Comment by dathinab
3 days ago
Ironically drop box and one drive folders I can still somewhat understand as they are "backuped" in other ways (but potentially not reliable so I also understand why people do not like that).
But .git? It does not mean you have it synced to GitHub or anything reliable?
If you do anything then only backup the .git folder and not the checkout.
But backing up the checkout and not the .git folder is crazy.
I use backblaze and have repos I dont push for this reason so I am a bit stunned lol
Microsoft makes no guarantees on onedrive, you are responsible for backing up that data. Of course they try hard to keep it safe, but contractually they give no promises
> drop box and one drive folders I can still somewhat understand as they are "backuped" in other ways
No they are not. This is explicitly addressed in the article itself.
Parent is using "backuped" to mean "likely in some cloud (latest version)". And that may explain why BB excludes .git folders.
You are using it to mean "maintaining full version history", I believe? Another important consideration.
> You are using it to mean "maintaining full version history", I believe?
No, they are using it to mean “backed up”. Like, “if this data gets deleted or is in any way lost locally, it’s still backed remotely (even years later, when finally needed)”.
I’m astonished so many people here don’t know what a backup is! No wonder it’s easy for Backblaze to play them for fools.
4 replies →
The author of article explicitly ignored both of them come with versioning so it is not just sync, you have old version of files too
normally this folder are synced to dropbox and/or onedrive
both services have internal backups to reduce the chance they lose data
both services allow some limited form of "going back to older version" (like the article states itself).
Just because the article says "sync is not backup" doesn't mean that is true, I mean it literally is backup by definition as it: makes a copy in another location and even has versioning.
It's just not _good enough_ backup for their standards. Maybe even standards of most people on HN, but out there many people are happy with way worse backups, especially wrt. versioning for a lot of (mostly static) media the only reason you need version rollback is in case of a corrupted version being backed up. And a lot of people mostly backup personal photos/videos and important documents, all static by nature.
Through
1. it doesn't really fulfill the 3-2-1 rules it's only 2-1-1 places (local, one backup on ms/drop box cloud, one offsite). Before when it was also backed up to backblaze it was 3-2-1 (kinda). So them silently stopping still is a huge issue.
2. newer versions of the 3-2-1 rule also say treat 2 not just as 2 backups, but also 2 "vendors/access accounts" with the one-drive folder pretty much being onedrive controlled this is 1 vendor across local and all backups. Which is risky.
Oftentimes the important data that needs restoring is in the checkout: uncommitted and unstaged changes that represent hours of work.
I have multiple drives that started out as their own os. Each of them has a Dropbox folder in the standard location. Each of them has a different set of files in them (I deduped at one point), with some overlap of different versions. I no longer use Dropbox, so none of these are synced anywhere.
They don't need to be in my case, I'm only using them now because of existing shortcuts and VM shares and programs configured to source information from them. That doesn't mean I don't want them backed up.
Same for OneDrive: Microsoft configured my account for OneDrive when I set it up. Then I immediately uninstalled it (I don't want it). But I didn't notice that my desktop and documents folders live there. I hate it. But by the time I noticed it, it was already being used as a location for multiple programs that would need to be reconfigured, and it was easier to get used to it than to fix it. Several things I've forgotten about would likely break in ways I wouldn't notice for weeks/months. Multiple self-hosted servers for connecting to my android devices would need to reindex (Plex, voidtools everything, several remote systems that mount via sftp and connected programs would decide all my files were brand new and had never been seen before)