The Washington Post (reg. req.) has published a Reuters wirestory on the use of local caching (this is new?) by Napster for the delivery of their streams and downloads (Napster, IBM Unveil New Music Delivery System):
The most popular tracks in the Napster system are stored locally, enabling customers like The University of Rochester and Penn State University, to reduce their computing infrastructure's vulnerability to overuse.
Hmmmm ... but how effective is this overall? After all, Napster supports a streaming model, where the same song can be streamed over and over again to the same listener. That's an efficient use of bandwidth, not.
And why does IBM have to be involved? Why not structure the downloads through something like BitTorrent where the student's computers carry most of the load?
via Furdlog
UPDATE
More at C|Net News (Napster, IBM aim to save cache).
I use Rhapsody, and it caches about 500MB on my PC. I suppose if it stores a cache on the University server, that would cut back on Internet traffic, and storing another cache on the user-PC would cut back on intranet traffic... (I don't know if Napster has a local cache like Rhapsody's)
Posted by Scott Matthews on March 10, 2004 06:30 PM | Permalink to Comment