I tweeted about this a while back and several people had good replies: https://twitter.com/aarondfrancis/status/1445408384472211464

For context, this was after facebook was down for a day. I think the odds of Google deleting all my photos is pretty low, but the odds of me losing access to my Google account for legitimate or non-legitimate reasons is much higher.

https://github.com/jay0lee/got-your-back

2. buy a new drive every year and copy it forward

3. store a copy with relatives

Note that a cloud provider:

1. can shut off your access at any time, by accident or on purpose, and you have no recourse

2. will sell your data to anyone at any time

3. will give it to any government for the asking

4. will leak it to any hacker

5. will allow employees of said cloud provider to browse your data

6. will scan it to use it to sell you ads

7. will scan it looking for anything they don't approve of this week, and take "corrective action"

8. if you've got terabytes, good luck trying to download it

2. Copying Error ( Not sure if that is still a thing on Windows )

3. HDD failure.

A cheap, ZFS, consumer friendly, Dual 2.5" HDD DAS would have been great. Unfortunately no one is making one. May be there is no market, May be they just dont care.

Or burn it to an Archival Disc. Apple could have make an iOS Time Capsule. But they are too busy to push their service revenue.

IIRC ZFS is more trouble than it's worth on tiny systems due to the RAM requirements... generally speaking if you care enough to want ZFS specifically you'll get a slightly larger QNAP / similar NAS (I think the smallest ZFS capable qnaps are 4-bay), or just DIY the hardware and install FreeNAS (or similar) on top.

It might work its way down to the consumer segment eventually but given cloud is pushed heavily (and often you get a decent chunk of storage for free with M365 / google / whatever), needs no end-user know-how, and provides a touch of geographic separation in case of a house fire or other minor disaster it's hard to see tiny ZFS boxes competing vs. cheap cloud storage.

The 2nd thing is that you are not using the NAS / DAS frequently in this scenario. The memory only get you extra performance. The only reason why you want ZFS is file integrity.

[1] https://news.ycombinator.com/item?id=11897571

For 8, cloud storage services like Backblaze B2 can mail you a hard drive with all your data on it. I believe with B2 you can also return the drive afterwards and get your deposit on it back.

Encrypt everything

Check drives (health etc)

Etc

1) Someone can use those to train DeepFake models and scam your family

2) There could be private photos and videos there that you aren't aware of

3) Why not?

For 3 the biggest problem with encryption is you introduce a single point of failure, if you lose access to the keys your photos are gone. On a simpler note I'm a big fan of keeping things as simple as I can. Encryption is not simple. It can be fun to work with and I don't want to ignore that but unless you want to do that for fun I don't think it's worth the complexity.

[^1]: https://cryptomator.org/

(no affiliation)

Sadly it has no gui (that I know of) but something like runtestic can help with automated incremental and encrypted backups, where you only want to keep the last n versions (and for example a weekly snapshot)

If you use Gnome, I would encourage you to use it and give feedback here [1]

[1] https://gitlab.gnome.org/World/deja-dup/-/issues/192

Yep, always good to have a local copy.

> Note that a cloud provider: >1. can shut off your access at any time, by accident or on purpose, and you have no recourse

Absolutely, which is why steps 1-3 help here.

> 2. will sell your data to anyone at any time

Not true. Neither Google, Amazon or Microsoft do this. The current privacy policies state that. If you want to be extra cautious, you can create a cloud storage account or an account with personal domain and their privacy policies are even strong. Imaging Google or Microsoft "selling" a business's images in Photos or Drive.

> 3. will give it to any government for the asking

Unfortunately, true.

> 4. will leak it to any hacker

Will leak it is a bit too strong and harsh. 'Could be leaked' is better. Although, I have a basic trust in their security as opposed to not even using their services for this fear. Of course, I would never upload any 'private' or 'confidential' images there.

> 5. will allow employees of said cloud provider to browse your data

Again, false. 'Will' here is too strong. None of the employees have direct access to your data -- neither emails, nor images. And if you want to be even more confident, go for Google Suite or Microsoft 365 account and this is guaranteed to not happen.

> 6. will scan it to use it to sell you ads

True, but you could mitigate it by using the paid tier accounts.

> 7. will scan it looking for anything they don't approve of this week, and take "corrective action"

No arguments here -- although the paid tier accounts should guarantee this does not happen.

> 8. if you've got terabytes, good luck trying to download it

The better solution is to keep a local copy using steps 1-3 and also upload them to "your chosen service".

I've seen enough reports of employees charged with viewing customers' private data to be skeptical that this can't be done. Also, people keep finding security flaws. If outsiders can penetrate the system and download terabytes, it ought to be even easier for insiders.

> Will leak it is a bit too strong and harsh

Not at all. I used to design airplane parts. The principle is not "could it fail" but "when it fails, how do we ensure the plane won't crash?" It's a completely different mindset.

> guaranteed to not happen

If you say that in the airplane design business, you'll get your knuckles thwacked with a ruler. It is always when it happens, ...

> current privacy policies

Which can change at any time.

I clearly judge risk differently than you do. My approach comes from 3 years of designing airplane parts. I wish this approach was taught more often.

I know that if my data is not in the cloud, it will not be compromised by the inevitable failure of the cloud to protect it.

Right, but your analogy is akin to saying "I won't travel by any Boeing plane because the plane could crash". Different risks are fine as long as the mitigation plan isn't "I will never travel by planes". Just my 2 cents since I don't design airplane parts.

> If you say that in the airplane design business, you'll get your knuckles thwacked with a ruler. It is always when it happens, ...

Sure, different industries, different risks. The risk in this context isn't death. I agree "guaranteed" was a wrong choice of word here, however I have basic trust in their enterprise products to know my data is safer and secure than doing my own thing.

> Which can change at any time.

Which is when you can decide if you still want to continue -- or purge your data (since I assume there is a backup).

Recent photos live on grandma-/normie-friendly services like Google Photos, and once every year or so I download everything and store it on platter drives locally with par2 archives to guard against bit rot.

Ran a RAW + JPEG error detection script: on the order of 10^1 - 10^2 of photos were no longer valid out of on the order of 10^5 - 10^6 photos.

Turns out rsync verifies on the way, not after write.

2. plug it into my system via a USB drive adapter

3. format the drive

4. run the program I wrote to copy the files I want to back up to it

5. unplug the drive

6. plug the drive in next time I want to do a backup

7. run my program again, which will only copy over changed files. You could use rsync if you prefer

8. unplug the drive

Error modes? I've had disk drives fail now and then. I have a 10 lb sledge hammer I use to render it unreadable by anyone but the NSA, and throw it in the trash. Buy another one.

I suppose I'm not really understanding your question. But it really is this low tech and simple.

I've never heard that, is there any evidence that this has happened?

I have a bunch of external drives. They go back to ~100GB parallel-ATA drives in external USB enclosures. I think I have ~10 generations of disks in play right now. I've spent less than $2,000 on drives in ~20 years. I probably should have purchased a few more disks over the years since I went >18 months between generations and sometimes 24-30 months.

I wrote some simple scripts to copy the contents to a new drive, compare hashes of the copy to the source, and store the hashes for any new files on the new copy. It's mostly automated. The only thing I've thought about adding is perhaps using PAR2 to add some bitrot protection.

I keep the most recent disk at home and copy files to it throughout the year. (I keep it unplugged most of the time.) When I buy a new disk I run my copy scripts then take the old drive to a relative's house and store it in their gun safe. (Lately I've been the oldest drive from the safe and marveling at the old technology. >smile<)

Some of my older drives are no longer recognized by the drivers anymore.

I've had several drives completely fail, and one I dropped that never worked again.

I've used CDROMS, DVD-Rs, zip drives, and even cartridge tape for a while, but the hard drives work the best. I still have the cartridge tapes, but no way to read them anymore. Same with zip disks, can no longer read them.

I've found that the greatest benefit of cloud storage is not as a backup, but as our primary place to keep and look at photos. We started uploading and organizing everything in Flickr a few years ago, and now we spend more time enjoying our photos with Flickr's web and mobile apps than we ever did when they were scattered across our laptops and backed up to an external hard drive.

I still have the photos stored locally on my laptop and an external hard drive, and I fully expect that I'll need to upload them again to another service (maybe Google Photos, if that still exists) when Flickr finally fails as a business.

I have a copy of a handwritten family tree and other notes written by great-grandmothers, cousins, and great aunts, some of which survived close to 100 years. If they had computers back then, I doubt any of the data would be available now. The photo books I have shared with relatives contains copies of these letters and charts and family photos, and even if half of them are destroyed or lost in the next 50 years, the remainder will survive, hopefully for much longer.

It amazes me how much faith other genealogists have in cloud storage or paid services like Ancestry, owned for the past 10+ years by a series of profit-focused private equity firms. Policies change (for PE's benefit), subscriptions lapse, people delete accounts, or online services get shut down, as some Ancestry users discovered the hard way:

https://slate.com/technology/2015/04/myfamily-shuttered-ance...

Another problem with these services: Even though I can share a link to an online tree or a shared cloud folder, more than half of the recipients hate to deal with them because they are forced to learn a new UI or get prompted to death to sign up for a paid service they don't need. The other half eventually forget the links or the information that was in them. Paper or books is the only proven long-term storage medium for genealogy data.

Thanks!

https://www.familysearch.org/wiki/en/GEDCOM

GEDCOM has many drawbacks, but it's the de facto standard, it's been around for nearly 40 years, and it's plaintext.

Core data and pictures and family stories gets put into photo books and widely shared with relatives, including some cousins I barely know. I've done three so far for different branches, and will do some updated ones as well as new research is completed.

This is a good point. I was 100% anti-cloud for my photos, but have to admit the barrier of bringing out the external HDD and hooking it up to a PC means we hardly ever look at them.

My next questions were then how do you arrange a project for the intended audience (context and information is important as random photos of people 100 years from now won't have much value). And then how do you create it so the paper book lasts for 100 + years.

Your idea on Blurb was was interesting to me. Do you know the quality of the paper of the book will survive a long time or did you look into that? Also what reasons did you settle on Blurb (if you don't mind me asking)?

Thanks for pointing me in the right direction - I was about to suggest printing out individual copies and putting them in classic photograph book with plastic covers and adding her own blurbs.

The books we've made through Blurb look high quality to me. They offer several choices for the type of paper at different price points, but I didn't do any research on that. We went with Blurb because it has some relationship with Flickr for pricing and an interface to import our photos from those already organized albums (see "Prints" from the drop-down menu on Flickr). It used to have a decent web app for creating a book. Now you have to download and install their Windows-only desktop app, but it's good.

If you are making a book for your mom, pay for the higher quality options if offered, such as hardback with a dust cover over paperback, and the heavier paper if offered.

The U.S. Library of Congress recommends acid-free paper for archival storage (https://www.loc.gov/preservation/resources/rt/perm/pp_x1.htm...) but AFAIK this option is not available for commercial photo books.

I also use Adobe Lightroom, not just for editing, but for storing photos. It's a great way for me to look at, organize, and even publish-to-web, from either my desktop, laptop, or mobile. As for it being a cloud solution, I'm more comfortable with handing my photos to Adobe's cloud than to Google's (or Facebook's!) Although it contains far far more photos than I plan to put in my Blurb books, it too is a "best of" sort of thing.

My "source of truth" is a local hard drive dedicated to photo storage. When I shoot, my camera is set to RAW+JPEG, and I save both copies. The thinking is: RAW has the most data, but do I know for sure that someone digging through these photos in 100 years will have proper Sony RAW Codec to view them? Probably, just JPEG is a nice fall-back.... I organize photos using the following file structure

  /{celestial_object}/{continent}/{country}/{state_or_province}/{city_or_locale}/YYYY.MM.DD__album_title 
  .. e.g. /earth/north_america/usa/california/los_angeles/2021.01.18__visiting_steve
  .. e.g. /moon/mare_tranquillitatis/2021.01.18__first_shoot_with_new_celestron_lens
  .. e.g. /jupiter/2021.01.18__galilean_moons

I'm looking to get a NAS and set up RAID, but even then I'll still mirror the disk periodically and keep a copy in a safe place.

But I do encourage everyone to consider printing photo books.

Back up the media to that disk, and then take it down to your safe deposit box. No absurd monthly fees, no Terms of Service, it's off-site so if your house burns down, you're covered.

"Your what?" Yeah, the banks still have those. Mine costs $15 a year. If you die, your next-of-kin will have to bring in a death certificate to get access to it. They'll have lots of other hassles, believe me.

Forget being trendy and having it in the cloud. How often are you going to access it? "Never", right?

You do have to go in every few years and replace it with the latest & greatest, and you'll probably have more media to backup anyway.

Lesson: never rely on a single backup for what's important to you.

Disks are not designed to be usable after 5-10 years of sitting on a shelf. Portable drives aren't particularly robust either, I've worked with many people with lost data on a USB attached drive. I'd use at LEAST two drives, ideally in two places if that's your only store for anything important.

I'd recommend using a cloud service (amazon.com/Amazon-Photos/, https://www.flickr.com/, https://photos.google.com/) or similar AND keep a local copy on a disk. Plan on replacing the disk every 5 years and paying whatever monthly/annual fees for enough photo storage. Include in your will full details on the online and offline access to the data. Sure it's a pain, sure there are costs involved, but it's not some insurmountable task. Much like your taxes, will, deeds for property, bank account info, etc. In all cases you want the original documents and pointers to the company/account.

Bitwarden (among others) handles designating a successor for accounts in the case of death, you can include username, password, and any encrypted volume info there.

The fact that it's news shows that it's rare, not common. The whole point of the news is to show things that don't happen every day.

I've never understood how people got to a state where they think "I saw it on the news, so it must happen all the time."

Not as many as used to.

I was at the branch inside the regional headquarters for one of those too-big-to-fail banks a couple of weeks ago adding my wife to an account, and I asked the guy if there was a way to transfer the contents of my safe deposit box at another one of the bank's regional headquarters to this one because it would be more convenient. He said that his bank (one of the five largest in America) doesn't open new safe deposit boxes anymore.

Still, if you can find one, they're invaluable. For reasons beyond the scope of this discussion, I'm glad I still have the one I have, even though it's a thousand miles away.

I wonder if the smaller banks are still doing them. Did credit unions ever have them?

"She sued the bank in Los Angeles Superior Court, seeking $7.3 million. Bank of America sought to have the case dismissed, citing language in its lease agreement stating that the renter “assumes all risks” of leaving property in the box."[0]

[0] https://www.nytimes.com/2019/07/19/business/safe-deposit-box...

Chances are the deposit box are way better.

No, I'm not closing my safe deposit box, but that article made me realize the danger of banks just deciding someday "You know what? Who needs those stupid things any more?"

Of course, the same thing applies to cloud backup services. Google is constantly telling me I should buy more storage.

Eventually, you don't want to rely on a service that's unprofitable for the company providing it. Unless, like an insurer paying off your life insurance policy, it's something that they can't out of, without destroying their reputation. Or bringing the government down on them.

citation?

and nothing ever goes wrong with cloud backup schemes, over a very long time period?

But I see little if any point to offline-only backups of something like family photos. Usable and backed-up is what I'm after, and preferably not just usable but very convenient.

Maybe MS has some solution, they have a 1 GB level.

I did used to use Mylio [1] which is a really lovely solution that uses a sort of internal P2P approach - your images don't go near a cloud host if you don't want them to. It just keeps your "vaults" updated with each other - a vault can be on a NAS / desktop / mobile. The issue I had with it was that the mobile scenario is a bit shonky - required you to have your phone on at the same time as one of your other vaults in order to sync mobile photos, and that proved too fiddly to maintain.

[1] https://mylio.com/

- subscribe to iCloud 2TB plan

- all photos on my wife and my devices are auto uploaded to the cloud

- i run a synology nas that downloads and organizes all the photos in a directory structure YYYY/MM/DD.

- a hyperbackup task backs up the (encrypted) photos to Backblaze B2.

That way i have my photos in apple's cloud, on my nas and in another cloud.

What do I use to download/sync the photos from iCloud to the NAS, you ask?? It's an awesome project called "iCloud Photos Downloader" - https://github.com/icloud-photos-downloader/icloud_photos_do...

I have it running in a docker container on the nas and it periodically syncs and keeps itself up to date, no hassle. https://hub.docker.com/r/boredazfcuk/icloudpd

Unlike other mentions in the comments what you don't want to have is custom scripts, git, backups tools, strong encryption, cronjobs running on some server that will go away eventually. A labeled, unencrypted HDD in a closet in two different locations will most likely do the job just fine. If you want to throw in your cloud backup that you take care of that will be a nice bonus but I wouldn't rely only on a "tech" solution.

Where the HDD approach usually falls apart is merge conflicts. And by that I don’t mean only in the git world. This is an inherently difficult problem.

Someone extracts a subset of the hdd collection onto their laptop to do some editing. Then forget to store back the result on the hdd for the next 6 months. Next time when doing backup, mix up the hard drives, only saving the photos on the second one. And so on. Even basic stuff like deduplicating photos or filtering out only the good photos can leave a lot of confusion. Quick resolutions like “just keep everything” completely reverts such a cleanup (1). A strict master-slave scheme helps but requires access to the master hdd for any writes.

Cloud doesn’t solve this entirely but shortens the window of conflicts by convenient and always in sync access to both read and write.

Someone really need to invent a git for photos, for humans. That can deal with the longer partitions which loose hard drives usually lead to.

(1) on the topic of cleaning up photos. Doing this before backing up is strongly recommended. The smaller your data the easier it is to manage. You also will enjoy your old photos a lot more. Nobody wants to click through the 100 angles you shot of some food from a local cafe 5 years ago. Keep only what is truly relevant.

Posterity is something we've been thinking deeply about. Our infrastructure is designed to support successors, similar to GitHub[2]. But since cloud storage is expensive, successors will have to choose between exporting a local copy[3] or paying for the newly accumulated data.

We are still designing this "feature". If you think we can do better, please let me know. We would be grateful for any feedback. :)

[1]: https://ente.io

[2]: https://docs.github.com/en/account-and-profile/setting-up-an...

[3]: https://ente.io/faq/migration/out-of-ente/

My personal backup is the usual 3-2-1: 3 backups, 2 places, 1 offline. I have one copy on my local harddrive (that I work with), one automatically synced copy via seafile on one of my dedicated servers (which also maintains a few months of history in case I accidentally delete something) and I have one external, offline harddrive at a relatives house, that I sync to every half a year or so. Since I'm paranoid, my dedicated server is backed up to an external storage every night as well via borgbackup. If you don't want to spend a few bucks a month on backblaze or another service, just use a local NAS - as long as you have one harddrive offline and external as well (in case of a ransomware attack that crypts all files).

Important: My files and backups are fully encrypted and it's imperative(!) that you backup all documentation, all config files, all settings, all cronjobs, all executables that have something to do with the backup and restoration process unencrypted with every backup - in the desaster case, nothing sucks more than trying to find the right settings again.

Case in point: I originally used a custom shell script and encoded the files with openssl. However, the default hash scheme was changed between openssl 1.0 and openssl 1.1 (or something like that) and when it came to restoring after a harddrive failure, this took me like a weekend to sort out.

As for posterity: it's up to you if you encrypt the external drive at a relative - if you're fine with a burglar having the images and you cannot be ransomed with them (e.g. due to nudes), just write what is on the harddrive clearly and you're fine.

What is the plan: when decryption fails (and before you identified that it's a versioning issue with openssl, in your case) you'd reinstall an old linux to a random computer and work from there? How many config files and settings are even involved in your backup process and how can you be sure you haven't missed anything?

I hope there are dependency-free solutions for this - a winzip-encrypted .zip file that asks for a password should work everywhere even in the future?

  > if you're fine with a burglar having the images and you cannot be ransomed with them (e.g. due to nudes), just write what is on the harddrive clearly and you're fine.

  > That's security by obscurity

I can thus checkout the project on a new machine and just initiate it (giving it the right api keys etc) without issue.

OP is talking about digitized videos, so asking to re-compress the videos is a https://xkcd.com/1683/ in the making.

Yea you are technically correct, but if the distance from the original is just a handful of encodes, good luck noticing any lower quality that's not simply due to poor encoding settings. And when a proper encode can be a 10th the size with hardly any drop in quality, in a video file you might view <12 more times in your life, does it really matter.

It's raid 0 (2x6tb) with 2 separate backups done nightly/weekly, one in the house and one done in the garage.

The speed of access is unmatched to any cloud system, in the long run it'll be cheaper than online (assuming £10 per month for 6tb of storage).

I made the plunge after trying a few cheaper options plus cloud. I'm happy with the investment and haven't looked back, it's been rock solid. Synology.

The same system could probably be had for less now, the nas box also acts as my docker server which has been running without issues for 2 years. Most importantly it's not a full time job maintaining it, I spend maybe an hr a month if that checking up on it.

Comes with duplication detection and everything you'd need without learning shell can be done using the nice gui. It's just nice having a system anyone can use, no shell scripts in place at all. It just works.

On my phone however, I simply send my photos to Google Photos. Actually, Lightroom exports end up there too. This makes it easy to see and share photos with family and friends.

I would love to hear how you view and share your photos using your NAS. Are you using something like OwnCloud or NextCloud? Or does Synology offer this type of software built-in?

Provided the app is installed you can easily share the photo through WhatsApp or others on your phone. In terms of public access / link I have disabled this however its possible to expose through a cloud service (Synology cloud).

The photos app works well when exporting however I wish it has a backup feature however the flip side is my family wouldn't want all their personal photos sitting on our family nas.

I've since been able to recover them through another external drive casing but I haven't found a replacement idea on what to do.

Exactly, I want my family to be able to get the photos if I suddenly get hit by a bus. Noone is gonna be able to deal with your custom server/cronjob/etc. crap in case anything happens to you.

Over the past decade I have migrated back and forth between cloud storage and a NAS. Both of them need upkeep, either to keep footing the bill or to check the drives and replace them if they die.

I think anyone who is serious about this needs something you can put in a closet/safe/bank vault and a decade later it will still work, and the best bet is optical disks, they were suppose to last hundred years +, but that industry is kind of on it's last legs and we were never sure how accurate the longevity claims are.

The timer runs every Sunday at 5am. It's never failed.

I should probably add it as a secondary destination in hyperbackup however need a nice way to autorun an rsync server on Windows. Then I could combine the alerts in Synology should something fail.

I store all my photos on a desktop computer and use BackBlaze to automatically keep everything backed up.

It violates the 3-2-1 rule but in the 10ish years I've been using this approach I've had tons of harddrive failures, had to restore some backups due to my own mistakes, but I haven't lost a single image.

I used to do complicated sync stuff but I found that the more moving pieces there were the more I would lose photos for various reasons.

Would love to do this: - Back-up local machine via Blackblaze - Periodically delete all local photos, save somewhere in BB - Turn on regular back-up again - keeping both the local stuff (now much smaller) but also the previous bulk save in BB.

There's a way to do it that I find somewhat risky: pay a bit extra for permanent storage. Normally Backblaze will remove deleted files in 30 days (I pay for a year). If you do permanent then you can delete your originals from your local machine and Backblaze will keep them presumably forever. That approach makes me nervous personally :)

Getting to the photos was horrible, so made an alias of the folder that contains the files (which is in the library bundle).

It’s all on a VM so it sort of out the way.

My reasoning is that I don't trust Google to not lock me out of my account at some point, so having both a local and a remote backup gives me piece of mind. I periodically check the offsite backup to check that it's still all working. Total cost for about a terabyte of files (it's not only photos and videos) is about $6/month, which is pretty reasonable.

[0] https://github.com/gilesknap/gphotos-sync

[1] https://restic.net/

[2] https://www.backblaze.com/b2/cloud-storage.html

Can it show me emails, photos, tweets, etc of the related meetings/events?

Can I take a photo or upload a photo of someone and provide as many details as possible? In relation to me personally and them alone as a person?

Can it be a celebrity or a person from the past (no longer living) and it could try to see a connection, related things, etc?

Would this also be helpful for parties, weddings, funerals, and more?

Could I ask it to try to find a relation between two people?

What if you connected it to something like Lexis Nexis? Be able to display connections in different chart types? bubble, connected graph, or graph commons: https://neo4j.com/blog/creative-network-science-graph-common...

Can a timeline or photos help to tell a story? Act as a guide or reference? Add more clarity to situations/events? Could we offer this as an addon to weddings and other events? Can we create a linear timeline of photos, dates, times and locations? Could we also show related influences? Like songs, albums, family history or tree. Topics to disable: lyrics, alcohol, ?? How to access/merge other apple photo accounts? AI photo collage, web based, download as video or PPT

At the moment we're focused on how we can tie together shared events and memories of your personal story (hence the name) into one place. Currently this is primarily from photos, but as you've suggested there is a lot one can do with how those photos are connected to time, location, each other, and other possible connections. This is what I see missing in most photo backup tools, since they're primarily archival and aren't designed to tie ephemera together into a permanent catalog of your experiences.

It is just ZFS snapshots with replication. I wrote my own shell scripts for this that takes daily/hourly snapshots. This scales well with multiple TB's, bandwidth is no issue as only changes from last snapshot(s) are replicated.

After 10 years, I've replaced only 2 hard drives of about 30. I do upgrade FreeBSD once every second or third year. Nothing else, it just keeps going.

This project has probably costed me max a week in configuration & setup & maintenance over the 10 years.

I have 3 separate use cases: casual access to shared memories for which I'm the family steward (photos, videos), motivated access to personal important documents, and motivated access to other personal files.

I'm not about to leave an unencrypted drive full of personal info in a family member's house, as the risk of theft is strictly greater than just having one copy. My current plan is to use EXT4 + Luks to satisfy the last 2 scenarios, which I think stands a reasonable chance for anyone slightly techy (most modern Linux distributions will simply prompt you for the container password when you try to access the encrypted drive) and is likely to enjoy long-lived support for at least a decade or 2. I have a techy person in my family, not sure I'd do this if I didn't... For casual access to shared memories, I plan on leaving an unencrypted partition. While it lasts, these media are also available on a family-only photo album I put up on AWS.

I'm considering using a laptop as the vessel for the encrypted drive, with a suitable Linux distro pre-loaded and instructions on the desktop/printed out and kept in an adhesive document pocket stuck to the machine.

For pics that don't fit neatly into an album theme I'll group them by date, "Family pics 1995-1996" based on their quantity.

Then I'll print photobooks, adding info where possible about location, date, and people. Basically, recreating physical photo albums. I'll print 2-5 copies of each album, depending if I expect to send copies to other family members or not.

Having physically-printed photos works both as a last-ditch backup and as easy access for anyone to browse. I nostalgically remember spending time as a kid flipping through my parent's and grandparent's photo albums - sometimes with them, but many times on my own. I want to ensure my own kids can have the same experience.

I have a server at home with a 4tb drive shared out with Samba and of course sftp (xfs). There is a 2nd internal drive formatted with btrfs. Once a day the btrfs drive is mounted, drive 1 rsyncs to drive 2, and drive 2 creates a snapshot, then unmounts. Later that night, I run an borg backup and rsync (depending on sensitivity of data) to a time4vps (https://www.time4vps.com/?affid=1881&_ga=2.199835630.1135648...) 1TB storage server. Currently, that's enough for the stuff I want to backup.

Online: A big BTRFS "raid 1" (block-based mirroring) array. This is what Plex runs off of. If I lose a drive, I just replace it and rebalance. I run btrfs scrubs weekly.

Offline: I have a WD external hdd with a plain old ext4 hdd that I rsync all files to weekly. If my btrfs array shits itself, I just copy back over from this hdd.

Emergency: If the above process doesn't work (say: the external hdd is dead when I go to restore from it, or I can't physically find it, or my house floods and everything in the basement is lost, etc), then my fallback is:

I sync all truly critical stuff that can't be replaced (so, basically just family pics and videos) to S3 as deep archive storage class. Costs pennies. I don't do any fancy tar/archiving, encrypting, or compressing (since video and image files are generally already compressed). I figure the risk of people snooping at our family pics is smaller than the risk of me needing to restore from a backup and not being able to because of stupid reasons (lost gpg key, some kind of problem with tar, etc).

This is small, since most of my disk use is from things like rips I've made of BluRay/DVD's, FLACs from bandcamp or rips of music CD's, etc. Things I can recover with appropriate effort.

I have a raspberry pi with a USB drive attached. The OS is ubuntu, so I can use ZFS. The drive is backed up to backblaze S3 using restic in a cron job. The backup and ZFS health is monitored with healthchecks.io

The data sync across all the devices including the raspberry pi is done using resilio sync.

I view my photos when I am home using an app written by myself: photograf (https://github.com/ptek/photograf) It runs on the same raspberry.

I know there's some holes in my plan (ext4 isn't foolproof, etc.) but it's been an ongoing process.

You have got me thinking though about my backup of family photos/videos. My backups are all too clever, encrypted and inaccessible at the moment. I feel inspired now to make sure photos/videos are easy for my family and future generations to access.

The ability to search photos easily over time and the weekly "this week X years ago" reels are great value.

My family are all taking photos on their own devices during a vacation, holiday, gathering, etc. and then someone creates a shared Album on Photos (iOS) where everyone just dumps all the photos taken. It's a complete mess in my opinion. WE have a shared album "Christmas 2018", "Holidays 2019", "Xmas'20", etc. depending on who named it.

Everyone else seems to like the ease of upload and ability to find past pictures. Getting the photos out of the apple app seems like a pain to me. I feel like there has to be a better way, but it has to work for the lowest common denominator (Grandparents, etc).

We have a couple people on Android and are essentially cut off from the family photo stream. When that happens everyone's mentality is, "get an iPhone".

Works across all grandparents and I used one of their auto upload a face (I'm sure that's the technical term for it) to automatically send a photos of the grandkids to the grandparents. This makes the whole system worthwhile, I get comments from them constantly about how much they love the auto photo albums.

If we have pics from someone else, I usually create an album and ask them to upload their pictures there. Not everyone does it - some people just share over SMS and then it's likely those pictures will get lost.

It's cheap, easy and ransomware proof.

This means obviously importing all photos taken via my DSLR into Lightroom, but also syncing all photos taken on our iPhones via the Lightroom Mobile App.

Lightroom Classic keeps all the (compressed) photos in the Adobe cloud for easy sharing and browsing, but also writes them out (unaltered) to a directory on my Local NAS.

This NAS gets automatically backed up via Arq Backup (https://www.arqbackup.com) to an encrypted Amazon S3 bucket. Additionally, once or twice a year I create a versioned copy of the NAS via Carbon Copy Cloner (https://bombich.com/) to an external hard drive. This hard drive is stored offsite somewhere safe.

In a nutshell, for around $12 a month + a NAS + a hard drive, we have all the convenience of the Adobe Lightroom cloud combined with a local copy on the NAS, a cloud copy on S3 (in case Lightroom cloud gets corrupted) and an offsite copy (in case our place and the whole internet burns down :-)).

In short, I have an old desktop-turned-NAS with ZFS that will tolerate 2 physical drive failures and access my photos over a network share. It runs a weekly cron script to copy an iterative, encrypted borg backup of the photos offsite over sshfs to my VPS. I can mount specific date's snapshots over sshfs back to the nas/another machine if I ever need to recover anything without having to enter the password on the VPS.

It's a good question on what would happen if I died. Thanks for getting me to consider this. My spouse wouldn't have a clue with the VPS or how to fix the NAS but could pay someone to easily copy the data off. I should leave a note taped inside the case. For now I'll plug an external ntfs drive into the NAS and if needed she could plug that into any other device.

Life is already complicated as is.

[0]: https://github.com/mholt/Timeliner

[1]: https://twitter.com/timelinize

You’ll obviously need to leave keys, instructions etc for your next of kin. Or put a hard drive in a safe deposit box.

Then I have a custom script that encrypts and replicates all the borgbackup repositories on Google Drive (where I have unlimited space because my previous university still hasn't deleted my account).

Ideally I would also replicate everything on a drive at some relative's house, but I haven't had time to set up this yet.

Overall I am rather pleased with this setup. This is for general backup of everything. Specifically for photos, I have a script on both my and my wife's Android phone that copies all photos and videos to a shared directory on the MacBook, and she periodically organizes all the photos there. Then they fall in the main backup.

If I were you, I'd delete 90% or more of those photos and videos to save running costs. I don't know how enthusiastic you are about watching photos of yourself when you were shitting your pants as a toddler, but, speaking for myself, I'd consider being urged to watch even a small fraction of what you have at family reunions some form of torture....

In addition to the other good answers, a box full a printed photos will stand the test of time.

While I can print photos, it's typically far easier and cheaper to have Walgreens do it.

2. Encrypted external offline USB drive onsite

3. Encrypted external offline USB drive offsite

4. Upload of an encrypted archive to the cloud storage

It's all manual so sometimes some backups are months old. Haven't lost a single digital photo for almost 15 years now, and have gigabytes of obsolete photos...

> It's all manual so sometimes some backups are months old.

Do you at least have a script for backing up the live copy to onsite? Like plug in the drive, enter encryption key, then run script?

Then I share my Amazon Prime backup with my wife. All our photos are backed up there but not videos.

For videos, I upload edited videos on YouTube as unlisted or private. And some of better videos as public on Youtube, Facebook, and Instagram.

And finally I try to share as many photos and videos in emails, Whatsapp, with friends and family. Sort of distributed back up in the worst case scenario.

(Also never delete any photo or video unless you really don't want anyone to see it ever. I used to extensively prune my library but wish I had not. Storage will likely keep getting cheaper. And AI will get better and it will be able to create better slideshows and edited videos.)

When talking percentages, sure, Glacier is $4/TB and B2 is $5/TB, so Glacier is 20% cheaper. But if you have to actually retrieve your data, it will likely cost more than any savings you've accrued. If you take a look at the S3 pricing page and your head isn't swimming afterwards, you probably don't understand the complexity of it. For example, they have data "retrieval" fees that are sometimes per request, sometimes per GB, and sometimes free (they promote that of course). But there's still that pesky 9 cents/GB data transfer fee that's on top of the per-GB retrieval fee, putting you at 10 cents/GB to retrieve data. Glacier also needs special tools: S3 tools can't access data stored in Glacier.

If you really want to stick with Amazon, I'd suggest S3 One-Zone Infrequent Access. It's a little more expensive, but you can use standard S3 tools to get your data back. You'll still have that 9 cents/GB retrieval cost, vs 1 cent/GB for B2.

I don’t want to download the old photos that are just on iCloud to my device, but there seems to be this python script which I will try to use tonight: https://github.com/icloud-photos-downloader/icloud_photos_do.... I assume that you could combine this script easily in a cron job with RClone which is rsync for the cloud: https://rclone.org/

So I gave up. Unfortunately I still have to do manual export all media from i$$$ devices, move them to linux/android-based system. "move" really means that, because I never pay icloud again.

I have no prior experience of such stuff and managed it (off to a Synology and from there to Backblaze).

Getting stuff out of iCloud in an automated and regular basis is surprisingly hard.

We use Apple Photos as the primary, of which I export a backup once a while (at-least yearly for sure). Because Apple has’t solved what everyone ask — a way for two or more users to merge Photos, I use Dropbox[1] on the other devices (wife, daughters) to collect into “Camera Uploads”.

I split the content into “Screenshots” (PNG), Videos, and Photos with Hazel[2]. Photos to imported to the Primary Photos Library, Screenshots, and Videos are either deleted or sorted weekly (few minutes), monthly (30-min-ish), quarterly(30-min-ish), and yearly (about 3-5 hours on a day around the end of the year, with the other digital clean-up task) during my digital chores that I do. My daughters produces a lot of them that I need to throw out. She can position her iPad and video-shoot while she talks her way cleaning her desk for an hour.

The photos and videos are now in my primary device iMac (for now), with a second backup on a 2012 MacMini running many other digital errands. The 2TB iCloud+[3] is good enough for now.

I would definitely love to clean this steps and move to a more non-proprietary format (read non-Apple) to Open Source tools/process.

I tried Google Photos because it was easier to combine and I even had one of their large plan but it turned out to be one of the messiest thing I tried. Stopped using them[4].

I will be reading and learning from the other interesting comments in this thread. Thanks to all.

1. https://www.dropbox.com

2. https://www.noodlesoft.com

3. https://www.icloud.com

4. https://brajeshwar.com/2021/how-to-delete-all-photos-and-get...

The price of existing cloud storage is too high, and some of the companies(Shoebox, Canon Irista) doing the business gradually shutdown the services, this is a money losing business, it’s not the efficient way to manage huge amount of assets centralized (flicker CEO’s open letter sent last year confirmed this), they have to either make it more expensive, or make you the product. Cloud service is convenient for the user, people don’t need to buy expensive hardware, don’t need to be the professionals to maintain that, don’t need to worry about the energy fee to keep it run 24x7, but things are changing, single board computers are getting cheaper, more powerful and more energy efficient, storage are getting cheaper with larger capacity, software are getting more intelligent, people are having more and more concerns about the privacy, it’s now viable to host the Photo service, your private cloud, at your own place.

There are tons of open source alternatives there, but non of them provide competitive features like google photo, even though there are some backend services like photoprism that does a very good job on indexing the photo with AI, however the mobile APP is missing, and google also has the ecosystem to show photo/video on google home and chromecast.

We've been building Lomorage (https://lomorage.com) trying to fill the gap, it's easy to setup a self hosted service(Still need a lot of work to make it easy for non tech users), cross platform, has iOS and Android mobile APP, can setup multiple accounts, some basic AI search(not polished and need more work), there are other features missing, but it's stable enough for daily use. Would appreciate you try it out and give some feedback.

1. a cronjob on my android phone (via termux) does an rsync to a VPS

2. the VPS has a cronjob to sync everything with S3

3. the frontoffice tool for people to access those photos is my open source Dropbox like frontend that is bring your own backend: https://github.com/mickael-kerjean/filestash My wife and I got an account and family members can access it through shared links.

The S3 bill goes to a shared account so that If I die, the VPS will probably be quickly removed but S3 should stay in there with my wife paying for it.

I keep all my photos in the "native" format, which is basically JPEG and whatever my iPhone is spitting out these days. They are currently editable (which for me is mainly crops, with occasional other fix ups). I scanned all my old negatives, APS, and slides and just had them all issued as the lowest compression (highest resolution) JPEG (I didn't have a huge choice of formats).

But I have a bunch of videos from film tape, and cards. What's the best equivalent (supporting simple crops and edits) that might reasonably be editable in future?

If I were being properly rigorous I'd be burning three copies and putting two of them in different physical places.

Printing out the occasional photo book of our favorites is also a part of the strategy - if our grandkids want to see photos of our lives after we're dead, they won't need to worry about lapsed hosting bills or ancient storage technology. They can just look at the book.

Obligatory reference to my rant about data preservation:

http://howicode.nateeag.com/data-preservation.html

IIRC, when I settled on the 25s, they were a slightly better price/GB than the 100s. Maybe that's no longer true (or possibly never was).

But printed images are the ultimate. They require no extraction whatsoever.

https://videohubapp.com/en/ - MIT open source: https://github.com/whyboris/Video-Hub-App

Non tech solution: I just print everything. I refill canon cartridges, making the cost of ink negligible and buy cheapish 4x6 photo paper, and I just print like crazy. After some basic calibration, the prints are quite good enough. I guess longevity of the aftermarket inks is still to be seen. Wife enjoys just picking up some prints to send friends and family. She snaps a pic of whatever she sent, and I reprint those.

The chance of all of these versions becoming unavailable is low enough for me not to worry. If I end up under a tree some day nothing changes in the short term, the data will be available to those who have not yet met their tree. In the longer term I'd expect someone to either pick up where I left off or move the data to some storage which befits their level of expertise and needs. I don't want to force them into any contract with any provider, nor can I decide what they deem to be worth keeping.

https://github.com/prettydiff/share-file-systems/tree/master...

I'm glad my backup and storage setup has stood the test of (thus far a relatively short) time.

Most videos are from mobiles, and I use syncthing on each device to automatically back them up to NAS. The NAS has an external USB drive to which back ups are directed, plus another NAS in the shed as a secondary on-site backup - which also has an external USB HDD that gets backed up to. All of the backups other than mobile device to NAS via Syncthing are done with crontab'd rsync scripts.

I then irregularly rotate portable USB-caddied HDDs from my parents and in-laws houses and backup to them as well. This rotation is well overdue, thanks for the reminder.

I don't do cloud. It feels too ephemeral in the context of my history with online services - I change my mind and switch around a bit, and lose interest here and there. I'm the kind of person who would host their own email server if I wasn't already gigabytes deep in a grandfathered free custom domain plan from Google. (I do actually host my own email server using Citadel, but it's a domain that's very rarely used and entirely non-critical).

I have a primary SSD for my OS and applications, a secondary SSD where I store my latest/active Lightroom catalogues/files and a large 4TB (spinning disc) hard drive for local backup.

After a shoot I'll import the files onto my secondary SSD and then manually copy them over to the local HDD once I've done my picks and culled things a bit. That HDD is automatically backed up to Backblaze so there's always an original copy of the files either local or in the cloud. Any images that I edit/pick/export will be manually uploaded to Google Photos (usually in it's own album) and I try to annually pull down files from Google to go into another 'jpg/exported' folder on my HDD that's backed up to Backblaze. My wife and I also have our iPhones backing up to Google photos automatically plus we have Apple storage/backups - so we tend to have double coverage on most things.

Some of it is a bit overkill probably but I'm particular about having access to all my photos and I've definitely gone back to look at and edit old photos/shoots so it's worked out ok for me so far.

I sometimes wonder how much of the way things are is because that's just how the US legal system is, and how much of it is because that's the worldview of the childless twentysomething singles who built these systems. (No judgment: I know when I was a twentysomething childless single building things, I had a person-centric world view, not a family-centric one.)

(An extreme example of this problem is the oculus quest, a device that is begging to be shared between family members, but it requires a Facebook account to use. Further, it is a breach of the terms to have a family facebook account. Nobody in our family with a facebook account - i.e. the older people - wanted to sacrifice theirs to the quest.)

It's weird/sad that a family doesn't have any sort of legal personhood, but a corporation does. A joke solution here might be to incorporate my family and have all these digital assets belong to the family corp. I'm not even sure if I'm joking.

For the cloud backup, I think that you will be saving money by using backblaze [1].

As for the solution for posterity, if you're worried about the possibility of next generations not being able to pay, the only solution I see from the top of my head is to print the photos. Given the amount that you have this would be a daunting task, so only print the ones that you find the most valuable and look for a "safe and stable" (whatever this means... :D) place where to store them.

[1] https://www.backblaze.com/b2/cloud-storage.html

The cameras have Wifi cards and sync to the desktop that syncs to Backblaze.

On new years day each year, the desktop gets archived to a external HDD and put in a safe.

That partition is backed up daily to two separate external USB hard drives using rsync.

That gives three separate copies. One of those copies is also rsynced to a separate laptop, so a total of 4 copies.

I don't store archival stuff like videos, photos, music in my personal directories and scatter them all over the place. An archive is an archive and needs to be centralised and then distributed outwards from there.

Is there a good solution for posterity? For example, once I die

Unfortunately, you can't control posterity from beyond the grave. What you consider priceless is something that your descendants maybe can't wait to get rid of. I learned this lesson when my brother died. I started off trying to keep as much as possible of his stuff, but as I sorted through it I found that it was worthless to me so I ended up discarding about 95% of it. I saw that when I died, somebody else would do exactly the same thing with all of my carefully-hoarded information.

- But what about searching... That moment when you're looking for a specific moment or latest photos of your dog? or pool to find a specific pool.

- Dupe detection.

Those unique features that for example, Dropbox is an inferior tool for the job.

I put photos and videos in separate folders, and then subfolder for year, subfolder for month. Within the per-month folders I rename those photos / videos from their defaults to something that describes their content, eg:

"Daughter cracking the shits in egg and spoon race"

Or

"Daughter getting chased by duck".

The ones we want to re-live.

This takes long time reviewing each and every photo and video though. I'm only up to 2014...

https://photoprism.app/

Then there's Dropbox. It automatically uploads all my raw data, and syncs it across multiple devices. So at least I have local and cloud backups of all my raw photos/videos. A picture taken on my phone camera will immediately appear on my PC desktop hard drive, as long as it's on. The downside is searching through photos can be a slog compared to GPhotos, but that's why I use both.

I used to use Picasa many years ago, but haven't reinstalled it on a newer computer. I've been meaning to run it through my Dropbox collection again to make it easier to sort through locally. IMO, it's still the best photo management software. As far as I know, it's still the only one that does facial recognition locally and without a subscription.

https://picasa.google.com/

They have a decent enough Photos app for browsing / sharing photos.

For offsite backup, I send to B2 which is ridiculously cheap, but I pay for, so I’m not the product. I haven’t put Cloudflare in front of it, but that’s something extra folks do.

Recent Local + Recent in Cloud(DropBox) --> Copy on Home NAS (Synology) + Local NAS weekly Backup (External USB) --> Continuous Cloud Backup (BackBlaze) --> Offsite Copy Annually

Copy all important and historical data to a cloud storage solutions. Currently using DropBox

Export photos to home NAS (Dual Drive) and NAS Backup to external drive weekly. Daily backup all local drives to BackBlaze

Once a year, copy all photos and important data to a larger encrypted USB HDD for one year storage off site at parents home. I swap the drives about once a year. This is my I need it quick recovery copy should the house burn down.

For encrypted storage, I use only local OS encryption. Got burned by TruCrypt/VeraCrypt not supporting new OS.

My wife is the one who takes most of the photos in our family. For her computer (MacOS):

1) Back up her computer to USB hard drive via Time Machine every day or two.

2) Back up her computer to a USB hard drive in her office (via Time Machine) when she takes it in, about once per week.

3) When her drive starts getting full, move photos to a larger USB media drive attached to my computer. (Every year or so.)

For my computer (MacOS), including USB media drive:

1) Back up to a "main" and "redundant" USB backup drives continuously (every hour or so) using Time Machine.

2) Back up Mac hard drive, but not media drive, to "main" USB backup drive every night using SuperDuper.

3) Back up to the cloud continuously using BackBlaze.

4) Swap "main" USB backup drive with an identical drive stored in a safe deposit box (every 3 months).

This provides a nice combination of convenience, redundancy, and recent off-site backup.

1. Pictures are either manually copied to a staging folder on my storage server (for DSLRs) or automatically synced there via Syncthing (for smartphone cameras). For the latter, Syncthing is set up to preserve deleted files in case of accidental deletion.

2. The storage server runs zfs and takes 10-minutely snapshots of all datasets, which are replicated to 2 servers using zrepl: one sitting upstairs and another I rent from Hetzner halfway across the world. Replicated snapshots are kept for 2 years. Everything uses zfs' native encryption, but the replication targets do not have the key.

3. For the really important pictures and other documents, I create 2 additional backups: an encrypted backup to Backblaze B2 via rclone and burning them to M-DISC blu-rays. This is sort of a last ditch thing in case a zfs bug renders the primary backups unreadable.

From then on, I backup to Google Archive Cloud Storage over restic [2]

[1] https://olimex.wordpress.com/2020/03/13/bay-hdd-sdd-is-easy-...

[2] https://cloud.google.com/storage/docs/storage-classes

[1]: https://www.microcenter.com/product/634744/wd-4tb-red-plus-5...

Our local "copy" is basically our personal laptops. This has required manually upgrading our SSDs over the years though because we have around 1TB of photos and videos.

Our off-site copy is backblaze. We don't keep a non-cloud off-site copy.

After that I decided to migrate as much of my hardware as possible to 19" racks. So next, I upgraded to a Synology NAS RS819 (4-bay, 4x4 HDD = 16 TB raw capacity, 12 TB in a RAID5 setup). This lasts for quite some time, as I'm not much into movies. Importantly, the NAS is normally physically _not_ connected to any networks except when I transfer a new batch of photos for archiving.

Some historic material is on a DAT tape in a fireproof safe but that is close to its expected end of life, which is fine as there are HDD copies.

To view the photos, I access a subset of commonly viewed media on another 19" server, so the NAS is truly for archival use only.

What would be nice, but is not yet implemented is a second site to add redundancy and protection from disasters (of course a backup should be offsite per definition). The two sites could synchronize over rsync. But I am still too concerned of the attack surface when putting archival machines on a network. A solution could be the following: some friends recommend Google's "cold storage" as a tertiary cloud storage, which is affordable, and they use scripts to encrypt their data; I think that is a good idea.

The question regarding mechanisms to manage things in case of one's death is an excellent point raised by the poster - I bet a lot of valuable information assets have already been lost due to people passing away when their relatives are not IT savvy to rescue them (or the deceased didn't take precautions to leave passwords and instructions behind).

For an offsite backup, I should start printing 2 and storing one copy at a relative’s house…

Nightly update to Backblaze B2 using Synology Sync.

Considering a third local copy stored at a family / friends and updated yearly.

This requires almost no work on my part... just the occasional copy over of pics from phone / camera to my network drive.

I use SFTP to upload new pictures from phones etc. to the pi. There are many apps that can do SFTP, like e.g. Foldersync on Android.

I have a shell script on the pi that runs once/day and uploads new files to S3 storage, after encrypting them with gpg.

I occasionally get a new USB stick and retire the old one, keeping it as a backup copy.

The whole thing actually works quite well and means:

- Local, unencrypted files I can access easily

- Encrypted files in the cloud, that no external parties can read

- Physical backups that I can put in a safe, or whatever

It’s here: https://github.com/ragnarlonn/savethepictures

backup server galvanically disconnected, manually plugged in to synchronize, then scrub and shutdown, small window for catastrophe, chances of both servers getting killed at the same time is so small I'd probably be dead too.

main and backup servers are in two different locations (different power substations).

I just use rsync for doing the backup.

Family is connected to main fileserver. It snapshots once per week, so if they destroy something, I can get it back, it keeps two years of snapshots.

Once every 5 years I buy a new fileserver, the old one becomes the backup (assuming it is still healthy and not experienced any problems during it's service).

I don‘t want to solve the problem by

A) paying even more money every month, much less even more than for Google Drive & Photos storage itself!

B) uploading my personal photos to a completely untrusted provider (at least I can trust Google to some degree to not leak my photos to any stranger on the internet).

The first is that we know it’s more expensive than Google Photos storage. That’s in part so that we can actually fund development. As a small operation we can’t run it as a loss-leader or make up revenue from other channels like Google could.

The second is that we’ve been somewhat vetted through Google’s application process. We’ve also been building things for the web for 20+ years. Admittedly, trust is hard to earn from folks you don’t know, but we are decent people.

To be honest, Pixaver probably isn’t right for most folks on HN as most here are pretty technical. (I’m just a daily reader here, and spotted this thread, so I thought I’d share it.)

We’ve geared this more for people like my mom who don’t want to set up a local hard drive or do anything even remotely technical. For someone like her, it’s a couple of clicks to having a second backup without needing to do anything additional. :-)

I’ll mention that you’d be interested in more options around backups to @shelkie. Like I said, I don’t think we can make this happen now, but it’s good to have a sense for what others might find useful.

Just curious -- where are photos stored for Pixaver? Are you using cloud storage?

We’re using Wasabi for cloud storage. So far it has worked pretty well for this project.

TrueNas is configured to automatically backup to Backblaze B2 (which is off-site).

After editing, all the exported JPGs are stored on Google Photos and shared with family. I like Google photos because of the content search and face detection features.

So that's 1 on-site medium and 2 off-site mediums. I used to burn the photos to archival disks and place them in the bank, but it got tedious.

Edit: I use Backblaze because it's the cheapest. Most Photo storage providers don't have support for Raw files. (If anyone has a recommendation let me know).

2. Periodically I take photos and videos off our phones and store them on my desktop PC, imported into Lightroom. They are stored on 5400rpm HDDs in mirrored RAID config

3. I get terabytes of OneDrive space with my Microsoft package (I can't remember the name of it right now!), and it automatically uploads all the media from my desktop PC

4. I also have Seafile installed on the desktop, and it automatically uploads everything to my Seafile server, which uses Azure Blob Storage

I'm not close enough to the end to worry about posterity. If I die tomorrow then my keys are lost and all of my data is entropy. I expect to be motivated to have my data pass on in legacy but that will only happen if I have a kid who is willing to accept maintenance of familial digital archives. We're the first generation.