How do you organize your files, Fediverse?
I have accumulated lots of stuff over the years: photos and videos, school and university documents, job files, private development projects, video projects, music, screenshots — all of which I want to keep backed up.
The problem I have is the directory structure. How would you organize all this so that stuff is always easy to find and duplicates won’t spawn?
I’ve tried Johnny.Decimal, and it didn’t work for me. Do you have other ideas? #AskFedi
@kytta I think I've taken to naming files as descriptive as I can and then just using a fuzzy finder...
Files are a mess.
@inhji it had too little categories per area and too many projects per category for me to manage stuff effectively. And, even though Johnny.Decimal promised to make my structure flatter, it only made more nested directories than I've already had. Other than that, I am way faster in navigating alphabetically sorted names rather than numbers. Just not my cup of tea 🤷♂️
@kytta I have 6 folders called "backup_of_old_pc", some of them are subfolders of a folder with the same name. I also have 10 folders called "photos_from_phone" and more than 10 folders called "Robbie". 😎
@kytta I just have a strange memory of what is where, even though my structure would make no sense to anyone else on my pc :p
@kytta I don't really have any special system other than just folders. Documents go in documents. I have a folder for Records which are import documents that never change and are for just reference (bill payments, lease agreements etc).
Other than that, just folders. I'd say the main bit that I could improve on would be photos as there isn't really any organization or system.
I ended up making dump folders where i put everything i am too lazy to sprt throught. I deal with duplicated imaged by using digikam and for text files i have used meld a bit
@kytta I think Perkeep's way is the way. Files aren't hierarchized, they are categorized.
I want a filesystem that stores files by their relationship to places, time, and people. Like a database.
@kytta may not work for everyone, but I burn DVDs every so often with important stuff I want to preserve. Then, everything from that year-or-two(ish) goes into the disc and I know where to find it.
Or go frantically hunting through 3-4 discs till it turns up.
I do tend to organise things, in an obvious-to-me-if-not-everybody-else kind of way, so that helps, but I guess I haven't reached so many files that I need a better system to find them 😅
Also, how does this compare to, say, a hard drive? Perhaps I could switch to storage hard-drives that I use exclusively for my yearly backup, which would (unless there's some chemical breakdown happening their too) significantly increase the lifetime 💽
(Or tape storage; I would ideally go for tape storage but of course it's much too expensive 😅)
Yes, they're bad and all, but they have incredibly cheap archival S3 plans. Paired with rclone (or a similar program), you could back your files up there. Saving and recalling costs more than storing, but if you plan to never have to take stuff out, it's peanuts (I think it was $5 per TB per year at Google, if not less)
I have a pair of hard drives, one of which is always off site. This is an air-gapped, offline backup.
Fancy modern filesystems like btrfs and zfs can keep you informed of your data's integrity. If you're willing to use multiple disks, you can get magical auto-healing.
For most normal people, automatic online backups are fine.
@badrihippo @kytta @kai It’s interesting to look at how docs are archived at the Library of Congress and at Standford University’s archives: #microfiche. And no, they are not obsolete. The decision to stick w/microfiche is based on the fact that film outlasts all forms of digital media w/tolerance for harsh conditions like mold, damp, & being dropped.
@kytta once upon a time I came across this rather comprehensive approach to naming things. I haven't used it but it stuck with me for its approach:
@kytta Two systems that have worked for me at various points:
1. Similar to Johnny.Decimal, but with three letter abbreviations and no numbers. I had only three levels for them the depth. And all folders were watched to find duplicates through a script (that ran on boot). This did not have any year based system.
2. Year based file system with a similar folder structure for each year.
In both cases, I’d used rsync for backup.
Please do share what you eventually choose!
Then mostly by topic/project and/or chronologically named directories inside. Johnny Decimal looks too rigid to stick to it. Maybe it works for personal files or abstract, administrative/managerial positions. Most projects comprise several kinds of files, and sometimes a high number of them, so I put those into different directories inside the project directory. Otherwise there would be pools of files, and I don't want that.
hello. maybe this link will be useful for you: https://karl-voit.at/managing-digital-photographs/
it is about managing files, not only about photo.
Are the folders I'm using for our shared family file system.
@kytta i usually make my things very hierarchical, e.g. my digital records are by type (e.g. shipping, banking), then organization/company, and then the PDFs. it seems like it leads to me avoiding duplicates better than if i had things in more vague directories, if that makes sense.
@kytta This question is really 2 in 1. How the data is organized and how it is stored/archived. W.r.t media, you’re aware of the lifetime limitations of optical media & flash media. So you’re using a hard drive to store your backups. Backup tech has become a sad state of affairs. Hard drives are cheaper than cassette tapes, but they have moving parts… heads that can crash. Not ideal.
@kytta W.r.t organizing data, free-world principles aside, no system impressed me more than the iTunes app. Each track has lots of metadata & the metadata is used to maintain the file in a heirchy of genre:artist:albums. If you change the genre, iTunes automatically moves the file (creating a directory if needed). Outside of iTunes, the tracks are simply in a well organized directory tree.
@kytta Although iTunes is designed for music, some ppl use it to organize non-audio files, like books. It’s just a matter of abusing the music metadata fields for non-music artifacts. That said, I don’t mean to propose iTunes as a serious recommendation. It’s just interesting to regard it as a model of how data should be organized.
@kytta Ideally, files should have metadata & there should be a tool that enforces consistency between the metadata and the file structure. My personal free-world hack is to use #procmail for this, surprisingly enough. A script can pkg any file as an email and create faux email headers from the metadata, and then a procmail script serves to direct where the file goes.
@kytta Note that a magnetic HDD cannot withstand as much g-force as a tape. But OTOH the cost per unit of tape space has gone much higher than that of hard drives.
Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.