 |
What is a Diku?January 9th, 2009 |
I wish someone who has a good memory of these things and was there, would document that the key game design features of a DIKU are, if everyone is going to refer to MMOs as DIKU derivatives.
– a comment from Daniel Speed on Broken Toys » Wikicrap.
Glossing much here… Edit: this article is getting updated on the fly as people add comments and reminiscences.
DikuMUD was derived from AberMUD, which was similar mechanics, but had more of a scavenger hunt mentality in some ways.
At its core, it is a class-based RPG with the principal classes being fighter, healer, wizard, thief. (Later codebases added more). It was heavily based on the combat portion of Dungeons and Dragons. Advancement handled by earning experience points through combat, reaching a set amount of points, returning to town and “levelling up,” which unlocked new abilities. Classes were immutable (though eventually systems such as remorting, etc were added). Rewards for killing things also included equipment, which affected your stats and damage capability. If you reached the maximum level, common cultural practice was that you were invited to become a game admin (this practice dates back to much earlier, and existed in some form in MUD1).
Combat was generally on a fixed rate, with “faster attacks” in cruder systems consisting of actually running the same attack multiple times in a row (so you could only do damage on multiples: 1x, 2x, 3x, 4x, 5x, 6x, 7x, 8x, 9x). More advanced systems added true variable interval attacks. Tactics were centered around controlling which target the mob was attacking, and using special state-affecting attacks that did things like trigger periods of indefensibility (stun), periods of damage multipliers, etc, using stances.
Much of the gameplay consisted of moving about solo or in groups attacking monsters for XP and loot. Grouping was a typical strategy because it was a large force multiplier, permitting players to kill targets much more powerful than they were alone. Because of this, an array of systems including level limits on equipment, on grouping, and even on monster attacking were in place. A command called “consider” told you whether the monster was too easy or too hard.
Monsters spawned originally on time intervals called resets. At first, the whole zone reset at once, then resets evolved in Diku-derived codebases into per-monster timers. (A zone was a collection of world data, including rooms, monsters, and items — each zone could at first have 100 of each.). There evolved the practice of “rare spawns” and eventually “rare drops” as well.
Weapons, potions, and the like were all based on simply on performing spell effects, in the fashion today referred to as a “proc.” They were hardcoded back then, however. Players had the typical array of D&D stats, with the addition of “move points,” which were literally spent by moving from room to room and based on the weight of stuff you were carrying and your strength. You had to rest to recover these.
Death in Dikus involved losing all your gear, because everything stayed in the corpse. It also could set you back levels, as it cost you a fraction of your experience points. You respawned back in town at the central spawn point, and in later codebases at your “guildhall” (a class-specific spawn point and levelling trainer). You then had to do a “corpse run” to get back to your body naked and reequip your gear. In an inheritance from aspects of AberMuds, you could scavenge gear from the “donation room” which was a place where excess gear nobody wanted (usually from outlevelling or from trash drops) went when it was donated or “sacrificed” by other players.
This was not the only means of trying to keep the economy balanced. Every item had a cash value (the value for which it was sold in NPC shops). In some Dikus, you got a fixed limit on what you could save with your character, said limit based on a maximum cash value. If you saved your character at a point where you were in excess of this limit, you were saved with nothing. In others, you simply could not save your character state. (State saving was manual, and you had to go to an inn to do it). There were also systems whereby your save time was limited, because you had to pay “rent” at the inn where you logged out. The rent was proportional to the stuff you had, which drove people back to the game to keep earning gold.
Despite this huge sink, it was common for the economy to spiral out of control (termed “mudflation”) and for the admins to wipe all items or even all characters. It was also common for the mud to crash, and corpses and stuff on the ground did not persist, since Dikus were a character state system.
DikuMuds came with stock areas, the best known of which is Midgaard, the main city. Many mudders from the time period would judge a DikuMUD based on whether Midgaard was the first thing they saw. Midgaard came to also feature stock tutorials called “mud school” which saw you through the first fight. Zones were built by editing text files, though eventually forms of online editing tools (“OLC,” for “online creation”) were added. Because map building was relatively asy, many Dikus were based on popular hack n slash fantasy fiction such as Forgotten Realms, Fred Saberhagen, Wheel of Time, etc.
DikuMUDs did not come out of the box with any quests, because they were not a programmable game engine. They were about combat and levelling up. There was no crafting either. They did come with good chat features, grouping, etc. “Clans” were a common addition — you would call them guilds today, except that they were formed by admin command, not formed freely by players. (Honestly, I am not sure where free-form clan formation came from. I know we did it on LegendMUD, and we did it in UO, but I don’t even remember which came first!)
Believe it or not, Dikus DID have simple pets out of the box (usually summoned and non-persistent). They also had hunger and thirst (with a requirement to eat and drink regularly for good health), containers, inventory, in-game messaging and bulletin boards, chat channels (at first just via “shout”) and so on. They were quite sociable, because of the grouping requirements, the corpse mechanics, and the move rate factor. Towns tended to have fountains in them so you could drink and rest, for example, and people would gather there.
Eventually Diku games added questing engines, then scripting languages (see Worlds of Carnage), etc, and diversity developed. Because they were functional games with content out of the box, many “stock muds” were created, which had little differentiation from one another. If you knew some C you could customize the game some, and the single commonest means of doing so was to add more classes, more levels, and more player races. It got to be a common sight to see games advertising “20 classes, 30 races, and 500 levels!!!!!” without actually offering different gameplay.
It is important to realize that Dikus were the least flexible codebase at the time. The other dominant codebases were built to be programmable platforms out of the box: MUSH, MOO, and LPMuds all had significantly greater capabilities and flexibility. They were “scriptable” out of the box, had online creation support rather than requiring you to create content in flat text files, and most importantly, the core rulesets for a game were written “in the platform,” using the softcode tools that were available. Because of this, tracing the history of a given mud feature backwards will usually find that it didn’t originate on Dikus.
That said, Diku codebases did eventually popularize many of the major developments in muds. Procedural zones had been done before; in Dikus, you saw 1,000 room procedural dungeons. Instancing, public quests, player housing (in the style we know today, as opposed to UGC systems), the modern scripting system model, the modern persistence system, and aspects of zone-based PvP all were developed or hugely elaborated on DikuMUDs. In addition, once scripting hit, DikuMUDs arguably saw the flowering of quests to a level unseen in other codebases (this is of course, a matter of opinion!). Diku derivatives gained things that got going first on other codebases, like banks, auction houses, PvP systems, player-formed clans, moods systems, player government systems, overhead ASCII maps, and even tags for sound effects. Many Dikus had strict enforcement of rules regarding roleplaying, and even required players to stay in character at all times, or submit essays as applications to play. Despite this, the core remained hack n slash, with many terming Diku gameplay “roll-playing.”
If Abers set the template, Diku was the root from which a huge portion of muds sprang, because they were so easy to get running (though hard to customize). As the initial code was release as open source (though not under what we would today call an open license), many variants were made and also released, and many of these then also resulted in derivatives, etc. When I did the tally in the late 90s, Diku-derived muds accounted for around 60% of all muds running. Considering there were at least three other major codebases and traditions with radically different architectures and significantly more power and flexibility, this was quite an achievement. By the end of their dominance, the Dikus were beginning to rival other codebases such as MOOs and LPMuds in flexibility, whilst still retaining a simpler core architecture than either.
Because they were template fill-in-the-blank muds, most of them were very similar, and had to differentiate solely on their worldbuilding and fiction. However, few altered the basic combat equation. Among other terms “tanking” “nuking” and the like were common. (Thieves were sort of a nuker, in some ways — they were used to initiate combat with a “backstab” attack that did up to quadruple damage — they then they had to get out of the way! They were also used as scouts because they could move without triggering aggro). In fact, “kiting” also took place quite a lot, by leading high level aggressive mobs into low level areas.
Everquest was created by players of DikuMUDs (specifically Forgotten Realms ones — Sojourn, Toril, Duris), and even had the same wording for many server-generated messages (“it begins to rain,” which was completely superfluous for a 3d game!). It played so similarly to its inspirations that some wondered if it actually was a DikuMUD, with graphics added on. See here for the resolution of that (false) rumor. Meridian 59 had DikuMUD players on its team. UO had three Diku players on the original core team (and a couple folks from other codebases). Of the early MMORPGs, UO played the least like a Diku, whereas the line of inheritance from Diku to EQ and thence to WoW is completely undeniable.
In the end, the central elements of phase-based combat, combat states, cool-down based special attacks, tank-healer-nuker triad, and basic aggro management are what you play today in WoW. A Diku player from the late mudding period would feel completely at home if you just gave them slash commands and a text box. They’d be astonished by the number of quests, would think the crafting system was insane, and would think that the entire PvP system was either a rip from an EmlenMUD or was teleporting you to HoloMUD, in the case of the battlefields. Even raids would feel a lot like a “group of groups” tackling a level 50, albeit at a scale smaller; I recall participating in a “raidish” encounter on Worlds of Carnage against a dragon in a tower which nearly “wiped,” consisting of a half dozen maxed out characters and an assortment of lower ones. My character, who was the lowest there by far, happened to get in the final blow and was the sole survivor, and gained five levels in one fell swoop.
While it is true that many of these core elements stretch back to AD&D, the fact is that as a pen and paper game, AD&D played quite differently when in the hands of an expert group. Similarly, even though there were plenty of of hack n slash tropes in place in AberMuds and various LPMud mudlibs (a mudlib is more or less a ruleset… that’s a whole other article!), or in standalone games such as the Wizardry and Might & Magic series, there was a very specific “feel” to how Dikus played, and this remained even across all the many children codebases (the DikuMUD home page has a family tree of these).
In the end, if you play modern MMORPGs, you are playing what to a veteran feels strongly reminiscent of what they saw when they headed out of Midgaard and towards Haon Dor to go kill giant spiders, in 1991.
Hope that answers your question! Everything above is from memory, so dinos, feel free to jump in and correct me.
You can follow any responses to this entry through the RSS 2.0 feed. Responses are currently closed, but you can trackback from your own site.
) was to add ANSI color support. OMG THERE’S COLOR ON THIS TERMINAL!?!!!! Ah, the good old days!
[...] landscape. What about you with more playing experience than I? What do you want out of a non-DIKU [...]
[...] a key part of the Holy Trinity that to date every DikuMUD (muds descending from Diku code bases, as Raph Koster’s magisterial analysis describes), and the core combat mechanic hasn’t really changed that much in the intervening [...]
[...] a key part of the Holy Trinity that to date every DikuMUD (muds descending from Diku code bases, as Raph Koster’s magisterial analysis describes), and the core combat mechanic hasn’t really changed that much in the intervening [...]
[...] looks for news). I wasn’t planning on playing it; it’s nice to see alternatives to the Diku-esque model of most recent major MMOGs, but I don’t think Darkfall is for me. Apart from anything [...]
[...] business model, a persistent world to share with others, the character advancement system, and the DikuMUD-derived base gameplay. The community and social systems are a major reason that players are happy to play [...]
[...] deriva de la mecánica de juego de los DikuMUD (este artículo explica en qué consiste un MUD y este articulo de Raph Koster explica detalladamente en qué consiste esta mecánica. Véase también este [...]
[...] the “diku era” (or frankly, the WoW era at this point what with the new role of quests and all) has to [...]