4 ms·

Google: the company may be taking a different approach by performing the encoding and decoding on its servers. Spider Oak: I wish there were some open source c

by stfu 13y ago

Google: the company may be taking a different approach by performing the encoding and decoding on its servers. Spider Oak: I wish there were some open source cloud client similar to True Crypt. Something that were really easy to use, working on every system and open source. Spider Oak's client seems to be still mainly a closed source product.

ippisl 13y ago
Syncany looks promising. It's open source, includes client encryption and support for multiple options of storage(google,amazon,rackspace,ftp and others),and it's only 8K loc of java. Of course it would still need a security review.
zorlem 13y ago
Have you tried duplicity [0]? It encrypts the archives using GnuPG, uses two separate keys - one for encryption, the other for signing, and can store the backups using a multitude of protocols [1]: > Currently local file storage, scp/ssh, ftp, rsync, HSI, WebDAV, Tahoe-LAFS, and Amazon S3 are supported, and others shouldn't be difficult to add. It is libre software and cross-platform. A number of front-ends address the issue of ease-of-use. [0]: http://duplicity.nongnu.org/ http://duplicity.nongnu.org/ [1]: http://duplicity.nongnu.org/features.html http://duplicity.nongnu.org/features.html
- dmix 13y ago
  I prefer Tarsnap. cperciva posted why Tarsnap was more secure and cheaper than Duplicity a while back: https://news.ycombinator.com/item?id=459334 https://news.ycombinator.com/item?id=459334
eblume 13y ago
I'm actually hacking right now on a personal project that might be relevant. It eschews the 'sync' metaphor and focuses more on the 'share' metaphor, though - it is not so useful for files that change frequently, but pretty robust for sharing arbitrary sets of files or data between a group of users. (It will also include user and access management). The whole thing currently sits on top of BitTorrent although I suspect I'll end up writing a forked implementation at some point to support some extensions I'm working on. Active, dynamic syncing is great for some kind of data but not so important for others, and I feel like that's the hurdle that needs to be overcome. More interesting to me is painless, fast, robust, and configurable sharing of relatively static data. (Some versioning may be included but it won't be a first-class system.)

by stfu 13y ago

ippisl 13y ago

Syncany looks promising. It's open source, includes client encryption and support for multiple options of storage(google,amazon,rackspace,ftp and others),and it's only 8K loc of java. Of course it would still need a security review.

zorlem 13y ago

Have you tried duplicity [0]? It encrypts the archives using GnuPG, uses two separate keys - one for encryption, the other for signing, and can store the backups using a multitude of protocols [1]: > Currently local file storage, scp/ssh, ftp, rsync, HSI, WebDAV, Tahoe-LAFS, and Amazon S3 are supported, and others shouldn't be difficult to add. It is libre software and cross-platform. A number of front-ends address the issue of ease-of-use. [0]: http://duplicity.nongnu.org/ http://duplicity.nongnu.org/ [1]: http://duplicity.nongnu.org/features.html http://duplicity.nongnu.org/features.html

dmix 13y ago
I prefer Tarsnap. cperciva posted why Tarsnap was more secure and cheaper than Duplicity a while back: https://news.ycombinator.com/item?id=459334 https://news.ycombinator.com/item?id=459334

eblume 13y ago

I'm actually hacking right now on a personal project that might be relevant. It eschews the 'sync' metaphor and focuses more on the 'share' metaphor, though - it is not so useful for files that change frequently, but pretty robust for sharing arbitrary sets of files or data between a group of users. (It will also include user and access management). The whole thing currently sits on top of BitTorrent although I suspect I'll end up writing a forked implementation at some point to support some extensions I'm working on. Active, dynamic syncing is great for some kind of data but not so important for others, and I feel like that's the hurdle that needs to be overcome. More interesting to me is painless, fast, robust, and configurable sharing of relatively static data. (Some versioning may be included but it won't be a first-class system.)