The deadline was yesterday and I wanted to make a point to not look up the results before today – that failed simply because Google was a lot faster in sending out the mails than I or #gsoc were anticipating, so my supposed to be last mail bag for the day included already a mail with the title "GSoC 2016: Congratulations, your proposal with Debian Project has been accepted!"…
"That was a mistake." was hence the first thought crossing my mind.
I was clearly only referring to fetching mail of course…
At the beginning of the year I was evaluating jobs to fill my gold chests short of dungeoneering (all the Goliaths slain in my hometown years ago) which coincident with me getting mails about GSoC in Ubuntu thanks to me still being subscribed to ubuntu-soc after my GSoC2010 project: MultiArch in APT started there and was moved before the start of the GSoC to Debian. (The subscription to soc-coordination elapsed at some point as I was frequently kicked for bouncing mails for some reason).
That planted an idea: Why not do something meaningful for an OpenSource project… but I wasn't completely sold on that idea: I was a student in 2010 already & for all purposes firmly established in OpenSource development, so that felt a bit like stealing an opportunity from another student.
Weeks passed without much thinking about it until I came to Mozillas GSoC page as part of my adventure in Firefox extension wonderland. Nothing peeked my interest right of the bat, but I made a bookmark and thought about revisiting a few days later with my extension being done.
So here I was in the mid of march kinda seriously considering investigating GSoC2016 further then Michael asked how it would look on the job front and if I had considered GSoC. Well, I had, but not for Debian because I thought that would be wrong for stealing reasons mentioned above – but I knew which project he had in mind because we all agreed for years on the "something should be done" part of the project. The problem was mostly in the "time and work" part of the project – and the "please lord, let this cup pass me by" part of course.
I was talked into trying it anyhow, but in an attempt to have an easy way out for everyone (including me) I asked for opinions about me running first. Didn't work out that "well" with two non-negative responses. So I had to propose in full with project description and a fully fledged out proposal from me. All set and done I also talked to other students about the project on deity@, in a few mails: The first one and a few more "just" referring to it mostly.
Eventually the proposal deadline passed and business as usual continued. Including the occasional ordering bug which was a good reminder on why something needs to be done "sometime" and also why nobody is doing it as it eats time en masse… In this bug for example I ended up putting "just" a day after I was able to reproduce it mostly because after comparing the extended logs (= with additional debug options) I had a rough idea what could be wrong, which wasn't only correct, but also easy to fix. Usually patches tend to be a lot more involved. In exchange I still haven't found a testcase for it which isn't a 'real' system…
Anyway, time passed, the announcement of accepted students came and I am one of them. Woah! Thanks a lot to everyone involved! I am obviously very happy to be one of the chosen few (again), but I also have a lot of respect for the task. Its not going to be easy… so I better start soon! 🙂
The first month is official bonding time. There isn't much for me to do here given that I know my mentor, the code and Debian for a while now, so my timeline is using this as a head-start to flesh-out protocol and implementation. This is something to be talked about after the first week through as that blogpost runs way too long already.
And before coding, I actually have one thing still t do to bond me closer to Debian, which I have carefully avoided for reasons until now… lets see how that goes.