Hello fellow developers. I've joined this list quite a while ago, mostly to keep a pulse on the Fedora development community, but also to look to become a package contributor. But before getting to that, a few words about myself. I've been "into computers" since mid-80's, started off with a 4.77 Mhz 8088 (IBM PCjr). I learned Unix in the early 90's on an IBM AIX system, where I picked up C programming and sysadmin experience. Which eventually took me into the world of Linux (I think it was kernel version 0.12, came on a boot disk and root disk pair I grabbed off a BBS, long before there was easy general public Internet access). Anyway I've been focused on Red Hat based distros for the past 15 years, and at my current employer I oversee about 700 systems installed at customer locations (where I was the resource responsible for packaging our applications and creating system build images). Any way, what I'd like to give back to the community is a really nice backup system called Snebu (Simple Network Encrypting Backup Utility). I initially developed this more than 8 years ago since there wasn't anything else that fit my needs -- I used it to back up my personal systems, and also in some lab environments. I've read plenty of rants that have been posted about how backups are either too difficult to set up, or don't support multiple clients, or require a repository encryption password to be placed in plain text on clients, and other issues people have. With that in mind, I believe that Snebu can be just what people want. Before going through and submitting the package for formal review, I'd appreciate some feedback on what I have packaged up so far. The current release is at https://github.com/derekp7/snebu/releases/download/v1.1.0/snebu-1.1.0-1.f..., and the project web site is at https://www.snebu.com. The main features that it has that are interesting: It maintains a centralized package database on the server (using SQLite3) tracking backup sets and metadata; actual files are stored in the filesystem as lzop compressed files using a file hash for the file name which leads to full cross-system file level de-duplication (so no proprietary file formats); uses a snapshot style backup strategy; it uses GNU tar as a serialization format to shuffle backups to the server which leads to how the public key encryption support was added by developing "tarcrypt"; and it works in single-system installs, client-push or server-pull backups, with no agent required on the client. Another interesting project that I may spin off is the above mentioned "tarcrypt" command. This acts as a filter for tar files, which adds RSA key data to the header (passphrase protected private key, public key, HMAC signatures, etc), compresses and encrypts the file contents while keeping standard tar headers in place (with the additional encryption metadata added via extended PAX headers). The details on this project is at https://www.snebu.com/tarcrypt.html. So far tarcrypt is part of the Snebu repository, but if there is interest then it may eventually be spun out as an independent project. BTW, the current .src.rpm file for Snebu mentioned above has passed through a valid build using the Fedora "mock" utility, and passed rpmlint. The only error rpmlint shows is: snebu.src: W: spelling-error %description -l en_US de -> DE, ed, d Not sure what that error is saying, as the text at the end of the message doesn't appear anywhere in the .spec file. Thanks, and I look forward to your feedback.