/hydrus/ - Hydrus Network

If you have the space to spare, you could just import everything from your folders without deleting the originals.

Basically, the problem is that when you use a custom folder structure, there's a very likely chance that humans or other programs are going to move, delete, or change the images contained within. Monitoring all of that would be a difficult, complicated, and computationally expensive task, when hydrus is already a rather expensive program. It's just not worth the time or effort.

So instead, hydrus takes the easy way out by using its own folder structure. And honestly, that's not a bad trade off. If everything is tagged well, then you can find a picture about as quickly as you could with folders. It's still easy to post to imageboards or what have you. And you get all of the extra features Hydrus has, like subscriptions.

I personally would be okay with a long and expensive boot up/re-calculating phase if it meant I could keep my existing file structure. I do understand the challenge of this however. There would be a lot of monitoring going on for changes and things would need to accommodate pointers to files rather than file data itself. While I want this thing I'm not asking for it.

Things like Direct Connect come to mind where it stores hashes of files live like that and recomputes on changes.

I like the idea of hydrus adding functionality on top of an existing structure since it would be low risk to adopt without a space cost either, it is easy enough to export though and if your files are properly tagged or tagged with your own metadata used soley for exporting, on import, you should be able to reconstruct some kind of similar structure already. I am concerned about storing my entire library in 1 database file though and interoperability with external tools, for example sharing your media library over smb/cifs/sshfs, etc

I don't mean anything other than to voice my thoughts on this.

>>1818

This is pretty much how I feel.

I don't mind if it takes a couple of more hours to search through all of my stuff once a month or whatever.

I used this for a day.. and I can already see how absurdly powerful this is. All great technologies have a learning curve and this is no exception. I've tried doing something similar in the past with shimmie2, but the lack of native tagging meant that I had to resort to saving the tags to the filenames resulting in truncated tags. It was automated… but very messy.

Truth be told, it took me about 5 hours of poking and prodding with Hydrus.. but now I understand the general flow and can use the program pretty much flawlessly and I LOVE it. Once you get used to this you won't want to have your previous directory system because your sorting methodology is vastly inferior to an advanced tagging system that leverages the power of crowdsourcing.

The beauty of this also is you can add your own "local tags" to simulate a personal categorization system thereby simulating the physical aspect of a virtual filesystem.

If you have 600gb, you better be backing up your content and having them in separate folders creates more problems than it solves unless space is an issue and you are unable to backup your data so you are relying on the spreading out your data so if some of your pictures – or nuts – gets destroyed, you'll still have others.

So the next time you want to look for "purple_haired beefcakes", you'll know that instead of searching hundreds if not thousands of folders.. you'll have a vast array of files prepared for your viewing pleasure in 24 keystrokes and 2 mouse clicks.

(Forgot to add my conclusion)

>>1852

options > default tag import options > add > domain

>Is there no way to have it simply monitor my existing folders?

That was my big question when I first started using hydrus. "This would be great if I didn't have to commit 100%!" is the thought process, right? Well, I complained about it bitterly here to the dev, but kept using hydrus anyway…

and now I realize how dumb it would be, and how much workaround shoddy code hydrus-dev would have to implement to get it to work.

There should be a sticky on this board and in the docs: "new users: use hydrus for at least a month before suggesting the ability to monitor existing folders".

I think the biggest reason people want that feature is because they're uncomfortable with the idea of having to exclusively access images through a program rather than what they're used to. People want to have hydrus' tagging system in addition to <crappy way that doesn't work but is what you're used to>.

Give it time. Hydrus dev is extremely accomodating to feature requests, hydrusnetwork is extremely stable, updates are extremely consistent. Having to drop what you know for some other system can be scary, but once you get to know it you'll probably never even use the ability to monitor existing folders.

>>1887

>Type in a few letters to make results appear

Is that supposed to happen automatically? For me, that bottom list area doesn't update unless I press the tag/file scope buttons. (And in fact, I didn't know what it was even for until just now.) It's like the element isn't updating in response to changing the contents of the text field. Shall I go report it as a bug?

>Page up/page down

Yeah, I'm an idiot, thanks.

>>1888

Ah, never mind, I'm a double-idiot. Turns out I had misinterpreted one of the settings for auto-complete and had basically turned it off. Move along, nothing to see here.

>>1882

Personally, my issue is thumbnails are way too small, even at 200x200. I'd like the option to make them at least twice the size, filesize be damned.