Seafile advantages

Hi,

as a general question:

What do you think is the difference between Seafile and Owncloud or what advantages does Seafile offer that Owncloud does not?

Why did you choose seafile ?

Thank in advance! :slight_smile:

a.) Seafile has high transfer rates (more data over the same bandwith)
b.) Seafile is lean (2 cores, 4 GB RAM is plenty enough for 100 users - not counting online editors, AV, full text indexing, …)
c.) Seafile’s desktop client are much better, not to say unparalleled
d.) Seafile supports end-to-end encryption in practice, not just in theory

Enough reasons?

2 Likes

I chose it mostly because I like the architecture. From what I understand it is more similar to git and git-annex than to owncloud/nextcloud. It’s more than glorified FTP server with rsync.

Seafile does file indexing, versioning, diffing and it can detect corrupted files on the server with complete certainty. Because of the diffing capabilities it performs delta compression on library updates, only the changed bits get sent through the network. Also, it will almost always deduplicate files on the server, files are split on unique “blocks” with an ID that uniquely represents the content, if the content you’re trying to add is not unique, the ID will already be on the server, so it just references the existing data instead of making a copy. Because the “block” thing, files on the server are not really readable as-is, if you browse the data directory you’ll find a lot of oddly named files all with the same size, each one is a block, so to read the files directly on the server you’ll need to do some linux stuff to mount the library and read it. Each “library” is like a git repository, it’s in most cases completely isolated and can be migrated from one server to another without much pain. As I mentioned, files on the server are stored as blocks and you can’t do much with them, you can do some linux things to mount the library and then read it, but only read it, Seafile expects to be the only one reading and writing these internal block-like files, you could of course sync the files within the server and use them that way, but you’ll consume more space.

Because of its similarities with git one would think this software would be ideal as a “decentralized could” but actually I haven’t seen anyone use it this way, nor it being advertised in the “community edition” of the software. And yes, seafile is split in enterprise and community editions, I don’t have a problem with this since most enhancements in the enterprise version trickle down to the community version at some point, and it helps fund the project.

Sorry if this sounds a bit too technical. Seafile is really easy to use and manage! Because of how similar it is to git in architecture it really knows its files, so synchronization is top notch, it never misses a file, and if it finds conflicts it will make a copy just in case. Besides, because of the versioning thing, you can roll back to previous versions of the libraries with ease in the case you mess up/get infected with ransomware.

3 Likes