Szóljatok, fanfárok - itt az RL blog első exkluzív interjúja!
Jeff Lait volt olyan kedves, és a Temple of the Roguelike után velünk is megosztotta gondolatait.
Abban a roppant kényelmes helyzetben vagyok, hogy az alanyt nem kell bemutatnom. (Aki valamennyire is követi, mi történik RL-körökben, legalább a nevével biztosan találkozott, és egyébként is minden ott van a fent említett - itt olvasható - velős életútinterjúban.)
Így elöljáróban csak mosakodnom kell. Az interjú több levélváltással, az idei 7DRL Challenge-től az IRDC-ig tartó periódusban, még véletlenül sem lineárisan készült (azért mondom, mert le se tagadhatnánk). Ennek ellenére szinte végig egyetlen, jól körülhatárolható (ős-)témán rugózunk: 'WTF is rougelike?'.
Impact of 7DRLs..
lvl.up: You managed to play and review almost all completed games of this year's 7DRLPC contest. I guess you had a good chance to develop a fine 'high level view' of the State of the Art - tendencies, exceptional new talent, fall-backs (if any). Do you think the Challenge play a highly significant role in shaping the present and future of the genre? (Which inevitably takes us to the next question: 'Quo vadis, rogue-like scene?')
Jeff Lait: Playing the 7DRLs swiftly makes it clear why we don't try to rank the completed games against each other. The state of the art for 7DRLs has moved forward since we first started the contests, I think. The two metrics I'd use are polish and diversity. On both grounds I've been seeing steady improvements year over year.
There is an unfortunate tendency in game genres towards 'grognard capture'. Roguelikes, already very much in the niche, really suffered from this back in 2004 when the idea of the 7DRL took off. The goal of every roguelike designer was Nethack++ or ++Angband. The goal was always to complexify the existing games, games, which it should be noted, have very selective audiences. I personally fell victim to this same gravitational pull with my own primary work, POWDER. My early design was largely based on Nethack. I had the explicit goal of not just recreating it, but to taking it to an even greater extreme.
The main appeal of 7DRLs, as a roguelike designer, is that the games finish. Traditional roguelikes do not have a completion date. They require your attention for years after you write them - POWDER has taken my time for over five years now. This level of commitment makes it impossible to try out experimental ideas. The large time investment almost forces you to trod the same tried-and-true paths of hack and slash.
It is hard to talk about 'movement' or 'direction' in something like the 7DRL, but I would say that I think comparing the initial 7DRLs with the latest shows that the developers have begun to realize the creative possibilities that the roguelike genre offers.
Krice, somewhat accurately, refers to 7DRLs as 'popcorn roguelikes'. His point, I think, is that they tend not to have the same play time or depth as their larger brethren. The limited development time means that they focus on novel gameplay rather than deep, balanced, content. There is a risk that a diet of these light-weight roguelikes might take the wind from the sails of more serious efforts. Audiences used to seeing something new and flashy in the first few minutes may not be willing to give the latest variant of Angband sufficient time to appreciate its more subtle nuances.
I do not think we are facing this problem, however. I think the 7DRLs are more of a rising tide that floats all boats. While the 7DRLs themselves might not be able to sustain long interest (unless they are further developed, as quite a few have been), they do leave new audiences with a thirst for roguelikes that feeds into the traditional scene. They also, I think, help many of the idle speculation of rec.games.roguelike.development to actually be specced out into real code and tried on for size.
So where is the roguelike scene going?
One interesting direction, I think pioneered with DoomRL, is the actual successful integration of ranged combat as the standard form of combat. Historically, half the reason everyone concentrated on melee combat was that the interface to engage in ranged combat was too cumbersome. This has seen slow improvement with quivers, auto-target, etc, but in my (limited) experience DoomRL turned things on the head by starting from the ground up with ranged combat rather than trying to balance it in as an afterthought. The result is a sort of immediacy to the game that isn't present in most roguelikes. The 7DRLs saw this theme developed through AliensRL and, most recently Starcraft: Rogue Mercenaries.
Another is an attempt to tackle scale. Zangband, I think, was the first to go really overboard with overworld maps. However, the result has always been dissatisfying in my opinion. Usually it was still based on screen-by-screen division, and often the resulting world quickly betrayed the random nature of its source and became tiresome rather than epic. Two 7DRLs come to mind as interesting 'epic' scale examples. The first is Urban Warfare: The Escape. The second is Chrysalis. Both, I think, succeeded with their large play spaces because they didn't try to fill it with content. Too dense, and the
player will quickly see the procedural roots of the system. By leaving most of it as ignorable scenery, they could let you concentrate on the journey. I think it is an important lesson for authors of procedural content, like roguelike worlds, to remember. For your content to not look like just 'noise', it has to be presented at the right scale.
Urban Warfare did this by scaling the unit of tactical detail, the buildings, large enough to keep the number of tactical hot-spots limited. Imagine the same game with everything being random rubble every two squares - it would be frustrating and boring very quickly.
Chrysalis had basically three levels of terrain. At the tile-level was a random texture of grass, trees, or other colour variation. But, at a much larger, multi screen level, was a voronoi patterning of different terrains - grass turning to desert or to forests. The high frequency noise ensured you felt like you were
moving, the low frequency terrains let you know you were getting somewhere. Finally, at totally separate level of detail, was the enemy bases. These more closely followed the Urban Warfare theory of making the features match the active engagement size.
Finally, another series of 7DRLs are what I think of as the Frankenstein genre, despite no one actually implementing it in that manner yet. Of these I'd count the 10DRL Scrap and the later Hive Awakening. Both have the concept of upgrading oneself with leftover parts from defeated foes.
lvl.up: As far as my scope goes, you seem to be one of the more avant-garde 7DRL designers.
Just for the record, plot-driven You Only Live Once turns in character integrity for narrative coherence; Fatherhood skips all monster-bashing, instead providing the player with three 'allies' (rather 'wards') to take care of (a sympathetic attitude perhaps less regarded by that majority of players who are inclined to see allies as disposable resources at best); SaveScummer integrates a controversial, but rather widely frowned method of cheating as a major feature.. How did these highly experimental ideas come to you?
Jeff Lait: The primary source is likely discussions on rec.games.roguelike.development
. Unfortunately, I do not know what precise discussion at what time prompted the plan, but I would put the blame in those quarters. An exception is LetterHunt which I know was suggested by Antoine.
I find it common in online discussions to find myself taking a more extreme view than is accurate to make a point. This usually prompts me to wonder if I should, instead, be arguing the opposite side instead. Two of the mentioned 7DRLs are simple antitheses of claims I have made at one time or another.
I've stated that 'plot' has no point in roguelikes. My argument is that all such text is wasted in a world you must revisit hundreds of times. I saw a choice of either permadeath, and having to read the same intro text a hundred times, or load/save, and losing the sense of finality that permadeath brings to the genre.
You Only Live Once was the solution I found to address both issues. It retains permadeath, but provides new text for every incarnation. It also steps away from character progression and replayability in favour of telling a simple
story. Some argue such barely-interactive-fiction is an inferior short story. I think games, however, can really add to a narrative by giving the reader agency in the world. I aimed for that wherever possible. My goal was for the player not to see each 'life' as another counter in the game, but an actual person. The people you meet early on are those who later become your pawns in trying to win the game.
SaveScummer, as the name implies, addresses the simple claim that save scumming ruins permadeath roguelikes. I set myself the simple challenge of making a roguelike in which save scumming was actually fun. My success is arguable - some interpret it as an excellent example of how save scumming ruins roguelikes.. :>
Fatherhood was inspired by the desire to avoid the 'teenage power fantasy' approach to roguelike design. Combat free roguelikes have often been proposed - often merely relabelling combat rather than eliminating it. I was somewhat of the view that they were not possible. My goal with Fatherhood, however, was not to go combat free. My initial vision actually included combat, and done right, it could be still included. The goal was to address an experience that I have only recently encountered myself: the titular experience of Fatherhood.
.. on all things roguelike
lvl.up: What kind of general reactions did your ideas provoke in the gaming community? (Isn't roguelike gaming an inherently conservative genre after all?)
Jeff Lait: I wouldn't consider roguelike gaming a conservative community. Yes, you do see a strong anti-graphics bent, a stubborn refusal to classify Diablo properly, and a tendency to clamour for 'hjkl' keybindings. But these are more roguelike features than conservative features.
lvl.up: Neither would I! (At least I hope not to..) Still, even the most open-minded people might stick to a dated set of principles - perhaps only because these dated principles define something even more dated they are so fond of.
I mean, even if members of the community are immersed into a wide variety of genres, and welcome all technical and gameplay innovations met there, this very genre (of roguelikes) is apparently quite resistant to changes. Yes, we have specimens armed with graphical interface, mouse support, musical score, LAN and MMO varieties, one or two even have full 3D implemented - still, the bulk of roguelikes seem to be mainstreamed by and along the tradition of hacklikes and *bands. To avoid toting big guns of definition theory in any offensive manner, I'd like to cite one popular source, Wikipedia along an entry from RogueTemple posted for those "very selective audiences" you mentioned. I see two things in common: both of these listings are highly valid - and (almost) every single line could have been put down ages ago. To keep citing you, fresh and/or experimental games "leave new audiences with a thirst for roguelikes that feeds into the traditional scene", which I am (perhaps falsely) inclined to interpret along the lines that roguelike as a genre is set to some kind of a tradition-bound spiral there. (E.g. "You have not played roguelikes until you have tackled Moria or NetHack!") As if the essence of 'roguelikeness' had been concocted so perfectly by preceding generations that any major intervention would reject a project from the family - or, on the other hand, these lines meet to shape a demand so effectively that they can (and will) remain so till the End of All Things (Roguelike). That's why I had the feeling roguelikes ('roguelikeness', if you will) seem to have a certain 'conservative' tint to them. Of course, similar calls for definition usually point towards that a tacit consent had been lost along the way - but again, doesn't these definitions hold fast despite all changes?
[Nota bene: Yes, this argument still comes back to calling the community names like 'conservative' in this respect, by the merit of stating that strait-jacket definitions and imposed limitations were and are succesful at satisfying a strong and valid demand for the majority of the respective developer/gamer community..]
Jeff Lait: I think I understand what you mean by conservative, and in this respect, I must agree. However, it really is a tautology then to call an established genre 'conservative'. Canon formation for roguelikes is complete. We thus should be unsurprised there are not any radical changes. Indeed, a sufficiently radical change should instead form its own genre rather than remain in the 'roguelike' definition branch.
I had a long running argument with Amy Wang on rec.games.roguelike.developmen
t on this topic. She accused the roguelike community of excessive conservativeness (which was even more true in the pre-7DRL days), and I argued that to be 'roguelike' necessitates being 'like-rogue' and so it should not be so surprising all the games are so similar. The best example of this is likely this thread.
There is a fair bit of flame-sniping, but it stays remarkably on
topic and on a quick re-read I think I still standby what I said
'Dated', to me, is a loaded word to use. The definition of roguelike is old - but why is being old considered a negative thing?
Game designers need a lexicon of terms to describe types of games: RTS, FPS, and even roguelike should fall into their vocabulary. Having a fixed definition for 'platformer' should not hinder the designer - there is no reason why any new game need constrain itself to one particular genre, or one particular definition. The purpose of a definition for roguelikes should, like a good grammar, be
descriptive, not proscriptive. Having strong clean definitions for game types like shm'ups and roguelikes should instead allow designers to be more self-conscious in their aping of previous games. Instead of including features of playstyles out of some gut-feeling that they were 'fun', they can understand how they relate and create an experience.
lvl.up: Well, history of roguelikes harkens back to the 70s. That was obviously a long time ago, with quite a number of generations passing by. How do you see the present situation? Is there an ebb or a flow? (And I'm especially curious about the population and age spread of gamers here, as their enthusiasm - constantly challenged by new and exciting developments in non-roguelike industry - fuels both community and development.) To put it in stark contrast: waves of fresh blood pouring into RL gaming day by day, or a dwindling number of aging 'IT crowd' dudes fighting a loosing battle there?
Jeff Lait: I think the roguelike community has remained relatively stable. It
might feel like it has shrunk, but I think that is primarily because of the rapid growth of the gameplaying demographic. It is always hard to tell on the internet what the churn rate is, or even what demographics the new comers have. Further, my own experience is largely limited to the USENET groups which represent a highly selective sampling of players.
I think I'll have to leave this question unanswered. It seems to be a 'Are roguelikes dead?' - I've seen that call too many times to be willing to weigh in on one side or another. I think it is safe to say that roguelikes are niche, but I see no shame in that. It would be a boring world if we all liked the same thing. It would also be horrible, as a creator, if every creator had to attract an audience of a million. The otherside of that coin is that the number of creative productions must be very small - broken into million person cohorts, there are a lot fewer timeslots to fit your own game into. On the other hand, if you have an audience of a thousand, we have room for a thousand times the audiences. This goes back to the unfortunate tendency for people to dismiss 'consumers' and call for everyone to be 'creators'. If no one consumed, being a creator is a purely personal project enjoyed by none other. While I do hold that it is important that one can justify one's project in those terms, it seems negative to not leverage that creative effort to the benefit of others.
The future of
lvl.up: Which 7DRLs would you continue working on, and how?
Jeff Lait: LetterHunt is the 7DRL that most deserves a second pass. The end game becomes too hard, too quickly, preventing one from really starting to enjoy the game for any long play period. I think the idea is good and there is a lot of good content that deserves to be exposed.
SaveScummer, I think, is complete as is.
You Only Live Once is complete, but cries out for a second chapter.
With Fatherhood, it is a bit early to tell. I wouldn't mind creating some new maps for it, but I'd also be interested in ideas of how to better address my artistic message which seems lost in the actual gameplay.
lvl.up: Last question. You are just back from the very first International Roguelike Develepment Conference. Could you share your impressions on the event and its potential impact on (the future of) RL gaming?
Jeff Lait: The IRDC was more than I had hoped. While we had planned
presentations, there was always the fear that no one would show up, or
no one would have a presentation. Instead, we had extra people show
up and present some very interesting presentation. Then, there is the
fear that we'd all sit around the table and stare blankly at one
another. Instead, the sun would set and growing hunger pains would be
what would drive the agenda.
I think a face-to-face meeting definitely has some advantages over
Usenet. It seems easier to generate consensus, for one. I felt we
got a lot accomplished in a short time on the Roguelike Question.
One major takeaway I had was that there is something called
'Roguelike Studies', and there is more than one person (myself)
interested in such a thing.
lvl.up: Thank you very much.