Data Persistence
Another post where I jotted an outline and then didn’t feel like fleshing it out:
*number of computing devices used by a person burgeoning (home, work, e-cafe, phone, library, car, tv, fridge, …)
*management of information becomes exponentially complex across geography and hardware/software boundaries
*hardware failure/loss often results in irreparable damages, and backups are rarely complete and up-to-date
*virtualized data solves these problems…
*information lives on numerous devices with on-demand access for more devices
*universal system failure much less likely
problems…
*memory availability across devices (1TB PC vs 1MB phone)
*network transfer sometimes unreliable or lethargic (cable upstream icky)
*privacy, authentication, information poisoning
*network disconnection
concessions:
*not all data needs to be universally accessible, only “unique” data: personal settings & documents – configurations, bookmarks, photos, spreadsheets, savegames, logs
*needs at least 2*workingset total system storage to mirror everything
*avoids transfer of huge amounts of data, and large documents can be previewed before committing to full transfer
*compression and versioning can trade CPU work for network bandwidth
*device peering can avoid complete network disconnection (cellphone<->notebook while on plane)