Classic Game Postmortem: Infocom's Zork

hi after having paid the Av tax for a while I'm ready to go my name is Dave leveling' and I was one of the original implementers as we called them of the Zork game on mainframe I was one of the founders of infocomm which took sort commercial a few years after we finished it as a hack project and I worked for infocomm for a long time - and wrote I believe nine games or co-wrote nine games so I'm here to talk about Zork though mostly which is I think for me the most fun part of our time at infocomm and before that so here I am now for those of you who don't know what a text adventure is is there anyone who doesn't know what a text adventure is yeah come on come on admit it admit it okay a text adventure is a game you play entirely by telling the computer what you do and the computer tells you what happens it's kind of like playing D&D except with somebody a DDM who's really stupid but but a DM who's totally without any qualms about killing you all wait a minute that's a real DM isn't it anyway so here are the people who wrote the mainframes or the originals work we were all at MIT I was a research staff member mark and Tim and Bruce Daniels were grad students and hangers-on and things like that and from the ways work started is really tied a little bit into its time which was the late 70s and only some of these were kind of like today in some ways we have Downton Abbey they had upstairs downstairs we have holders of slave they have roots so forth and so on they star wars both time Star Wars came out during the month in which we started working on Zork the original Star Wars now where did this happen this happened at MIT at the laboratory for computer science sometimes called Project Mac it was as you can see a rather sort of post brutalist kind of architecture but we had a nice set of offices on the second floor and on the ninth floor were all our computers the two gentlemen in the middle there are JC our lick lick lighter and Alva both of whom went on to be on the infocomm board and now actually work for infocomm as CEO for a while so they were respectively a former DARPA program manager that was lick who had foreseen the internet way back in the early 1950s and alves uh who was a huge proponent of the ARPANET when it actually became a real thing and he was our group leader now those are the tools we had a couple of micro No a couple of mainframes we had a pdp-10 that actually is the pdp-10 that was a Stanford back then and they had it was enormous it had two megabytes of memory it was enormous and disk space huge amounts of disk space we I think we had it was either three or four our p o-- fours with 92 megabytes each and one or two RPO sixes 178 megabytes we had we had a KERS of of computing oh and the thing down on the lower right is a terminal it actually is about the size of a desktop PC but it was our terminal to talk to the mainframes and all of Zork and many of the infocomm games later were developed on those terminals now this was the internet I'm sorry the ARPANET the network that's a map of of it it could not have more than 255 hosts because the the IP address was one byte and zero was reserved we were in the upper right of the map there the little dot labeled m6 I believe and offered as the pdp-10 that was our machine which was called MIT diem now the way we talk to it is through a router that's the big white thing it was a milspec router and you could drop it from ala copter from 20 feet and it wouldn't break it and in fact even had hooks so that you could pick it up with a crane and drop it just to test fit the thing on the right is another sort of router that people could dial in with a modem and that is how the user community of Zork came to be there were people all over the internet which extended notionally at least from hawaii to london but people could come in over these phone lines and then telnet to our machine you know telnet still exists and so that's that's how we got our user community once we started writing now why did we start writing sorry this is not the most legible of slides and I apologize but this is my personal map of adventure adventure was the father of us all it was originally written by will Prowler who was a guy at BBN which is now part of Raytheon and they were the they were the implementers of the internet they created that machine that you could drop from the helicopter and we'll crawl the road a bunch the software for it he also wrote the first part of adventure and then a few years later dawn woods found it and said oh this is really awesome but it's really just crawling through a cave which in fact it was it was crawling through colossal cave in Kentucky actually think the official name of it was bed-quilt cave and if you ever manage to go to that cave which some people have done they report that it is exactly like that part of the game so all the parts of adventure that described caves are real life descriptions now Don was more of a game type and so he said this is cool but I want to do fun stuff so he started putting in treasures and enemies and puzzles and magic words and all elves cheering elves cheering dwarves maybe anyway and anyway it was it made it a more fun game and it was like four hundred points or something and he wrote it in Fortran and he let it get out on the internet and for two weeks on the ARPANET and for two weeks all work on the ARPANET stopped and we were among those who stopped and so being at MIT instead of Stanford we felt that this was a wonderful game it was a wonderful wonderful game but it had some defects for one thing it had a two-word parser you could only say you know take ax to relax you know you couldn't say something like put the sandwich in the lunch box it wasn't that good and it couldn't had no adjectives it had you know it just was imperfect from our point of view and furthermore it wasn't written in a modern language from our point of view so we said we're going to do something better we have to do something better the honor of MIT is at stake not to mention it would be fun so we had some experience with games we had created a game called maze that was a really the first first-person shooter in some sense you wanted through a maze you fired at your friends and killed them or you fired at monsters and well robots and killed them that was on a smaller machine called an in LAC we had a there was a game called hunt the Wampus which we we did and then there was a crowdsource trivia game which was the reason why all these people on the internet knew to come to our site to look for interesting things they had all come in to play trivia and when Zurich appeared they came in to plays org so we sat around for a while saying you know we should do this and nobody really was ready to quite bail the cap and so I said dammit we're going to do this and I'll start so I wrote a torrid parser and then I went on vacation so mark and Tim and Bruce sat down and started fleshing out what what the world should look like in such a game and they did a sample game a sort of four room game that's unfortunately now lost forever not that it was a great game and the reason there is a Bandstand there is one of the puzzles involved putting a band in a band box and when you put a band in a band box it played hail to the chief' because this was early 77 and Carter had been not inaugurated as president just a few months ago and that was one of the more exciting puzzles in the game but notice put band in band box three word parser so again this is another slide that's not super easy to read but this is Mark blanks early notes and I think some of Tim Anderson's on top of them about the beginning of Zork and you can see a few rooms there in the middle there were a few puzzles patrol Cyclops who because Mark had seen seven voyages of Sinbad but not read the Odyssey I guess was scared by the word Sinbad rather than Odysseus or Ulysses or whatever there was the thief that became Tim's obsession who wandered around the whole universe of Zork looking for you stealing things from you messing up the classic let's solve a maze thing where you drop an object on the floor because that'll mark the room because mazes are all like the echo room was put in and there were pits in in the dark if you walked around in the dark you would fall in the pit and so one of the very first things that happened when I came back was I started playing this this alpha of the game and so I said to myself okay let me wander around and I went up in the attic of the white house and it said it's dark I hadn't taken the lamp and it said it's dark up here you could fall in a pit and sure enough I fell in the pit and so I walked across the hall to mark and I said mark I fell in the pit in the Attic and so he said okay do something about it and so I sat down and I said okay well there's a monster that gets you in the dark then so I invented Gru's and they are I guess one of the more iconic things from infocomm I stole them from Jack Vance a science fiction writer II released the name and in French it means crane so the first time I went to Montreal I saw all these signs that said bruise for rent you know groove glues on you a really and and I thought Canada was laid-back I guess it's not so this is a kind of a sketch of the kind of things we could do with the parser this is a later syntax this is the Zil syntax for doing for parsing sentences and the idea was that the verb put could be used in enormous number of different ways you could put an object in an object you can put it on an object you can put it down which was similar to dropping it you can put it under an object you could you know you could put something out you know put a fire out so put was there a lot of places and then over on the right you see the real verb it maps into for each one and so when somebody types a sentence in it maps in this case into putting a gold key handles adjectives put in you know and then in the wooden box and that's kind of you know like two percent of what the parser does but it's kind of cute and then behind the parser was what was the main loop of the game which is read in an utterance parse it then give it to somebody else to work with so the first person that got a chance to work with it was the so-called actor now you're usually the actor that means the avatar that's currently running but sometimes it can be another creature because eventually we got to the point where you could say things like robot walk east robot take the caged robot do this and in that case the robot would be the actor and there would be special actor code that would make the robot be able to do things and if there if he did it you were done otherwise it went on to the vehicle you were in because we had vehicles most famously the inflatable boat that you use to go down the frigid River then if that doesn't handle it you give it to the so called pre verb if there is one for the verb take just checks the tape bit for the pre verb and then there's the indirect object can handle it so the wooden box and that previous slide could handle it or the direct object could handle it and then it certainly just keeps going on there that eventually it gets to the so called demons which are timer things you could have timers like in five moves the bomb goes off or in five moves match goes out or in three moves and then three moves after that and three moves after that something happens or every move the thief does something and this is someone I honestly don't know who it is out on the internet who did a complete map of Zork and this is for the mainframes work it's got everything in it the so-called Chinese puzzle in the upper right the volcano I get yeah in the low in the middle laughs the frigid River the reservoir the maze just everything you could possibly want in a mainframes org and this this sort of it was from the point where we were nearly done with it so actually let me there's there's one one thing I want to talk about which was design which was a word we didn't really know and I've got a little video here of mark blank talking about how it all happened when the game was first being written we had no yeah just like the original adventure it had no real story I mean it had no it had a place and things to do but they were not at all tied together there was no intention to tie them together it was all different pieces of fun that you could do in any in any order through very little in the way of you have to do this to do that there are feet there became more of that and as a game got flushed ad it was like oh we can hook this up to here so where you end up at the end of this puzzle is actually here that's clever sand other different ways to go but none of that was planned in terms of oh we've got the strategy and we're going to fill these areas and you go from here to there I mean it all kind of happened and he's really understanding it a little bit what really happened a lot is that someone and that was more often than not mark would come up with a cool idea for something to do and one of the first ones he came up with was vehicles he said oh you should be able to have vehicles you should have boats and cars and planes and trains you should be able to get in a vehicle and do something and that's just complicated everything because especially when you have something like a portable vehicle like the inflatable blow boat you could take it somewhere there wasn't intended to go and you could try to use it and you just all kinds of unintended consequences would jump out and so we would run around fixing all the unintended consequences and then mark say oh you really ought to be able to talk to other characters and do something like that you don't order them around that would be fun - and so he created robot that you could order around and it turned out that that had complications elsewhere in the game because there were other characters like the thief didn't want to be able to order the thief around that that that would be fatal so just every time we turned around someone would come up with an idea I came I came up with sort of having D&D fighting in it and that had its impacts and just everything that we came up with made more work for us and after a while we started going oh my god this is really difficult now this is the final this is my final Zork map and it's too big to show really on one screen so I'm going to try to scroll it down here which is a little bit slow but it might might work oops goes back nevermind there we go okay I guess it got all the way and one of the things you'll notice about it that I found kind of interesting when I was looking at this again is that it's all very flat even though the game was up and down you could actually map it flat and when I think of later games like well EverQuest or the you know the various shoot-'em-ups which have many vertical levels that actually interact with each other at different points that was something we didn't do very much but this is a interesting flat environment but with lots of stuff in it it was close to I think was like 200 or 250 rooms by the time we were done so about the same time but somewhat unrelated infocom under to Zurich we were founding infocomm and this is five members of the board of directors of infocomm Marc Blanc is on the Left Joe Barresi later became president Allen lick and Chris Reeve who was the guy who wrote the muddle compiler muddle was the language we wrote Zork in the modern language and he wrote the compiler and he wrote and maintained the interpreter and was just you know the guy for all of that hairy stuff and when we started infocomm we thought what should we do and the idea of doing something with Zork kept coming up and we you know it was like yeah we should do sort Sark would be cool and then we would say but it's too big it's a mag it's a hold meg how can we possibly fit it on a micro because here's what was it there for micros today then rather trs-80 model 132 cave memory Apple 232 K of memory about a hundred K disks more or less so that doesn't divide well into a meg for one game so we we kind of put it on the back burner for a while and in the meantime Joel had gone out to Pittsburgh to work in his family's business and Mark had finished medical school and was starting an internship in Pittsburgh so they were both in Pittsburgh and since both of them were bored by what they were doing I spent a lot of time going to lunch and having Chinese food and talking about what they could do with Zork to put it on a micro and so I'm going to show another little video here about how you have fun in Pittsburgh xur came ran on the mainframe and it was well over a megabyte and the machines we were looking at besides Joel's pdp-11 very few people had pdp-11 us at home Sheldon but you know the other machines are out there then were Apple twos and trs-80 model ones and those are machines that had floppy disks with a DK on them and maybe 32k of memory so we were thinking about how this might be done and we actually came up with an approach that was based loosely on what the researchers at UCSD did I had been playing with few CSD Pascal which was an implementation that used threaded code where reasonably high-level instructions were encoded down to one or two bytes so that you could have a fairly complex program that actually took up very little space they invented a language that was portable it was machine independent by basically coming up with machine independent instructions that would then be interpreted on different machines to actually execute the code this turned out to be a fabulous kind of idea and nowadays the concept is used all over the place you know Java being a good example of a language that's done exactly that way and for similar reasons but it was extremely slow and painful because again rather than running on these slow machines to begin with and I'm talking about on on a bigger machines on on PDP 11s and PDP tens I mean the problem isn't you know now you had not only these slowish machines you're dealing with but now you're interpreting a language on top of it so you've added another layer of performance hit on that so so using that approach of just using UCSB Pascal we were thinking that's just going to be horrible but we had the idea that if we wrote our own language our own virtual machine language where the instructions of the language themselves of the language itself were very tied to the game in order to we were running a general-purpose language we're writing something to run Zork and so an instruction in this language rather than add two numbers although we did have one like that because we have to keep track of scores for instance might be move an object from this place to that place and that would be literally three bytes in the game which is very very small there'd be a byte you know eight bits that would say it's the move instruction and there'd be a byte that said it's this object and a byte that said it's going to that place by encoding the code like this you could actually just create a fairly small interpreter that would take each of our own codes interpret for whatever machine it happen to be running on so that suddenly solved the problem of deciding whether we're going to support the Apple 2 or the trs-80 mono one first or or in fact the pdp-11 we could fairly easily do all of them at the same time so a little tiny little bit more detail about that you could have 255 objects which included all the rooms as objects you could have 230 something global variables things like the score that market mentioned you could have 30 bit flags things like is this lit room or an unlit room you could have 31 properties things that could be changed things like weight and the descriptor sentence that you you gave when the object was first encountered and you could have 64k of static code that it was static because a 32-bit machine and a game that was 64k of code and another 20k of non non code you know volatile stuff like string and not strings but things like variables and stacks and the interpreter 100 hundred K disk all of that couldn't possibly fit so we did dynamic paging as needed every time a routine was called it said is it in memory and if it wasn't it would pull it in off the disk which was kind of cool on the Apple 2 because you'd be playing on the Apple 2 and if everything was in memory that you were doing it was you know very quiet the disk wasn't doing anything then you would do something and the disk would start going and that could either mean you were in really good shape or you were about to die so the next step after this was for them to convince the rest of infocomm that this is what we wanted to do this was not terribly difficult since between the two of them they owned 40% of infocomm and once you add in me and Tim and a few other people it was easily 50 percent but it was not even that hard people thought hey this could actually work and so we set about doing it and the first thing we did is we started to create a tool chain and originally we had the model language with some additional macros and we called that Zil that was the language that we wrote in and on the other end there was assembly code that would be run by those interpreters that Joel was talking about you'll see that in the second slot there there's a thing called a compiler we didn't have that initially but if you look at zou code into assembly code here on the two sides you can see they look moderately similar but the assembly code looks like assembly code and we didn't want to write assembly code assembly codes terrible it's disgusting and so we because we were so incorrigibly lazy we decided we had to have a compiler and so mark wrote part of it I wrote part of a Joel wrote part of it suddenly there was a compiler and it meant that you could write in the Zil and not even have to look at the assembly code except just throw it at the assembler and then run it which was great we've been much faster turnaround for for bug fixes and stuff not to mention we didn't have to learn a new language and this is actually this is examples of muddled versus little objects because the original mainframes work was written in muddled and that's the stuff on the top and zil is on the bottom the major thing you see that I just wanted to point out is there's a little cleaner looking that's about that's about it but it also looks a little like XML or maybe I don't know it it wasn't so we said we want to publish this well we don't want publishing ourselves that's terrible that would be no fun at all we did publish the pdp-11 version which I think sold 20 copies 30 copies I don't know some small number nominee had pdp-11 s as Mark said but we wanted a real publisher and personal software looked like a great bet because we knew some of the guys there they liked Zork the mainframes ork and we they had had a big success releasing a thing called visicalc which was the first microcomputer based spreadsheet we thought and they had also some games they had a version of adventure out so we thought they were great they would be great so we went to them we brought it brought an apple to or maybe maybe trs-80 version over there offices and they tested it because they knew how to play it and they said hey this is just like Zurich that's awesome and we published it so this is what they did for the cover and we said you know that doesn't look like us that cover doesn't look like us for one thing we never ever in the game described who you are what you look like what sex you are what race you are whether you're fat or thin rich or poor you know I'll go scruffy I believed was occasionally used but we never did anything more here we've got Conan here with a with a sword and a trol and all that kind of stuff it didn't quite do it for us but we didn't want to publish and so they put out the trs-80 and Apple two versions and sold so it's sold reasonably well but then this account took them out in here actually here are some screenshots but what happened to them is visit Cal took them over completely they eventually change the name of the company to visit Corp and when the time came to do the next game they weren't that interested here's the sort users group a map of zorp one Mike Dorn Brook who later came to work for infocomm and was our first tester of the zorf one game said you know the first thing you do in these games is you draw a map maybe I should give them maps and you know maybe they pay for them and maybe I should give them hints and maybe they would pay for that too so this is the zorp users group zorp one map and you can if if you were a careful student of one of the previous slides you would see that what's in this is a big chunk out of the middle of the mainframe game it was of live 40 to 50% of the mainframe game we couldn't fit it all which left us room to do other games and it became short one I think I was the one that actually sat there with the map it probably it was a copy of my map and I just sort of been a big circle and said okay I'll be our point and it turned out to be pretty close and this is again not a terribly readable slide but this is an in visit clues hint book which is another thing Mike did where you had a developer marker that would answer your questions about puzzles until finally the last answer was like the complete spoiler thing that would do it the subtle hints less subtle hints and then an abject spoiler so I was talking at the fast-forward somebody said were you going to talk about what you did wrong and what she did right and I said oh really I guess I said a little bit but this is kind of what we did right and wrong with Zorich one the parser was right the z-machine was right having a tool chain was right the size magically came out just right and some of the puzzles and this is personal opinion they're great there was the thief there was a puzzle where you had to bring light into a dark room that was that I thought was really good there was flood control dam number three which was one of my favorites and the famous inflatable boat which had the great feature that here you've got this pile of plastic can you recognize eventually if you're lucky that it's an inflatable boat and you blow it up and you put things in it and when you're done putting things in it and sailing down the river you dis inflate it and suddenly it becomes a bag of holding from dee dee because the inflatable boat and the uninsulated boat were different objects so all the objects that you put into the in the inflated boat were not part of your weight that you did weren't encumbered any of that silly D&D stuff and you can walk around with the uninflated boat it go anywhere and then when you reinflates were back so it was a bag of holding I thought that was pretty cool what we did sort of wrong is we didn't this was a midnight programming project we didn't design anything you know we just had fun pure dungeon crawl yeah pure dungeon call crawls are great some people they're kind of passe today but but it was fun there were too many puzzles that were magic words and I kind of blame adventure because it had a bunch of those too there were there were puzzles that you just had to say the right magic word like to scare the cyclops away you had to say Ulysses and the Cyclops would run away now we mitigated that later by having alternate solutions to some of those more silly puzzles and we killed you a lot I mean we were bad we killed you all altogether too much Graham Nelson I think it was who has written a lot on adventure gaming said Zork hates its player and there were ways to get stuck you know the thing you use something that you need later and you're you go somewhere and you can't get back to the other place we were learning how to write adventure games and it took a long time to learn to do them better so then came infocomm as a publisher because personal software got sucked into VisiCalc and they said yeah games there now we're going to we're going to we're going to do busy stuff busy databases busy editors busy party planners visi everything and it even change their name to visit Corp and they society we don't want to do another game we didn't want to do Zork to forget it we're not interested and so we said well I was going to do it well we're going to do it so the very first thing we did is we found an ad agency that shared our sensibilities artistically I guess you could say and we raised some more money to pay for being a publisher and pay for them and the very first thing they did was redesign our logo to the logo that's that was what we had until I guess Activision changed it after they you know in the dim and Misty horrible feature the company was called G Rohini Russell and these are some of their sketches for possible logos not as good as the one they finally came up with but I liked it and this was the original package you think wow that's not very interesting by today's standards and it isn't but in those days most games were sold either in baggies or sometimes in boxes small you know boxes that just basically held the manual and you lifted the top off or you put it back on they were not very exciting and the the stores didn't like them too much so this was lovely blister pack you couldn't get in it without making it obvious it was beautiful it was the look of zorkin infocom forever after that and we redid zorp one well actually we redid Zork one from the Barbarian to this and loved it much more than the barbarian and we did sort two in the same style and Mike Dorn Brook reminded me that we actually put them in a zorb lands work to counter display that could be in a in a store at the same time so zort too was in theory everything else left over from Zork one that was that was the goal and you can see if you look in the the far left all the things with stars were taken from the mainframe game but hadn't fit in one as it turned out a couple of them ended up getting pushed yet again to Zork three but that was pretty much it and these are my design notes for writings or two I mostly works or two during that period and Mark worked deadline which was our first mystery game but he helped a bit on zorp too and came up with something new the idea was ok Zork one was totally a dungeon crawl let's do something a little bit more interesting so there was this character called The Wizard of Oz and it became the goal of the game not to crawl the dungeon and collect all the treasures but to find a way to stop him from annoying you and in some cases you could actually kill you if you weren't careful and get rid of him so the dungeon crawl part of the game became a subset of defeating the wizard which was in some sense an advance in the world of of narrative and story a little advance but in advance more design stuff there's this there's the the infamous Diamond maze which I created which required specific baseball knowledge that most Americans don't even have and people who aren't Americans kind of go what I was going on so that was that was one of my triumphs and puzzle design another exciting puzzle and Zork to that mark road we were trying to come up with a different type of puzzle than find the key to the door which is what most of the work whether it was a key there was a button or was something you'd find something in one place and he'd take it someplace else and you'd use it so they're very mechanical I think for the Bank of Zork the the key was that the was that the puzzle was about the way you would move had something to do with what would happen afterward not something you had so the ideas who are trying to come up with her I was trying to come up with a puzzle were to figure it out it wasn't so much what you were carrying or what you were doing but how you would move through things would affect what happened when you went through the the shimmering wall depending on where you came from and they think that puzzle was I think it was hard because it wasn't obvious that was going on and and I guess one of the things that was hard for me writing it was wanting to describe it in a way that would lead you maybe to think about that kind of solution without giving it away so you'll notice that even mark cannot explain how that puzzle actually worked I remember the first time I tried it I eventually stumbled on the answer and I came out to him and I said mark what happened you have to end up in a vault and it involves going through as he said kind of the right sequence of rooms to get in it remembered where you had been and that determined where you went the next time you did something and so he explained it to me and he draw drew a little diagram and I said oh ok ok and then someone else tried it and asked me because I had played it and I said I'm not sure I can explain it and I tried and they said oh really and so we went to mark again and he explained it again oh really I don't think anyone even mark ever has successfully explained how it actually works but it was a foot it was a fun puzzle just really hard so here's some screenshots from zoar to the ones on the left are from I think from Apple twos and the one on the right is from an Amiga so much much later machine that didn't even exist at the time but they give a sort of flavor of what is going on they the worst worst thing is that wizard down at the bottom left you can see him casting the word the spell fantasize on you all of his spells began with F's and the one be obvious one you're thinking of wasn't in the game and fantasise i think was like an acid trip or something here's the Zork users group for to map all of these maps and hint books were illustrated by a friend of ours a guy an artist named Dave our dido who did some really fun work with them you can see zort - has has a lot of stuff in it on the upper left you've got the Bank of Zurich on the bottom right you've got the baseball mais on the bottom Center you've got what we call Dallas land which was a sort of Alice in Wonderland type scenario and those were all from the mainframe game and then in the bottom left you've got the wizards workshop which was new you've got a glacier puzzle that was new there were a lot of things that were added you've got a very nice puzzle with a topiary garden and a unicorn and a princess and all of that was was new so as it was you know about half old half new stuff and so again the right-wrong thing you know Bank of Zork was too hard baseball maze was a bad puzzle we have adventure game physics you could inflate an entire hot-air balloon by burning a newspaper but you know that's what adventure games are for bad physics things that were good we actually introduced magic we had the wizard you know we were self-publishing we and we were having so much fun self-publishing because we got input into the packaging no barbarians you know get rid of them and then we moved to 55 Miller Street which was the first time we had actually been anywhere in our other than in our own homes and wheeler Street brought people together working at Miller Street is great because we were actually together rather than together sometimes or together at someone's house or something like that we can actually work together and say hey come over here look at this I don't understand why this isn't working or live an idea and actually write it let the other person walk over to your your office and see it and so that made a big big difference I think in terms of productivity well certainly in terms of productivity but I think also in terms of quality of the games because there's much quicker feedback whether it's a testers or with collaborators when I work from the fav I think it's hard to do any creative project entirely by yourself of this sort because because it's hard to get the feedback I think deadline which was for the most part written at my my apartment and Medford I think would have been a better game if if there'd been more opportunity to have other people there while I was writing it and actually claim within saying it at the time so I think you missed something working that way so having the office I think was good it also gave ever anything more a sense of you know permanence is this was a real thing that we could keep taking seriously I mean it wasn't just like a lark that we were doing bit and we were trying to build something there's the sword 3-car you'll notice the sameness of one of the threes or covers aside from the three and the little tagline on the top it's called establishing a brand I think and again some sheets of marks design of sort three you'll notice our designs were very simple it was just handwritten online printer paper or you know sheets of note paper or whatever and you sort of show them around and people would read them sorry that seems pretty good why don't you do that so we were actually doing a little bit of design and Zork too and work three because we had new puzzles so work three however had another thing that was different instead of just design he had a different idea of what you want to do in a game it's not a dungeon crawl you know who I pick up the microphone 50 points I pick up the gold bar 100 points I kill neutral eight points it was more about how you exhibit the characteristics that make you not only a good person but in the end a fitting replacement for the dungeon master because instead of an antagonist the true antagonist like The Wizard of Oz Zork to the the antagonist is sort of is the dungeon master the guy who is in theory running the great underground empire is responsible for well really all the bad things would happen to you in any case he's in charge of the place and so by the end of Zork 3 you're you in theory should be fit to replace him and that was what Mark was trying to write so it's not killing things it's cooperation it's trusting people it's doing things that make you a better person as opposed to a better account in the Barbarian type again some more screenshots mostly you'll see this was over time all that really seemed to evolve a little bit was the look of the the scoreline and that's probably because it's from an Amiga anyway and as work three and a map it's if you count them up you'll see it's a slightly smaller layout than Zork 2 but not very much smaller and again it uses the last bits of the mainframe game the thing in the lower right is what was called the royal puzzle in Joerg 3 originally the Chinese puzzle and in the upper right is the so-called endgame where you make your way down into the realm of the dungeon master and solve the final puzzle to win the game and most of the rest was brand new stuff and right and wrong it's a story the dungeon master was pretty cool you only began to realize after a while that was actually he was the same person every time you encountered him had this innovative scoring the winning score was like 9 points previously the been an advertising thing for adventure game companies to say we have an adventure game with 300 points 400 points 500 points and Mark kind of said hey how about 9 points that sounds good so Ori there's also a great time travel puzzle now I'm sort of indict I'm travel puzzles and did a few of my own later and I think there were three or four or five infocomm games that had time travel puzzles we all liked them sort 3 was too small that's just a personal opinion because I wanted more and there was one that so-called Chinese puzzle royal puzzle was really just a blocks moving game which is a thing you can do in the real world and I personally I mean all these rights and wrongs in my personal opinions not anything real I kind of like real world real puzzles in the real world but I don't like them in games as much that's it again that's just me so and here is the complete works of infocomm in time worker so you can see Zork ones or to happen sort one was 1980 so two was 1981 and at about the same time deadline which was our first mystery came out then mark was doing his work three while I was doing Star cross which was our first science fiction game and they came out of more or less the same time and just on and on and on all the way down to 1988 and 1989 for the last three games four games I guess 93 when we as it were the Cambridge office of Activision was moved to the west coast the people were not but the office was so what I mostly what I do here is thank some people the other three implementers Tim and Mark and Bruce and Joe Burrus Mike in Mike dorm Brooke and Steve maretskiy and everyone else at infocomm but Joel in particular for being in those interviews Mike and Steve for being the original infocomm testers the most prolific writer in the case of Steve Mike became our director of marketing eventually they contributed word infocomm individually than almost any other person that you could name and so they they get a thanks and then I want to thank Rick thornquist who I described as an archivist here who is provided videos and a lot of the scans that I've shown here and he is doing a book about infocomm not sure when it will come out but he's got some great material and any questions don't have to be okay I think I think the gentleman here on the left was the first one to get to a mic is that on okay that that's a two-part question since the mic wasn't on the question was what are my thoughts on stories and games and my advice for people trying to put stories in games and second what does Abreu look like the first one has a sort of long answer and short answer the the shortest answer is that writing stories is very hard and writing games that our stories is even harder unless well what happens is you tend to get on a railroad track you know you're you're programming the player to experience the story that you have in mind whereas the player may have some different idea and the example I always use which is getting a little old I guess is what happens if in the first act of Hamlet Hamlet says I don't believe him no ghosts man my dad in there you know I forget it I'm going to go back to England and go back or Wittenberg and go back to school the things are weird here in Denmark and I don't think Shakespeare had that particular choose-your-own-adventure branch worked out actually there's a there's a great webcomic forgotten it's called dinosaur comics but the author of it created a choose-your-own-adventure for Hamlet and it had a huge Kickstarter for and Hamlet could choose what to do Oh philia you could play Hamlet you could play Ophelia you could play Claudius you could do all kinds of stuff and so I guess in some sense my my answer is incorrect but anyway the short answer is it's really hard to do that and as for what grooves look like I think grooves were actually portrayed at some point in one of the later post infocomm Activision versions of absorb in a return to Zorich or the other one however they were it was incorrect it was not what they look like if they were in there and if I knew what a grue looked like that would be kind of a logical impossibility because if I knew what a grue look like would be dead okay first I'd like to thank you for inspiring me to learn how to type and then also inspire me for my 15 year career in the games industry so thanks for that my other question is I saw a screenshot up there which had an Emacs editor yes I'm curious Emacs VI preference let's see if I build entire games while they're actually back in the day it was before Emacs it was still in TECO when Emacs was a giant mass of Kiko macros that would fill this room if it was printed out and we used a variation of that giant mass a slightly less giant mass called our mode but it was very max like and Emacs was very much like it and so I am definitely an Emacs guy absolutely yes yeah yeah VI types out I do use the eye every now and then I can say : WQ okay so um I must say for what question you say that the beginning sock it was very difficult to make because of memory and talk free was more difficult because I wasn't attention coder so what do you think what was the most difficult soccer games to make until on maybe the most fun to for you well that requires a little explanation the next Zurich Zurich for was going there was actually a scene in sort three that was ostensibly from Zurich for and it involved you being sacrificed to some Elder God and an altar and that sort for became a different game called enchanter which was the first game in our magic trilogy chatter sorcerer and spell breaker and mark and I wrote in chanter and Stephen wretzky wrote sorcerer and I wrote spell breaker and in enchanter you see that same scene so you begin to realize that that's really a continuation of the Zork series and probably our marketing people in retrospect wish we had done it that ways or four or five Zork six sort 337 we didn't and so it that's a long way of saying my favorite game of all the ones I've written was in chatter I mean I didn't write at all entirely myself mark blank of course was a big contributor as well but it had a magic system in it that was new I was sort of like the D&D magic system or the Jack Vance magic system it had a magic system it had some great puzzles it had some good writing it had some fun characters you got to meet the adventurer from the earlier resorts because you were now playing magician so that was my favorite of all the games I wrote and it was the hardest would have to have been spell breaker because that was it was again another magic game but it was incredibly complicated it was one of the last games to be in the 64 K code limit we we came up with a way of making the code limit our 28 K later and it was it was hard to fit in hard to fit it all in and had a lot of complicated stuff going on that was for me that was the toughest one to do okay thank you hi I think that the language and the humor of so many infocomm games and Zork in particular is just indelible in my mind and I never thought I'd be able to thank anyone for it so thank you and all those involved and I do have a question so after I finished playing the games being a being a young kid I would spend hours just screwing around with the with the parser basically swearing at the game and saying things like take troll and the troll would say oh how romantic or the game would say already so I was wondering is that did that emerge from you all just screwing around with the game or was it something that was it was conscious there were several reasons for it one was we were a bunch of wise asses secondly people we would we could one the mainframe and and of course at infocomm we had testers we could watch people playing we could on the mainframe you could literally watch them playing in fact you could even inject things in the game as it was going on which marked love to do move things around in the maze was like that but we could see how they were interacting and if they tried something and got a default response like you know if you said taped role and said you can't take that very boring so someone me or Tim or Marco Bruce I didn't remember who who put that one in but somebody would come in and say okay if it's take on the troll then give a better sarcastic response like there was one there was one of my favorites which we didn't end up taking out but if you tried to dig with something other than an actual shovel it would say digging with whatever it was is slow and tedious and it turned out that a thing you could dig with was had the so called tool bit you know I was talking about those bits so the shovel had the tool bit now it also had was the case that your hands had the tool bit so if you said dig with hands or just dig because it would look for things with a tool bit and say oh he must mean that it would say digging with the hands is slow and tedious and after that that became an expression around the company for doing anything that was incredibly boring so yes it was entirely deliberate if we saw these things or if our testers saw them you know once we had a actually you know we had Steve maretskiy is one of the funniest guys in the world and Mike Dorn Berg who is incredibly meticulous both noticing these things when they were our testers and then we had a full-time professional testing core later so whenever they found something like that and pointed it out if it was room to put it in the game which they're sometimes loving we would put it in thank you it was just curious you mentioned Graham Nelson earlier when curses and then later inform was released I was just curious what the you guys the implementers kind of here's a guy who reverse engineered our stuff is making a basically a brand-new infocom game of course there is magnetic scrolls and all that stuff that this was like a single like Oxford mathematician poetry editor I was just curious what kind of the response to that was well let's see starting from the end and working back wheel of magnetic scrolls we had a good relationship with there were nice people and as for Graham it was a whole team of people in reverse-engineer that he was just he wrote a compiler of his with his own idiosyncratic idea of having to write how to write a language that reverse-engineered our stuff and it was like I I think personally I was incredibly flattered that they had done this it was like what happened when a guy at Digital Equipment Corporation broke into the machine at MIT you know hacked into the machine but in today's standards and stole the source code for the mainframes work and translated into Fortran so that it could be distributed by the deck users group thing we were flattered it was not like ooh competitors or what an evil person it was like an homage or an honor to that somebody cared enough to do that I think that's how I felt anyway okay it looks kind of wait a minute somebody had enough I'm wondering what you think about that today the computers can do so much more than they could back then the memory is so massive the beat is so much faster if you ever think about what you could create now and and wouldn't I mean obviously it'd be so hard to scale up but I'm just curious what you thought about that well I think trying to scale up the kind of game we did which was heavily based on well first the idea of being able to parse natural language sentences and second the idea of creating all these puzzles and making them interact properly and stuff you could keep doing that they're you know they're better English parsers out there or could be than the ones we did and all that and people are sort of playing with that in in inform and other other venues so you could do it I I'm not sure I'm not sure it's just point more than sort of an antiquarian pursuit but I love seeing things like curses to be honest it's fun to have change from then is so great nothing's really taken advantage of the capability of today's computers to that to the extent it could but it would take massive work to do it yeah well then today's computers spend most of their time and energy pushing graphics around and the result is pretty magnificent as graphics goes and in some cases it's pre big magnificent even in terms of story there have been some games recently that have really reintroduce story into you know and even sometimes call them adventure games like you know the Walking Dead and the last of us are kind of adventure games they're not quite our style but they have a lot of stuff in common with ours and they they don't really try to push the envelope too far on graphics but I think they've done a pretty good job of game play okay looks pretty quiet well thank you very much everyone for coming

Channel: GDC

Views: 55,259

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design, Infocom (Video Game Developer), Zork (Video Game), Video Game (Industry), Video Game Culture, Talking, Reborn, Gameplay, Rpg, Free

Id: FXdmo2j_CiQ

Length: 60min 30sec (3630 seconds)

Published: Tue Aug 25 2015