[fauigerzigerk]

The line between backups and live data has become somewhat blurred since file versioning was intruduced to protect against overwrites and deleted files can be recovered. Why were these features introduced in the first place if not to provide some of the features that backups have traditionally been used for?

Clearly, what happened here should not have happened. It doesn't matter whether you call it backup or cloud storage. Google promised to store that data. They failed to do so.

My backups are supposed to protect me against my own mistakes, not against Google's mistakes. Protection against Google's mistakes should be Google's job. They should have redundancy. They should have backups.

If they provide a storage system that does not reliably store data, they should put a big fat warning label on every single one of their products that uses this storage system:

Do not ever trust us to store your data! It could be gone any second. Always make offsite backups!

At the end of the day you are right of course. But the users's mistake is not actually to have mistaken live data for a backup. The mistake is to think that Google reliably stores data when in fact there is absolutely no contractual obligation for them to do so.

[BSDobelix]

> file versioning was intruduced to protect against overwrites and deleted files can be recovered.

That's rapid recovery NOT a backup. A backup should survive an exploding computer or a burned down house.

>My backups are supposed to protect me against my own mistakes, not against Google's mistakes. Protection against Google's mistakes should be Google's job.

Should be someone else job was never good enough for my data, and i think i have been proven right (once again).

>Do not ever trust us to store your data! It could be gone any second. Always make offsite backups!

Yes just trust yourself. Yes always make offsite AND offline backups, if your key/password gets stolen for example "insert storage provider" your data is also in danger.

>But the users's mistake is not actually to have mistaken live data for a backup.

Both made a mistake, one did not make a backup the other one is...well google.

[crazygringo]

> Why were these features introduced in the first place if not to provide some of the features that backups have traditionally been used for?

The key word being "some", not "all".

Yes, file versioning means you don't have to go back to clunky backups to restore an older version.

But it does absolutely nothing for data loss, which has always been the primary purpose of a backup. Accessing older file versions has only ever been a secondary purpose.

[fauigerzigerk]

The primary purpose of backup systems has always been to protect against data loss caused by failing/lost hardware or by accidentally deleting/overwriting data.

Overwrites and deletes should largely be covered by versioning and soft deletes.

Data loss caused by failing or lost hardware should be covered by a business relationship with a data storage service provider.

This service provider role is what's new and different when we're talking about cloud storage. And this is why I reject a direct comparison with traditional backup systems.

Yes you're right, the data should have been backed up to protect against data loss. But why is it the user's job to do that rather than Google's?

I think users should be able to have a reasonable expectation that their backup needs are covered by using a cloud storage service with versioning and soft deletes.

The fact that this expectation isn't met borders on false advertising.

[crazygringo]

> I think users should be able to have a reasonable expectation

Sure, of course they should. But nothing in this life is perfect. Google engineers will roll out a configuration change that has unintended consequences that results in data loss. Or your account gets falsely flagged for abuse and you get locked out. Or your computer gets infected with ransomware that accesses your cloud sync, creating duplicate encrypted files, deleting the originals, and emptying the trash -- so much for your version history.

Don't put all your eggs in one basket, create your own backup. It's "the user's job to do that rather than Google's" because at the end of the day, other people mess up, and it's your personal responsibility to safeguard against that if you want to protect against losing your data. It's always been this way and always will be.

[fauigerzigerk]

I'm not denying that it's a good idea to make independent, offsite backups of cloud data for many reasons (even though some cloud services have protections against your ransomware scenario).

But I think it's a bit unfair to blame users for using something like Google Drive as a backup system, because it does in fact have all the main features of a rather mediocre backup system.

Cloud services are supposed to relieve users of some of the traditional burdens of operating computers, such as making proper backups, copying files to multiple devices, keeping it all up-to-date and in sync.

These things are hard to get right. Most people's backups are utterly chaotic, unreliable, insecure, incomplete and vulnerable to some of the same attacks you describe.

[BSDobelix]

>Google Drive as a backup system,

That's the point, OP had it's data just once (on Google Drive), that's not even a bad backup, it's non-backup.

>Cloud services are supposed to relieve users of some of the traditional burdens of operating computers

"Supposed" was never good enough for my data, but now you have Cloud services AND your computer AND you phone AND your password/key AND the chance that they block your account for whatever reason, i just see more burdens not less ;)