Thoughts on Google music…
Google announced Google music on this years Google i/o conference, a launch that has been anticipated by the pundits for a long time. The service looks like it could be very big as it allows a maximum number of tracks instead of a maximum amount of storage to be uploaded. What was unexpected was the fact that, as Amazon had done before, Google launched without having a deal in place with the record labels. Google didn’t elaborate on this, but the general consensus seems to be that they probably don’t need a deal to begin with.
As the first invitations are pouring out (of course i have already applied for mine) I wondered how Google manages this. Google’s redundancy model usually protects against data-loss by mirroring your data across multiple datacentres. But even on Google scale, mirroring 20000 tracks per user (at 4-7 mb a pop) to 2 or more datacentres has got to hurt. Even if most users never come near the 20000 track limit, I cannot imagine Google engineers being happy about storing “Baby hit me one more time” 3 million times. So this got me thinking if Google maybe has some data de-duplication going on somewhere. Even more awesome would be if they were to include matching your content on your hard-drive against fingerprints stored online so that you wouldn’t need to upload your entire music collection. Oh well, one can dream..
No related posts.
