I've started implementing credentials for Obnam 3. They will allow access to the per-client chunk, which contain the encryption keys for all data chunks.
I started implementing an OpenPGP software key credential, but it was more work than I anticipated so I'm retroactively claiming what I did today to be a prototype that let me plan actual implementation work for the next few session.
Today in Obnam 3 development I finished adding command to mange a client chunk and its encryption keys. The code has now landed in the main branch.
Today in Obnam 3 development I added the beginning of support for "client chunks", which is where chunk encryption keys get stored. I have client chunks, but only just, and am lacking management commands for them.
Today I added data compression to Obnam 3 using DEFLATE (what gzip uses). Compression is optional at chunk encryption time and transparent at decryption time.
Today I changed the Obnam 3 command line interface to use the chunk repository by default.
Since March., 2025, I've been implementing some foundational components of backup software. For fun. In Rust. For about three hours at a time on Sunday. I take notes and post them on the Obnam blog. This may or may not become a backup application, but it might. If I start getting there, your help would be welcome.
https://blog.liw.fi/posts/2025/obnam3/
https://blog.liw.fi/tag/backup-impl/
https://obnam.org/tag/obnam3/
https://app.radicle.xyz/nodes/radicle.liw.fi/rad:zbWNQYkQ4QKgdSQcd1tjaemv6d6x
Today in my weekly development session for Obnam 3 I finished implementing a chunk store using local files, by adding subcommands to use one. This is not yet integrated to the previously implemented chunk manipulation commands, but that will come soon.
Today in my weekly development session for Obnam 3 I implemented the basics of a chunk store using local files. I ran out of time so didn't add sub-commands to use a chunk store, but I will continue next time.
I'm thinking about ways to storebackup. Obvious choices are local disk and a remote server with an S3 compatible API. What else should I consider?
Where and how would you want to store your backups?
Edit: Rewritten to clarify, as my original wording was unclear.
I finally got chunk encryption and decryption merged for Obnam 3. It uses the aes-gcm-siv crate for AEAD cryptography. I'm sure experts will now tell me all the ways in which I did this wrong, but I'm tentatively confident that I can do some other things for a while.
Maybe compression support?
Maybe a backup repository?
Not sure yet, but your input is welcome.
In today's Obnam 3 development session I got myself entangled in my own abstractions, but hopefully realized a way to avoid that in the future. I should write abstractions I want to use, not enterprise libraries.
In today's Obnam 3 development session I integrated the encryption things I did last time. Did not get to actually encrypting chunks yet, though.
In today's Obnam 3 development session
I took a step to the side to make sure I can get AEA-GCM AEAD encryption and decryption working at all, with a bit ergonomic wrapping..
Today's Obnam3 development session was not a success. I tried to do too much without having a clear plan.
My fourth Obnam 3 development session. Put into place the scaffolding to add chunk encryption, but didn't have the time to actually implement the encryption.
Small steps and tea is the way to reach goals while having fun.
(Obnam 3 is my Sunday project to experiment with implementing fundamental components for backup software.)
I did a bit of chores in today's Obnam 3 development session. There are documentation comments for everything, and command line help texts.
I've had my second Obnam 3 development session. I finished (at least for now) the chunk encoding functionality, with commands to encode and decode chunks, and inspect an encoded chunk.
My first Obnam 3 development session today. Got started with chunk encoding and decoding (in clear text, not yet with encryption). Write up for those curious.
I'm going to start Obnam 3, the third generation of my backup software, to explore the implementation of the ideas I wrote about last year. It won't result in usable software, certainly not this year.
Some more thoughts on backup implementation: snapshots vs deltas, file system snapshots, rsync and directories of hard links.
