That last point is especially valid. So often I sit at the PC thinking I can do another update/re-release of StarLines INC rather than start a new game with all the effort that entails. I guess you just have to bite the bullet now and then and start afresh, and once you are at the end of another product ( as I am now) you feel much better about it than doing another re-release.
Now if only I had a great idea for another game....

http://www.positech.co.uk

I have an idea about that last one too...
One word people : Sequels! ;-) generally you could
use some of the same code, graphics and whatnot to
continue to update and improve an already existing
product. But as it changes over time people would
buy each game as a new game, and in fact it would
actually advertise itself, like a franchise.

Keith

Which do you think is better with shareware games, time-limiting, or feature limiting to get people to buy?

To address your question:

What do you think will convince them to play more, taking away features that they haven''t seen or taking away the game that they have seen.

One is based upon building a game that will hook them, the other is built off making a game that will intrigue them and make them wonder what else is available.

If you feel your game has high replayability, I''d go with time limiting it, if it''s a role-playing game, or a game that has a definite goal in mind, I''d suggest going with something that would simply not allow progress beyond a certain point. Basically give them as long as they want, with all the features, but limit the level and the areas they can access to the first 1/4th of the game.