How One New Service Is Tackling The Death Of The File System | Hacker News Reader

How One New Service Is Tackling The Death Of The File System

fastcolabs.com

19 pointserikcaso14 comments

They are making the same old mistake of over-abstracting.

The fundamental difference between local and remote is the access latency. Until they can get remote content from there to here as fast as they can get it from a local drive, all this "file system is dead" will remain a non-technical marketing bullshit that is out of touch with the real-world application issues. Why the hell you would want to hide the fact that file.mov is 10 hops away behind a saturated link? You don't.

"Death of the file system." Jesus.

When I read this, especially the word "patented," the question that springs to mind is: why isn't Tahoe[0] succeeding?

(The same reason Linux on the desktop isn't? Other?)

[0] https://tahoe-lafs.org/trac/tahoe-lafs

Tahoe looks interesting, I hadn't come across it before. What exactly is not succeeding about it? It sounds like BT sync minus the closed source / protocol.

Well, I chose "not succeeding" over "failing" on purpose. The project has been around in one form or another for about 8 years, but it still hasn't hit the mainstream.

Make no mistake, I'm a fan, but I'd rather see all these commercial players making proprietary front-end improvements to Tahoe instead of reinventing the core wheel.

Whatever Tahoe is, maybe it's not succeeding because their web server is throwing Connection Refused errors?

I would rewrite the title to be "the death of being aware of the filesystem".

The filesystem is still there. They are creating an abstraction layer such that you don't care about it anymore.

This is indeed the goal, marketing terminology notwithstanding. We're trying to create a system to tie a file to a user (regardless of device), rather than to any device itself. And to do it in such a way that the user doesn't need to know (unless they want to) on which physical device (or devices or services) the file resides.

I see what they are getting at, I was recently traveling with an android phone, iPad, and macbook. I generally like to keep my photos organized in aperture but between android, drop box, iPhoto trying to import my pictures, galleries I had exported from aperture to Facebook, and being unfamiliar with how to manage photos on my phone there wound up being versions all over the place. I lost track of the canonical versions, and have sets with some reduncies. So if o want to get them all, I think I'm going to import all the photo sets into aperture and sort it out

An app that could help with this would be great. What's even more annoying is the confusion that comes with managed libraries I'm the Apple apps. Unless you're paying a lot of attention, those files are pretty much taken out of the file system already (the photos that is).

What Younity is attempting here sounds very interstating, but what a challenge! How can you do effective backups if you don't know where all of your files? Is the idea to have some kind of super cloud sucker thing that backs up all of your chaotic files and structure/state?

Good luck with this, I think you might have a great product

I don't believe this is news ... https://github.com/KnpLabs/Gaufrette (lib for php). only if they manage to buy their way in to the consumer will they have something of value.

Love the in-depth analysis in this article and seeing if people really don't want the file system. I love organizing in folders, but at the same time true file ubiquity could be a dream come true!

> "It is forecast that the typical American will have 5.8 devices by 2015"

http://xkcd.com/605/

So Bill Gates was wrong about a computer on every home ;-)

Boolean projections are fundamentally different than numeric ones.

Consider the distinction between cell phone penetration rates and device sales projections.

so basically a distributed semantic filesystem?

Reply on news.ycombinator.com