phomes blogs about the role of Gnome-Games, wondering about the purpose of it, and criteria for adding new games.
While I like that the quantity of games is controlled, I'd like to see a better breadth to cover a wider variety of games styles/types and appeal to a greater variety of game players, and for each type to see more variants playable within the game.
So, say there are a set of game categories, like "Card game", "Chess/checkers game", "Falling blocks game", "Bouncing ball game", "Invaders game", "Platform game", etc. and one game "package" for each category able to implement multiple variants of that type of game. So the "Card Game" category game might play several different solitaire variants, cribbage, poker, etc. The "Bouncing Ball Game" entry might do pong, breakout, billiards, foosball, and pinball.
It sounds like there is an issue currently, in that Gnome Games gets many proposed *new* games, but has a harder time attracting maintainers for already accepted games. There seems to be good motivation for getting *in* - namely, seeing your game and your name installed on every Gnome desktop - but less motivation for maintaining/improving the game.
A thought for addressing this might involve encouraging a "engine + implementations" architecture, where certain code, artwork, and so on is shared by multiple games. A requirement could be set that a game implementation must be no larger than X mb's, but that anything it uses from the engine (and shared resources) is not counted against it. This would allow people the freedom to create new games, but encourage them to reuse or update the shared portions.
To combine both of the above ideas, Gnome Games could allow anyone to propose a replacement for a given category, so long as it implements ALL the games and functionality of the current entry in that slot, *plus one* new game implementation, and must fit within specified disk, memory, and cpu requirements.