Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[KittyTac]

I say completely remove the "pause" feature in DF to accommodate for multiple forts, thus the fortresses could be both played in realtime and there would be no syncing issues regarding time. Sure, you lose the ability to plan more carefully, but where's the !FUN! in that?
This is totally possible, but it will require some feature sacrifices, and a really beefy server.

More about the removal of "pause": This would mean that the game would continue running in the background while in menus, designating workshop buildings, etc. The game will become significantly harder, but that could be a charm of the mode itself. There's also the question of how embarks should be handled, but a good placeholder would to just make the dwarves appear out of nowhere for the time being.

There's so many other things that could be thought of, such as player-sent caravans and player alliances, maybe even carrier pigeons to act as a way to communicate with other players instead of a traditional chat system. Of course, all these things could just be avoided for the time being to just focus on making a working build of a multiplayer fortress.

I think the game is fundamentally made to pause when designating things, and it's hardcoded. And imagine spending years during the Big Wait to make a working build of a multiplayer fortress, only to have it broken by the new magic release. Not worth it. And player-sent caravans are not moddable even with DFHack. Playing multiple forts at once is also not possible due to the game's limitation of only loading one map tile at a time.

Dude, anything's possible with enough work, time, and lots of alcohol.

And the sources for DF, of course. Which Toady will not give away.

Another problem is that only a minority of players use DFHack, and only a minority of those want multiplayer. Really, it's easier to make a simpler DF-like game.

[Rufflikerex]

Another problem is that only a minority of players use DFHack, and only a minority of those want multiplayer. Really, it's easier to make a simpler DF-like game.

But is it possible to make DF into a multiplayer game? I'm just curious because one day I might enter a state of insanity fey mood where I attempt this. I mean, sure it'll be a big waste of time and the end product might be a big buggy mess, buuuut it'd still be something.

What I would start with is to create a DF "server" (modified base-game) that just loads a world, but only progresses through time while generating events on an in-game daily basic just like it does in normal gameplay (basically letting the world of DF run without a player inhabiting it). Then I would have separate fortresses run in the single world (with all pausing disabled), of course the server would do all the calculations for this. Then once this is possible, I would create a DF client that could control the events in the fortresses, and then test to see if the 2 fortresses could co-exist without breaking the game.

If that were possible, then multiplayer could work, although there would be little point in playing it without changing countless features to make multiplayer different than single-player DF, but it be someplace to start. Plus the server would have to be extremely buff to handle even a single world.
Man, conquering hell with your bare hands would be easier than even attempting this, but nothing is impossible!

[thefriendlyhacker]

...
But is it possible to make DF into a multiplayer game? I'm just curious because one day I might enter a state of insanity fey mood where I attempt this. I mean, sure it'll be a big waste of time and the end product might be a big buggy mess, buuuut it'd still be something.
...

Possible?  Yes.  Easy?  No.  In fact, I would call it insanely difficult.  Probably harder than writing a new DF-like game from scratch that is designed with synchronized multiplayer in mind from the start.

Here's the thing.  With enough runtime hammering at the DF executable using dfhack or another memory editing tool, you could theoretically do anything.  We are talking about using turing complete tools here, so you could turn Dwarf Fortress into the flight computer for a space shuttle if you were mad enough.  But why would you?

Now, if you wanted to do something a little less ambitious, like writing a client that lets two or more players run the same fortress at the same time, then that is much more reasonably doable.  But that isn't what you are asking.

...
And the sources for DF, of course. Which Toady will not give away.
...

If we are talking about something as mad as integrating a synchronized multiplayer environment into Dwarf Fortress, minor little things like "reverse engineer the game's code" seem pretty straight forward in comparison.  That sort of thing won't win him a friendship with the big toad, of course, but we are already in the realm of the borderline impossible and utterly nonsensical, so why not throw in a few extremely difficult and ethically (and legally) dubious things while we are at it.

[Starver]

What *I* would do, in using DFHack to enable multiplayer, is not have a single core game/server servicing everyone else, but assign and hand out 'areas of influence' to each new playerfort-gameinstance. With full peek'n'poke of wider world data (movements of armies) all happenings within a given area of influence for a player would be allowed to happen exactly as a single-player game would, but the data read about these things would be divested outwards to the other players to ensure their wider-world statuses (beyond their own AOI) is tightly synched to the one allowed to generate it. Any beyond-AOI world-happenings calculated by a player-instance would be moot.  (Is there a way to neuter the attempt to move armies around, maybe?)

Departing from one AOI and arriving at another would require an Air-Traffic-Control-like handover sequence, including enforcing chronologies between two versions. I think DFHack has pause-control so it could force a game-pause on a game that is too far advanced (to send/receive a moving world-event across with all others involved), as the only half-way workable say to do it. This of course includes playerVplayer (or playerHELPSplayer) interactions like raiding and trading parties.

The beauty of this system (if it can be made to work) is that it would allow new joiners and retirees (just force a pause, reassess AOIs, then resolve all synchs and desynchronisations) and possibly even fudge unplanned link-drops from players, using the last-known-state (and possible span of chronology) to hand over the lost AOI to other player(s). If that player rejoins, handback of AOI may include pushing assumptions the live net of players had made about the progression of externalities (around the troublesome playersite) back upon the player. Probably rewinding their clock a bit, which might also push DFHack's API abilities somewhat.


That's how I would try to do it, based on my somewhat limited knowledge of the tools actually available, but I still think it's a fool's errand to actually attempt without a ground-up rewrite of the actual DF core, which is even less likely to happen.

[mikekchar]

I have to agree with thefriendlyhacker.  It would be dramatically easier to rewrite DF from scratch than to try to mod it to do this.  It might even be easier to rewrite it from scratch *even if you had access to DF source code*.

[Bumber]

Rewriting from scratch misses that point that it's not really DF anymore. You'd be sorely lacking a whole bunch of features.

[Shonai_Dweller]

Rewriting from scratch misses that point that it's not really DF anymore. You'd be sorely lacking a whole bunch of features.

I would hope that anyone with the time and energy to recreate the entire game (in its current incomplete form) but with multiplayer would realize at the design stage how idiotic they were being and make their own game. Still, there's some scarily obsessive people out there...

Well, I guess suing them the moment Dwarf Fortress (but with multiplayer option) is released would give the brothers some extra income, so not all wasted effort.

[mikekchar]

The conclusions being drawn are correct ;-)  In other words, while not "impossible" it is completely unreasonable to ever do this.

[Shonai_Dweller]

The conclusions being drawn are correct ;-)  In other words, while not "impossible" it is completely unreasonable to ever do this.

I would add "in the current version" to that. Improved modding support is a long-term plan, as is multiple sites loaded (or at least visible) at once. One day the simulation might be robust and flexible enough to manage it. Maybe.

[Sylverone]

How possible would it be to just create zones within which items and entities would be read, transmitted, despawned, and then spawned in your friend's receiving zone? Not exactly multiplayer, but both people could use the same world if they wanted, to make it somewhat parallel, and trade things and needed workers. Separate means could be used for viewing one another's screens if desired.

Or heck, I just learned about Parsec today, but if you wanted to just pal around one fort with multiple people, couldn't you use that? https://parsecgaming.com/local-co-op-online

[Fleeting Frames]

Hm. Are you speaking of spawning/moving things at specific location in different, retired fortress so that they can pick up the things transferred when unretiring? I don't know, but copies oughta be possible at the very least. Maybe could stick the items into the histfig unit inventory of the noble there and teleport them to receive zone on unretire? Maybe someone who has messed around with adv-mode can tell you better.

Glancing at it, seems parsecgaming would work as just another way to screen-share.

[Dozebôm Lolumzalěs]

How possible would it be to just create zones within which items and entities would be read, transmitted, despawned, and then spawned in your friend's receiving zone? Not exactly multiplayer, but both people could use the same world if they wanted, to make it somewhat parallel, and trade things and needed workers. Separate means could be used for viewing one another's screens if desired.

What do you mean by "same world"? If it's asynchronized, then it's more like "parallel playthroughs". If it's synchronized, then that's a very hard problem and might take as long as DF itself.

Assuming both players are using vanilla raws and the same version, this sounds possible but will probably have negative side effects. Any links to history (tracked kills, being a histfig, being an artifact, etc.) will be severed, and it will probably be difficult to restore them if they return. If it's asynchronized-parallel, it might seem possible, but histfigs and artifacts are created over time so you will still need to sever links to histfigs that don't exist in the other world. And what if somebody made an artifact in Aleph but not in Bet? Should it still be linked to them? This shouldn't be an issue if it's synchronized, but at that point you might as well just implement sending trade caravans. That's minimally difficult compared to implementing synchronized multiplayer.

[Particleman]

I cannot for the life of me fathom why so many people want multiplayer DF. Can someone please explain the appeal of this idea to me?

[Starver]

Cooperativeness - taking on a task to support others (or maybe show off to them, how well/not so badly you can do a boring/troublesome thing) who are arranging other matters of their own assignments. At levels of 'Grouping' your PCs in Adventure mode adventures, administering different aspects of a fort or even developing parallel sites with an eye to regional interactions ('intelligent' trade agreements between ore-rich sites and food-rich ones, maybe).

Competition - PvP on any of those levels. Personal combat (and/or hide-and-seek), land-grabbing a site, setting up rival settlements or even empires.

Or a hybrid. Stab your trade-partner in the back. Enable your competitor to gain the footing needed for the battle you intend to have.


Games with these sort of capabilities (or a particular subset you seek, at least) are common enough. The complexity of DF isn't often a part of them. I recall a Minecraft server being devoted to an adhoc 'pretend you are Urists' cooperative, a while back (MC was in Beta, but was actually ahead of DF w.r.t. minecarts/tracks) where you someone might devote themselves to fishing whilst someone else was digging out.


So it probably appeals to some. I never even really wanted to multiplay Minecraft. I stuck to making my own grand projects in solo-playing survival mode, wasting many an hour for nothing but a few screenshots, but if you build it I'm sure they would play it. If you build it right, anyway, which is the problem right now.


(edited for armokawful proofreading of what I intended to write and what I actually did)

[Cathar]

I cannot for the life of me fathom why so many people want multiplayer DF. Can someone please explain the appeal of this idea to me?

Can't speak for other, but to me the very idea of showing off my world and going on a virtual hike with friends is my definition of computer fun. Doing that alone or with NPCs is fun, but I'd like friends to be part of the trip. It's not a "must have at all price", it's a "wouldn't it be great if we had it".

Online sandboxes like ark, rust or MC are usually kind of shit in my taste because the world doesn't take care of itself, and relies entierely on player input to have its story progress, which in turn means the world has to be open to a lot of players, among which a non negligeable offset will be dicks and ruin my fun (I mean just have a look on Ark official servers).

But DF does take care of itself, generates stories and worlds to explore so it is accessible to one single player, or a small group of players. That's the appeal to me at least