Hello everyone,

You might remember last year I organised a Great GD.net Collaborative Coding Horror Experiment. It was great fun and you guys helped create some interesting games.

I'm planning on re-running a similar project this year. As per some of thelearnings from last year, I'd like to make some changes to the format.

The main change is to use real version control - my (hair-brained) idea to use the forum for that was definitely a source of issues. It seems like GitHub supports something called "Organisations", which should allow us to manage adding collaborators without making me a bottleneck (avoiding timezone issues).

I also want to remove most of the weirder rules governing the nature of the changes. The rule set will be something like:

* The code change should be relatively small and cohesive - add or change a single feature at most.

* Be kind to the features and code that others have contributed. Try to work with the existing code, rather than replacing it.

* All assets and code should be free of copyright restrictions.

So, what is this post about then?

I'd like to see if anyone is interested in trying this again. Contributors of all disciplines and abilities are welcome. Last time, we had a C++ and a HTML 5 version. I'd like to see if people want to try the same things, or if there is enough support for different language. A particular weakness of the HTML 5 version was lacking a library to make it easier, I'm not sure if there is something akin to SFML for HTML 5, rather than a larger framework.

Any thoughts?

If it is to truly simulate a real coding horror (and not just intentional coding horror for fun), have people sign up and then assign for example 2 hours to each person. Then plan a completely unreasonable number of target features, and try to implement as many as possible of these in the allotted time, and perhaps give participants points on the number of working features only. When typing speed becomes the bottleneck true coding horror will appear by default :)

I'm up for it! (the C++ version anyway)

If it is to truly simulate a real coding horror (and not just intentional coding horror for fun), have people sign up and then assign for example 2 hours to each person. Then plan a completely unreasonable number of target features, and try to implement as many as possible of these in the allotted time, and perhaps give participants points on the number of working features only. When typing speed becomes the bottleneck true coding horror will appear by default

Ha, yes, that would be an ideal way to foster some true horrors.

I say we set up a twitch script a'la Twitch plays Pokemon, and see how long it takes to implement Marching Cubes.

I'd suggest Unity as a good platform that diverse people could rip out a few features quickly, but from what I recall from a year or two ago the indie version isn't easily source controlled.

Version control is essential for such a project, but some light Googling suggests that it wouldn't be incredibly difficult. I played around with Unity for a bit, but it has been a while and I wouldn't be particularly confident in creating a starting project base.

In addition, I'm hoping to avoid "framework" style technologies where the barrier to entry is very high, in favour of simple., lightweight libraries that are easy to start working with.

If you don't mind a lack of graphical tools, you might consider Panda3D: while more advanced matters might call for a bit of experience (or community aid), it should be fairly easy to get started with, and I wouldn't expect significant problems with establishing source control with a Panda project. (That said, this is based on my experience with using its Python API; if you want to use C++ instead then my experience with Panda's C++ API is rather limited, and I don't know how easy or friendly to source-control it might be.)

Thanks for that, I'll take a look.

That said, it would seem there is limited enthusiasm this year. I'll probably kick it off anyway, just in case it is more popular when running than this thread has been so far. I'll stick to just C++ this time, probably with SFML again, though anyone who wishes to take whatever sprites I create and kick off an alternative language project is welcome to do so.

Quite a few of us do not use GitHub ... especially us Windows users who can not get it to work on our systems due to the ####ing authentication system ...