Saving defeats the purpose of one-shot runs, I believe. I could start out, work my way down the levels, and continue saving at each new dungeon level. Once I die, I can reload from the level's entrence, and continue leveling. With this approach, there is nothing stopping players from reaching ungodly levels, dominating the high scores. You'll also have to encrypt saves so people can't edit their first save as their most recent, giving them uber-level at dungeon level one, or even in the town.

Maybe it would be best to not allow saving at all.

~X

P.S.

When is an RPG not an RPG?

When it's DBZ. ;)

So... what's stopping you from saving again once it's loaded? I still think it's against the whole concept of the game.

~X

So... what's the point of saving at all?

Then just quit. It's not like the game has any long-term goals. It's all one-shot runs into a endless dungeon. There is no beating that sort of game, and the only reward is a score. I don't think there's enough incentive for the players to warrant saving.

~X

Xooxer wrote:

Then just quit. It's not like the game has any long-term goals. It's all one-shot runs into a endless dungeon. There is no beating that sort of game, and the only reward is a score. I don't think there's enough incentive for the players to warrant saving.

Unless you're an exceedingly vain person (which unfortunately most players are), the only meaningful incentive conventional RPGs ever have is the opportunity to advance to new areas, face new challenges, and just see new stuff in general... which is exactly what the idea behind temporary saving would be. Suppose there's no save function: you fight your way down to floor 99 (or whatever major landmark you're approaching), *CLICK*... your modem decides that now would be a good time to disconnect itself.

For any game which takes more than a couple minutes, if it's worth playing, it's worth saving. My main pet peeve with arcade-type games is that too often, getting the high score ends up being mostly a matter of being able to sit and play for the longest stretch of time. Such games--pretty much any game, for that matter--should be based on the player's ability to adapt to varied new challenges, not their ability to spend vast quantities of time plugging away against the same old challenge. You can spend time fleshing out the game and add more stuff to breathe life into later stages... but if there's no option to pause the game and take a break, that will still be irrelevant to anyone who doesn't have the time for huge marathon sessions.

Xooxer wrote:

Your modem going *CLICK* isn't going to secure you a save, and if you program it to, then it's just as open for abuse as before. One-shot runs should be that, or not that. Saving turns them from one-shot runs to a shot and infinite retries. Even if you removed the save when it's loaded, the client/host (depending on your saving model) can easily replace it and try again. You had a good thing there up until saving, then it just becomes a game of who can backup and hack their saves. It ruins the game for people not cheating, and turns it into a game of who can cheat first, fastest and best.

If it was a distributed hosting model, you would have a point here and I'd be strongly reconsidering the idea. However, while I do expect (and at least hope) that the game would garner some players, I don't expect in the least that it's going to be a huge runaway success; I don't see the game model presenting any problems under a centralized server approach. If someone can hack their savefile without having any access to it, then withholding a save feature isn't going to stop them from cheating.

Now if the game was geared more towards soloing, a bit more like a traditional Roguelike, then having to log into a main server would just be a hassle; the sensible distribution model would be to have everybody be able to download it and play it on their own offline. This would be the only case that would necessitate giving savefile access to anyone except myself and/or a dedicated host--and it would also remove most of the need for a save feature. If you're just playing offline on your own copy, pausing the game is a matter of minimizing it and leaving the computer on, but a lot more things can go wrong when you throw in an online multiplayer feature.

I do see some potential loopholes for abuse, but they've got nothing to do with the structure of saving or savefile manipulation, except for outright bugs that could develop. Players in dangerous situations could log out and wait for friends to come to their rescue, then log back in and join them (there would under no circumstances be multiple save slots available, to prevent this sort of abuse from being even easier; I suppose multi-keying would be a problem, too, but it would be a problem regardless of whether there's a temp save feature). I still want to keep the idea of being able to put an expedition on hold and return to it at the exact point you left off at, but I will have to think pretty hard about the details of implementing it; I may even have to limit it strictly to an emergency save feature, to cover brief disconnects, reboots, and server outages only.

Also, I should point out that there would almost definitely be saving of a different sort, not only for high scores but also because, as I said, I do have plans for some minor incentive system. This would consist of awards players would get for good performance, these awards taking the form of tangible "head start" benefits on future runs--a player who makes several successful runs to landmark levels would get to start out at a fraction of that level, e.g., a player who makes it to level 100 a couple times might get a combination of awards that would allow them to start out at maybe floor 20, being 20th experience level, having equipment equivalent to what you find on and around floor 20, etc. These would only be cumulative to a very limited point--once you get the relevant 2 or 3 awards for a certain landmark, you couldn't just keep making runs to that floor and bumping up your starting point except by making it to the next landmark--and would not impart any advantage beyond a head start. A player who's earned a free start at level 20 would not have any tangible in-game advantage over a player who just started and made it down to level 20; their bonus is strictly that they start there without spending the 15 minutes or however long it takes to get down there. Indeed, I think that with some of the features I have planned, taking a higher head start could handicap your run, in that you'd be getting less time to accumulate expendable items (potions, scrolls, etc.) before you start running into the harder stuff; players would, of course, have the option to shelve their awards if for any reason they want to use an easier level for their starting point. To some extent, it's mostly a token gesture--players are pretty spoiled nowadays and they pretty much insist on some tangible (relatively) reward for their efforts--but I think it would be a nice perk without spoiling the essence of the game. I hope that the game's party-oriented dynamic keeps things fresh--since getting really far depends on player interaction, it's not going to be a matter of the player going through the same routine every run--but if you're routinely getting stuck somewhere in the early 100s, nothing's gonna stop the first few levels from getting old really fast. Again, though, this would be independent of any in-dungeon save ability.

Xooxer wrote:

Then don't remove their mob right away if they disconnect in mid-game, like Lexy did with HRH. This would eliminate the need to save when their connection dies, as long as they log back in before their mob expires.

There was a reason I decided on saving the mob rather than just leaving it there, but I can't quite remember it. I think it had something to do with party members logging out in the middle of a battle, but I forget exactly how my logic went.

You can curb multi-keying by restricting access to a single instance on a machine. A cookie should suffice for this, or even a client-side save with good encryption.

Ooooh, I never thought of this. Having the server check the IP addresses always sufficed for me, but this is a really good idea.

You could offer a pay-to-save feature. Allow everyone emergency saves when the game or server goes down, but only allow a person to save on their own for a price, and only at certain locations, like the start of a level. Be it items, trophies, gold or whatnot, make saving cost. Each successive save in the same day would cost twice as much, and not to exceed a certain amount, say 2 or 3. Deeper level saves should cost more. Unused saves aren't accumulated. Maybe allow subscribers a free save plus 1 or 2 pay-to-saves a day as an incentive to subscribe.

Hmm... all interesting ideas. I'm currently leaning towards two ideas:

1. Saving requires a special item that can only be bought out of your starting funds, before you enter the dungeon. This both limits the number of saves and gives you a handicap (less money to spend on other supplies), and forces you to plan ahead.

2. Allow saving only after certain landmark points. Maybe every 20-25 levels; possibly every 10, depending on how long I end up making each level. Again, though, saving would end your current session and once you reload, it's gone... and you can't save again until the next landmark.

I like the party aspect you're describing. It would be sweet to allow players to form complex combo attacks by directing thier focus through members of their party. Perhaps a mage could enchant a party member's weapon to temporarily boost it's attack, or two fighters could meld their styles to provide almost impossible to defend against attacks. You know, the old "you go for the head, I'll go for the legs" tactics. Yeah, I can see a lot of fun potential there.

Mmm. I hadn't really put a lot of thought into that--most of the fighters are differentiated more in terms of what type of enemies they can take--but I definitely will have to add a bit more there. I think I know a few ways I can add some of this pretty easily without having to bend any class's role out of shape.

Spuzzum wrote:

With the first "pay for saving" model, people who like to play in tiny spurts will be more or less hopeless in the game, since they'll be severely disadvantaged and unlikely to make any real progress. With the second "checkpoint" model, people will have to play for longer periods of time in order to be able to save. If I had to vote for one of those two options exclusively, I'd vote for the second. You did mention that progress was fairly fast through the dungeon. I'd just be strongly against any method which required you to play for more than 15-30 minutes.

Ultimately it's not really meant to be played in tiny spurts over a long period of time, at least not as far as individual runs go; it does have a multiplayer, group-based focus, which means you don't want to take a lot of breaks on one run (since each time you stop and come back later means 1. you just let down several other players, shame on you, and 2. you have to hang around and try to hitch a ride with another group, which is especially hard when you have a reputation of adventuring 5 minutes and then quitting), and when possible you do want to break at the most convenient point available, not just in the middle of nowhere.

But there's also the considerable issue of what happens if you have to quit suddenly because of a real life issue. Neither the pay-for-save nor the checkpoint models cover that contingency (with pay-for-save you could allow yourself some slack, but even so, you can't expect the unexpected)...

I might combine the two in some way--set up primary save points every X levels, and allow additional saves anywhere with some limiting factor. I would be making an effort to balance different paces through the dungeon, so that an all-or-nothing crash run to drop as many floors as fast as possible is roughly equally viable with a slow, methodical approach designed to find as much equipment and loot as you can--and not only that, but allowing and even encouraging a mix-and-match approach, so players can build up supplies when they have the time to take a while between checkpoints, and make a mad dash for the next goal when they don't.

Its almost impossible to beat most games on BYOND. And yes the only reward for basically all games is a high score.

Braging rights tend to be an excellent reward. People will go to extremes to see their name at the top of a list that other people can see :P.