IPFS

[Like Ra]

The IPFS client can be downloaded from https://orion.siderus.io/#/download

Main implementation of IPFS https://dist.ipfs.io/#go-ipfs

Plugins for browsers:

https://addons.mozilla.org/en-US/firefox...companion/
https://chrome.google.com/webstore/detai...gkdhaomnch

Test:

[Tinker D]

Well I can see the picture and it’s is moving.

[Like Ra]

Good!

[Like Ra]

@Anne - in other words: It works and I'm in 😉 Could you, please reupload/reinsert your videos? I'll "pin" them.

[Like Ra]

Well I can see the picture and it’s is moving.

I can still see the picture after I stopped IPFS locally, so the global IPFS cache is working.

[Anne]

I can still see the picture after I stopped IPFS locally, so the global IPFS cache is working.

Because you used https://ipfs.io to display it, it is cached there. They run a node so people not running IPFS can use that website to view files on the IPFS network, it will stay cached for about 24 hours, after that the file must be available on another IPFS node or it will no longer be found.

Anyway, I started searching in my backups for my vids so expect them shortly. I'll make sure to post the hashes as well so you can pin them.

[Like Ra]

IPFS is available for QNAP 😉

[Like Ra]

I see some obvious problems with IPFS:

o- ipfs.io is SPOF. If it's down the links will stop working (some files will be cached by CloudFlare, but most likely it will show "can't connect to the backend")

o- I could not find any means to check what you have added and what this is (the Orion program keeps a local database with all the names and comments, the "ipfs" programs knows nothing but hashes). If you lose the hash, there are no means to find it out again without the original file, and it's not possible to figure out what this hash correspoids to without downloading the stream of bits, analyzing the content and opening it with the correct program.

o- Just like bittorrent, IPFS has no anonimity (yet?).

[Anne]

- ipfs.io is just something like an "exit node" so people not running ipfs locally can still get to the files. If it goes down I can still use the hash to get the file from another node or locally. The "more work" solution would be to run your own exit node, if it goes down the forum will probably be down too. I don't really see this as a problem, nodes can go down and the file will still be available, that's the one issue that ipfs is solving.

- indeed, files are identified by their hashes so unless you keep track of what is what you'll need to get the file and see for yourself

- no anonimity, that's not the point. It is possible to find all ip adresses hosting a particular file.

As with all technologies it was designed to solve a problem, with ipfs you can serve files with lower bandwith requirements and make sure not one party (like youtube) can make your files unavailable to the public. I could run a webserver on my server and host the files the traditional way but I'd burn trough a lot of data and if my server goes down or I reach my data cap the files will be unavailable too. So I opted for IPFS because when that happens I hope some nodes will have my files pinned or cached.

[Like Ra]

24 hour statistics (I "host" only a couple of files announced here):

% ipfs stats bw
Bandwidth
TotalIn: 6.5 GB
TotalOut: 11 GB
RateIn: 69 kB/s
RateOut: 1.6 MB/s

I expected mostly outgoing traffic...

So I opted for IPFS

I was going to play with IPFS anyway, so it's a perfect opportunity 😉