But consider the following. You know how to build the game on at least one platform, since you did so already in order to test it. The effort for you to provide a binary is close to zero.
...
Yep, sure, providing my binary would be close to zero effort, but keep in mind I had 3 minutes to go at the time haha, and I didn't want the upload form to be locked so that I couldn't upload at all. Even packing and uploading a binary that already exists within 3 minutes is a bit risky.
I also figured that since my binary would use shared libraries, as you say, the user would have to grab the dependencies anyway and if they do, an additional grabbing the game source, cmake, make and play is not much more effort.
But it is true that it keeps people away, but even if I would have provided a nice linux binary that would work for _all_ linux peolpe, it would _still_ keep all windows/mac users away which is kind of a big user group. This is a problem that is not easily solvable, and the more dependencies that are in play, the worse it gets. After all, who wants to install 5+ dependencies just to try out a gamejam game?
This seems to be a broader issue for game jams in general that incorporate users with different platforms.
I also see that it can be too much to request builds on multiple platforms by the developer, but as you say we could require that the game use only portable and available dependencies. Somebody of the community or the Game Jam Team could then take care of building it for missing platforms.
Yeah, indeed and that can indeed be a solution.
Another possible solution which would be cool but would take more maintenance effort would be if someone sets up a buildbot instance or something and every participant gets ftp-access to a folder to put the game's source with a cmakelists file and the buildbot automatically builds it for the major platforms. Then if you want dependencies, there is some time _before_ the game jam to request which ones you want, and have them installed on the bot. This solution would not really be worth it though if there are as few participants as we had.