~500k retrieved). These are available here (user IDs in the viddy_user* files, and the media data (UserID, MediaID, comment_count, like_count, view_count) in the viddy_medias* files).

The user IDs were obtained by doing a BFS on user's followers (script stopped after about 2 days of running), and the medias were then obtained by getting the media listings for each user. Obviously this is not every single user or every single video, but it is certainly a good starting point. As of Nov. 29, computerfreak is continuously running a script to get new user IDs and is pulling off the new IDs and fetching the associated media info.

Download Your Data

"We’ve made it easy for you to download your content before this date [December 15th 2014]! Please click the button [...] and log in with your username / password to get started."[1]

The link: http://www.viddy.com/download

References