HN Explorer

Ask HN: GitHub projects for the job search

5 points by kdaley3 ↗ HN
Hi,

I have heard various things about the role of github in the job search, but it seems to be more important. I am about to graduate school with a Mathematics degree and lots of informal coding experience but few completed projects (that I can share; most of them are research work done for hire).

I need to put something on my github. If you are an employer, or if you were starting out fresh today, what sorts of things would you write and release? I can make pretty much anything; I know how some of the most advanced 3d algorithms and machine learning methods work, and my research is in dynamical neuroscience, but finding ideas for code that people outside my field will actually want to use or read or care about is nontrivial ;P. I also want to control scope so it can be done and polished in a few months' time.

By the way I am pretty new to HN so I have no idea how to submit this to the right category. I just edited the title; does that work?

2 comments

[ 0.22 ms ] story [ 15.2 ms ] thread
viraptor
When I was looking at other people's github accounts during recruitment, I didn't care what the code is for. People have different interests - games, web frameworks, fun research, anything... As far as I was concerned, all that mattered was: does it look usable, does it follow any conventions/patterns, is it made to be used by others. I didn't care how advanced it is, how many features it has, most likely won't even try to run it - I just care whether the author cares.

For example if someone published a library that they expected to be used, there's a whole spectrum of quality. On one end there's a dump of messy code that's not even indented properly with a single commit "Done". On the other there's a project with a proper license, documentation, version tags, "how to use", potentially some travis-ci integration and other source code scanners.

Examples:

- throwaway project I don't care about: https://bitbucket.org/viraptor/cmv_encoder/src

- (mostly my) proper project I care about a lot: https://github.com/openstack/anchor

So I guess my advice is - just do something you actually care about. If it's only for people in your field - great!

kdaley3
Cool thanks,

I will think on that, that definitely opens up quite a lot of interesting projects which are loosely related to my research. For example, my professor is working on another project related to consensus dynamics in autonomous fish drone formations. I am not involved in that at the moment, but I could definitely create a simulation testbed related to such work (maybe for aerial drone formations) with a 3d frontend. Actually that's a really good idea.