Hardware virtualization and emulation based sandboxing for web rendering
Juhapekka Piiroinen
I have been playing with the idea of having the web rendering
processes running in an Linux environment inside QEmu/VMware -image. I
have been looking for the IPC documentation of the Chromium browser as
it would make sense just to separate the existing codebase to
different packages and then handle the IPC over TCP/IP. The benefit in
this would be that there could be a really good protection against
malware and other evil apps from the user point of view. As the idea
would be that each tab would have its own virtual computer and when
the tab is closed the virtual computer is deleted.
Another thing is that the firewall bypassing which is now reality with
web browsers would limit to the virtual machine and the client machine
would be safe, including the personal files. One benefit would be also
that the antivirus applications and system monitoring could be running
inside those virtual machines, which would mean also that you could
assign browsing related scanning resource usage to the minimum.
Or you could have the browser sessions for example in Amazon Web
Services which would mean that next time you would recover full
sessions on browser startup and the sessions would not be dependent of
the location/machine. This would also enable real background
downloading while the laptop is shutdown.
But anyhow, wild ideas I guess, but the thing is that is the current
Chromium IPC compatible with TCP/IP or any plans to enable that kind
of approach?
In this case when there is bit and pieces to be done, it makes sense
to ask a bit dumb questions rather than doing work which is already
done.
Kind Regards,
Juhapekka Piiroinen
Juhapekka Piiroinen
<juhapekka.piiroi...@gmail.com> wrote:
> But anyhow, wild ideas I guess, but the thing is that is the current
> Chromium IPC compatible with TCP/IP or any plans to enable that kind
> of approach?
> In this case when there is bit and pieces to be done, it makes sense
> to ask a bit dumb questions rather than doing work which is already
> done.
I guess I could clarify my dumb question a bit more :P ..so the
document in the http://www.chromium.org/developers/design-documents/inter-process-communication
states that in Unix Chromium used socketpair and in Windows it uses
named pipes. So at least in theory the Unix side should be easy to
change and I guess in the windows side some kind of named pipe to tcp/
ip service should be done. Better ideas?
- Juhapekka Piiroinen