No Google source code for Viacom, just 12TB of YouTube data
By Nate Anderson | Published: July 03, 2008 - 01:15PM CTAs part of the discovery process in its $1 billion lawsuit against YouTube, Viacom asked for an astonishing array of information: the source code for the search functions that power Google and YouTube, the source code for YouTube's new "Video ID" program, a complete set of every video ever removed from the site, databases containing information on every video ever hosted at YouTube, and a copy of every private video. The judge has now ruled, saying no, no, yes, yes, and no. Even with the limited discovery, some poor drone at YouTube can now look forward to weeks of copying information onto Viacom-provided hard drives. ...
Keepping the "no"s, the "yes"s are:
But Viacom did better when it came to accessing YouTube's massive databases. The first database requested contains all videos ever removed for any reason, which Viacom hopes will show just how many infringing videos YouTube has hosted over the years. The massive database will "require a total of about five person-weeks of labor without unexpected glitches, as well as the dedication of expensive computer equipment and network bandwidth." The total number of videos here is "intimidating," wrote the judge, but he granted Viacom's request anyway.
Next up is a 12TB database containing logging information on every video ever watched at YouTube. Viacom wants to see just how often infringing clips were viewed, then compare this against noninfringing ones to prove its contention that YouTube, in the early days, was an engine powered mostly by the gasoline of illegal content. The database will also show which username and IP address watched every video, a move with potential privacy implications.
The data set is large, but the judge noted that it could be slapped on three "over-the-shelf" 4TB drives. Request granted, score tied at 2-2.
Delicious
Digg
Reddit
Newsvine
Furl
Google
Yahoo
