> I don't want to do it locally, since that puts me on the hook for assuring physical safety and backups of my backups.

Regardless of where you store your personal data, the proverbial buck always stops at your desk when it comes to assuming responsibility.

Paying someone else to store your data gives you someone to hold accountable to what happens to your data. But that accountability is always limited and doesn't dismiss you from managing that relationship. For instance, the other party is still entirely free to bring terms and conditions to the table, which are meant to mitigate liablity as far as the party which will host your content is concerned.

Put more succinctly, even when you rely on a cloud service, you're still on the hook to assure safety and backups of your own data.

> I suppose I could roll my own solution, but frankly I get tired of maintaining my one-off projects like that after a year or two.

Yeah, I'm gonna prod you on that one. :-)

For e-mail, you could run an script once a month that harvest all your e-mails over IMAP and roll them into TXT, HTML, EML,... whatever. You could go for one big file, or discrete files. You could even pain yourself and fit everything in an mbox file. Roll everything in a ZIP or TAR ball. Then you'd store output file in whatever service you want.

You could spin up a VPS, provision the entire thing and set up a script that does all that as some service or cron job or whathaveyou. Now you've got an entire stack to take care off. Not what you want.

Maybe this is a good use case to go serverless. You could string something together using AWS serverless. Write an AWS Lambda in Python or Java. Should be one single script. Store your e-mail backups in S3 or somewhere else.

The idea of such a solution is that it just runs in the background and it runs in a robust fashion with not too many moving parts. So, here you have something running indefinitely as long as the bill gets paid. Or amazon doesn't cut you off. But at least you've now got your e-mail mirrored on two independent services, which is marginally better then where you are right now. The only other major constraint here is being able to restore your e-mail from that backup in case of disaster.

This is a project I started working on but never finished. The problem is that people didn't care enough to pay for backups because they (are conditioned to) believe that Google is already safe and won't go away. The cloud is this magical thing that saves everything and unfortunately they don't realize otherwise until it stops working.

Because it costs more than a few dollars a month. The net cost to users for Google is zero (or a token amount) because Google makes their money from data mining and selling your meta data.

Email addresses and inboxes are important to people. They are critical to our online identity.

Wiki mentions Gmail having 1.5 billion users. Imagine even 100M of them paying $100 per year. Makes Gmail a $10B product just on paid subscriptions, that is not a small amount.

Amazon cloud storage is $1.99/mo for 100GB. That's exactly what I get from Google(I pay $1.99 for extra storage).

Are you saying the software to perform the syncing costs more than a few dollars per month?

> Are you saying the software to perform the syncing costs more than a few dollars per month?

The operational costs could be more than that per user if you want regular updates. That's disregarding engineering time.

You need stable business with at least few people employed, to make sure the lights don't go out, and that people fix issues when they come up. Storage cost is probably only small part of the price.

You also don't want it to be VC backed because such businesses will not be worth billions. (and if it will try to become such you will have similar problems as google)

For a competing service maybe, but they're talking about backups. I find the bigger issue is that not enough people believe they need a backup for the "cloud".

> since that puts me on the hook for assuring physical safety and backups of my backups

Only in the event that Google locks you out, at which point you can create a new backup elsewhere. If Google has a copy and you have a physical copy, then Google would have to lock you out at the very same moment your house burns down for you to lose any data.