2 ms·
Years ago, I proposed using our research on Zap (https://systems.cs.columbia.edu/projects/zap/ https://systems.cs.columbia.edu/projects/zap/ - arguably the spir
by compsciphd 7y ago
Years ago, I proposed using our research on Zap (https://systems.cs.columbia.edu/projects/zap/ https://systems.cs.columbia.edu/projects/zap/ - arguably the spiritual ancestor to CRIU) to create simple mobile devices (say a tablet that provided a consistent storage environment) and enable the "desktop environment" to migrate from the tablet to a more powerful PC when docked with it (and migrate back to the tablet when one wants to undock).
We demonstrated it using a portable hard drive migrating desktop environments between PCs (https://systems.cs.columbia.edu/archive/pub/2006/09/highly-reliable-mobile-desktop-computing-in-your-pocket/ https://systems.cs.columbia.edu/archive/pub/2006/09/highly-r... - with the portable hard drive providing the consistent storage environment).
We demonstrated being able to checkpoint a desktop (in this case with the fs as well) very quickly (http://systems.cs.columbia.edu/archive/pub/2007/10/dejaview-a-personal-virtual-computer-recorder/ http://systems.cs.columbia.edu/archive/pub/2007/10/dejaview-...)
I argued that Microsoft's work on their Metro interface (i.e. UI elements introduced with win8 - the touch friendly stuff) could actually enable this model with tablet. If developers were able to write UIs for their apps that were able to switch between touch optimized modes and WIMP optimized modes (as lets be honest, they aren't the same and we haven't found anything better for desktop use as of yet than WIMP to the point that Samsung tries to make android play well in a WIMP environmnt with DeX).