I've been running a podcast together with two friends for over a year now. We have a recording and mixing process going, but unfortunately it involves using non-free operating systems and software all the way through.
It is probably going to be hard to get away from using the things we use for the actual recording part for a while, but one thing I feel should be perfectly doable in openSUSE is the mastering bit. In theory, Audacity could work for this, but it is old and clunky and doesn't do real-time filtering, making it annoying to use.
The idea is to play around with a minecraft-like block exploration game, written from scratch in C using SDL2 and OpenGL 3.0. Minimal dependencies, probably won't be a game as such before the end of the week, but the goal is to have a world generated and to be able to walk around in it. Why? Mainly to refresh my 3D knowledge, catch up with what's happened since I last played with that stuff, and to have fun. :)
* Source code
The use of the bitcoin testnet for trading cards for Hackweek got me thinking about a protocol more adapted for the tit-for-tat trade of virtual goods: the transaction only completed once both sides have contributed their share to the transaction. It would need some way for each side to verify that the goods traded are those agreed on.
It would be an interesting problem to work on (or find some existing solution).
almost 7 years
Has no hacker:
I would like to have a programming language that has the performance characteristics of C, but integrates some programming language features that I like from other languages. My first goal would be to start with a language that is very limited in scope and compiles directly to C, and step by step add to it.
There are a lot of programming languages, but very few that handles memory as an explicit resource. To be able to explicitly manage memory is a very powerful technique and essential for applications such as high performance video games.
I want to learn Erlang and see if it makes sense to use it in the context of High Availability (it would certainly seem like it). As a learning project, I want to try to reimplement my koi project (https://github.com/krig/koi) in Erlang. Koi is a minimal cluster manager I wrote in C++.
With rust 1.9 released, it should be possible to from now on bootstrap rustc from the previous version of rustc (so 1.10 can be built using 1.9 etc.). This means that it should now be possible to create a rustc package which no longer needs binary snapshots to build, meaning that we might even be able to submit rustc for inclusion in openSUSE Tumbleweed.
This is the first goal.
The public cloud is generally not a high availability-friendly environment, with unpredictable I/O latencies and machine stalls, custom tools needed for handling IP assignment etc. The idea with this project is to see if there is any way to leverage the Raft algorithm and booth together with pacemaker_remote for resource management to manage cloud instances and services running in the cloud.
For testing Hawk, we're currently using a Vagrant configuration, and for testing HA releases we've been using a set of scripts originally authored by Antoine Ginies as a Hackweek project.
My vague idea is to combine Terraform, Salt and a custom web frontend to make a tool that can be used by us as developers but also for others that want to try Hawk or play around with a cluster, for example for the UX team when testing modifications to the Hawk UI. I would base this on the work done by the SUSE Manager team:
I want to create a more modern mail storage format, which leverages git and tagging instead of folders to manage my mail.
This is inspired by having used notmuch and mbsync for a long time, liking the good aspects of this setup but getting frustrated with the problems. Mainly the issue of storing mail on multiple computers with eventual consistency (for example being able to read mail on my laptop when travelling but my desktop computer when at home).