[Rab] I'd would suggest that should be set up when the game is defined. If a game proves popular as an ephemeral it would have to be explictly killed off and reinstated in the slow burners section when next a slot appears. The important point is that the short-lived games section is explicitly flagged as such, and everyone knows that not only do we have have a licence to kill them off, but that we are expected to do so. Possibly even flag the game's creation date and give it a defined expiry date - three months or so tops, I'd say, pulling a figure out of my hat.

I still quite like the idea of doing it implicitly. If active games were sorted in order of sluggishness or otherwise (a bit like at Orange where you can sort them by most recent post, but based instead on the frequency of the last, say, five posts). Unloved games drop to the bottom, active games fizz to the top. Every week the least active game is automatically culled (unless, say, it started less than a week ago or it has been given some special protection like the Furcation game).

I make this suggestion in the secure knowledge that, as I can't code, I won't implement it, which is probably the worst kind of backseat driving. :)

What SM said about time (though this is still my first port of call in the Morniverse, and for mostly the same reasons he gave). I don't agree about typing games in advance, though - how would you know in advance whether something is a quick turnover or a slow burner?

I quite like Projoy's idea about automatic culling. Or, and this is completely off the top of my head, how about turning the process on its head? The opportunity to start a new game is always open, but every time someone starts a new game, the least active one gets killed off?

Hmmm... I can see benefits to both the explicit and implicit approaches; the former is easier to code, but might be a bit onerous on (particularly) game creators. Furthermore, I've noted an increasing tendancy to be over-prescriptive in the game brief, and I think further prescription about how quickly moves are to be made might actually be counterproductive and cause all movement to cease completely. For this reason I think I prefer the implicit approach. As I mentioned before, I want to mine the data of all games played so far to see if I can devise some kind of estimate of a games "flagging" point, in which case it can be put on an endangered list and expire if no further moves are made. One problem with always having a slot open is that, as experience at the Lockisseum showed, you get passers-by just starting things for the hell of it.

[rab] Passers-by just starting things for the hell of it... is that necessarily a bad thing? Some of the most memorable games have started like that ("wigwam" on Orange springs to mind, but there are others). I'm starting to think we'd be better off it all games were treated as fast burners and we encouraged a much faster turnaround for games, to keep people interested and stop us all (me included) getting too caught up in only one or two games and ignoring all the rest. I've heard people say that years ago, the MC servers did have games which were much shorter, only lasting a day or so.

Killing a game is easy creating a game that people want to play is more difficult, something different does not go down well, so players do not feel confident in filling a vacant slot. My worry over a timed game is once it has been timed out it will reappear as ...the Return Again of... which happened only recently. I would like to see more people prepared to start games and not stand on the side lines saying "what shall we do now, what shall we do now". With a regular turnover there will be flops and hits. The reason games hang around is the reluctance to replace them. I would not want to see time limits just more bravery with the sythe. Afterall what is the turnover of games on MCOiS where there are 18 active games.

[Darren] It's not a bad thing if it occurs only once in a while, but on the Lockisseum nuisance games were started on an almost daily basis until Dunx brought in the login system. I neither want the maintenance burden of cleaning up, nor to enforce logins for the starting of new games.

[Inkspot] Well, that would be all well and good if people actually killed games, but they don't so it seems that some form of encouragement is needed.