How does the "flip the bit" scenario work?

[timuckun]

In your "flip the bit" scenario. How does the system find everybody with the existing version of the operating system and distribute my code over to their machines so the bit can be flipped.

Once the bit is flipped the file gets a new hash, is that stored as a new object?

How is the new object indexed as the old object URL so it can be found again to perform another update in the future.

Does the old object stay in the system taking up useless space and perhaps a dangerous operating system which somebody might accidentally install?

What if the machine is offline? Does it eventually get the update when it goes online?

Is there any way to submit a status report back to the issuer of the "patch" to verify the patch was executed properly?

[dborzov]

Kolmoblocks and IPFS's data model is immutable, so the unpatched and patched versions would obviously have different hashes that never change.
The way that "flip the bit" scenario can work is, we can have an HTTP URL that would have the hash of the "most recent update" version of the distro. The phones can poll the URL once in a while and attempt to get the most recent version of the distro advertised by the URL.

When the new update is patched, the hash posted at the special URL changes. Phones now proactively start looking into how to get the new version of the distro. They will be able to discover that there is a kolmoblock that is small in size and has only one dependency block: the one they already have, the unpatched version. They would be able to download and execute it instead of downloading the new version all over again.