Post by @polotek@social.polotek.net

@polotek@social.polotek.net

2025-12-04 23:10 UTC

I use s3 for files. This is again something I end up introducing way earlier than most people. For basically the same reason as databases. I just don't wanna worry about losing anything. And if I make it Amazon's problem, that's one less thing for me to worry about. I also tend to see these things as keeping myself sharp on what it takes to make things "production ready". Every single time I do s3 I have to look up the right way to secure a new s3 bucket without just making it public.

Replies (3)

@polotek@social.polotek.net 2025-12-04 23:13

Anyway, this is where I'm currently at in my "why is it so hard to launch real personal projects" journey. I'm trying to figure out how we can see more "small software". Apps that are useful or fun but will never be a huge business that makes people money. The developer graveyard of abandoned side projects is littered with this kinda stuff.

Open ##2300287
@aredridel@kolektiva.social 2025-12-04 23:18

@polotek@social.polotek.net This. This whole thing is driving me crazy with self-hosting. I don't even mind backing up files unless I’m on a PaaS, and database would be fine if it was just one but no it's a zillion managed a dozen ways. It's such a barrier now.

Open ##2300296
@Xoriff@hachyderm.io 2025-12-05 00:22

@polotek@social.polotek.net I wonder if there's any way to hack something together throwing sqlite3 files around via s3. Maybe a build process where the last action of a version of your app is to upload its DB file to s3 and the first action on boot is to download it? Don't know if there's a way to do blue-green deploys like that. But for little side projects... :neocat_thinking: _edit:_ oh! Could also make for the hackiest/easiest backups ever. `setTimeout(() => {yeetData()}, 10*60*1000)`

Open ##2300298