The Early Days of id Software

hello everybody welcome to the talk its software's early days as told by me so welcome everybody I'm John Romero I'm co-founder of its software and I'm gonna take you on a journey back to the beginning of its software are you ready to be entertained everybody needs some entertainment today alright okay so I realized that some of what I'm about to say may sound insane but we were in our 20s and when we started we didn't know that there were any limits so I grew up in a wonderful small town in Northern California named Rockland the population was only 6,000 in the 70s I was massively addicted to spending loads of time in cold dark are caves playing all the games there and I was got really good at playing them in 1979 before anyone really had a computer at their house including me I went to the local college when I was 11 and I started learning basic from the college students I just went up just went up to them and I just asked him what the words mean their listings were and I wrote them all down on paper and I started experimenting with them on the HP 9000 mainframe there I was mostly interested in just making and playing games to keep me at home my parents got an Apple 2 plus I was done going outside after that I spent all my time programming games on that computer a few years and 20 Apple 2 games later and I finally learned 6502 assembly language from this book it was the language that all fast arcade games were written in back in the 80s then I could really make 80s games like this except that they kind of looked like this on a computer so not quite our cave games but home computer games which were on the Apple 2 and let's just say the Apple 2 was my personal home arcade as well as 1 million other Apple owners back then when I was a sophomore in high school I did some programming for the air force when I was 15 years old I can't tell you what was I was programming though it's classified after high school I kept making games and by 1987 I was finally working at one of my favorite companies in the industry which was origin systems my first job was porting 2400 ad from the Apple 2 to the Commodore 64 by this time in 1987 1988 I'd made 74 games and three previous startup companies Capital idea software on top inside out software where i ported my magic to from the apple ii to the commodore 64 and ideas from the deep and I was 21 years old so then I went to work at a company named soft disk at the start of 1989 I learned how to program a das PC there and it made a small game or utility per month for a whole year then I created a game product called Krieg gamers edge at soft disk and I had to hire a very small team of developers to make games so I hired John Carmack and Adrian Carmack not related into the department for programming and art and then Tom Hall came in at night to help us out since he was already a softest and he loved making games so this was the first time in ten years that any of us had worked with one other person or more on any game after making them alone right for a decade each and it was incredible getting together finally on a team so while creating our first game together floored ax John Carmack discovered the smooth scrolling trick on the PC so Tom and John stayed up until 5:00 a.m. and they made a little demo called dangerous Dave and copyright infringement and so they put a desk on a disk on my desk the next day I came in I saw the disk I ran the demo and I watched the screen scroll smoothly pixel by pixel and it was a massive Eureka moment for me it was like a bolt of lightning hit me and I'll elaborate on why that was in a moment but it software was born that instant on September 20th of 1990 so one thing led to another and we spent a whole week putting together a demo of Super Mario 3 for Nintendo we did this after work and intend to liked it but they decided not to publish it because they decided to only publish their games on their own nes platformer back then which was a smart move so no problem we just use the tech for different game which was the Commander Keen trilogy so why would a side scroller be a huge hit on PCs in 1990 well it was because no games on the PC had scrolled smoothly per pixel you know the PC had been out since August of 1981 even in nine years no one had figured out how to make smooth scrolling pixel by pixel like a Nintendo until the dangerous Dave and copyright infringement demo which led to Commander Keen does anyone remember the original duke nukem game alright it scrolled by in chunks of eight pixels like other games of that era and the reason why I scrolled with any speed is because I taught toddler plug 'el how to do it while he was coding dark ages and he made the game that he made right before Duke Nukem so the Commander Keen trilogy provided the start of the company its software and we made these three games in three months from September 20th to December 14th in 1990 so Commander Keen was a massive hit for us and it was so popular that people cost played as keen for years at events even today and Commander Keen pioneered the creation of game engines so we designed the game as an engine that operated on different data for different games it was the only way to get the trilogy done so quickly in fact in 1999 1991 when we were working on keen 4 we started licensing the engine for the first time so as the beginning of the modern engine engine licensing business development on our games went very smoothly and quickly because we stuck to some core principles that are important even today so through this talk I'm going to hide some of our core principles I'd like to highlight something else right now namely this photo has anyone seen this photo before so not in Europe it's a picture of John and I at the lake house in Shreveport Louisiana where we started his software so the funny thing is that people have asked me for years what was in this picture so I analyzed it recently and here's what you see so this is John me and John in September early September of 1990 we were working on the Super Mario 3 demo that we planned on sending to Nintendo so we both worked on this huge Dungeons & Dragons table that John had we used to play D&D on the weekends in those sessions led to ideas for future games like doom and quake so this is John and me Tom Hall took the photo we were working on the Super Mario 3 demo that we planned on sending in tendo and the computers were brought home from work on the weekends so this pic was taken on a Saturday or a Sunday on top of the monitor is one of those old Intel reflective astronaut plushies from a long time ago and on my left is my notepad which is a task list of bugs to fix and this is our high-level task list of what had to get done to finish the demo there wasn't really anything online back then there was no online really this is tom halls area where he was doing all the graphics for the demo he recorded gameplay on a VCR then he played it back and he paused the action so he duplicate the tied tiles exactly in deluxe paint too if anyone remembers that the TV set had a 13 channel selector on it and it was connected by an RF modulator so it was very old-school so its software was formally founded on February 1st of 1991 and we made 12 games that year Shadow Knights dangerous Dave in the Haunted Mansion rescue Rover hover tank literally did it tons of games we actually took two months per game but we made two games simultaneously so this was due to basically having 10 years of intense game development experience prior but it's also due to the first principle that we had which was no prototypes just make the game polish as you go and don't depend on polish happening later so always maintain constantly shippable code this is how we made so many games so quickly we had the whole game designed in our heads you know we just needed to quantify what needed to be done and then went about working on it until the game was finished there were no prototypes for our games so we just made them and remember we did have a small team of four people that could do this and large teams definitely require more planning so time for a quick story one day it rained really hard and cross lake in Louisiana rose and it was flooding everywhere and I needed to get to work we were furiously working away on our games and I had to get back into it and I just took a shower and everything and went down the road in the car and then I saw this whole road was flooded well basically I waited through the huge flood with water snakes in Louisiana all the way to the house and I took another shower and they could start coding and this was because we're all so excited to be making our own games you know 24/7 basically also note that during this time that of during this year that we made 12 games we also moved its software from Louisiana to Wisconsin to Madison Wisconsin and that takes a lot of time out of making games but we couldn't afford to have anything go wrong while making our games at such a pace and so we created another principle that kept us developing quickly it's incredibly important that your game can always be run by your team so bulletproof your engine by providing defaults on load failure so if you have 100 people on your team trying to develop a game that will not run you're paying for a hundred people to sit around and wait for it to get fixed so we recognize the importance of keeping the game always playable and we decided to bulletproof our engine by making all input into the game solid so game engines typically fail because they're trying to load bad corrupted or non-existent data so checking for this and providing a default for a failure case we'll keep the game running and it will quickly show you what's missing so if you fail to load a sprite into the engine just show bagel theme song is not working or loading up play something obviously wrong for the game like I know heavy metal or something missing a sound effect same thing just play something very obnoxious so after 1991 it software's first stage of company development was complete and another important principle was in effect keep your code absolutely simple keep looking at your functions and figure out how you can simplify even further so we wrote all of the games that we made up to an including quake in plain C no C++ not saying they shouldn't be used in C++ so so ok Stage two was about to begin in August of 1991 we decided to move to Madison Wisconsin Tom Hall and I visited at that time in August and we found it to be really nice his Tom Hall used to live there while he was in college so we moved all four of us there and we continued working on our games only four months later were found dead the snow victims of Wisconsin's brutal winters that we did not research so the moral the story is do your research we knew how to program an assembly language but not how to ask Tom Hall hey what are winters like up here so after six months of cold hell we moved to Texas so new principal great tools help make great games spend as much time on tools as possible I wrote a tile editor in 1991 named Ted for tile editor it Ted was used for 33 shipped retail games several of which were 3d games even like hovertank catacombs 3d Wolfenstein 3d spear of destiny raisa Triad corridor 7 etc so as January 1992 we decided to go all 3d based on catacombs 3ds promised that we had we had just made it a couple months ago so it looked cool it just didn't play very cool so in Wolfenstein 3d it took us four months of development time to make Wolfenstein and launch the shareware version with one episode of neon one episode of levels it took two more months to finish all six episodes of the game plus the hint book and in the first month the game sold 4000 copies all priced at $60 each so that was really pretty cool for us spear of destiny took two months it's a prequel to Wolfenstein 3d and it was retail only so get two games done that year soon after Tom Hall traveled to Kentucky to work for a couple months on Wolfenstein VR yes this was 1992 VR back in the days of Commander Keen I discovered a small three-person game company called Raven software in Madison Wisconsin I called him up we went over we introduced ourselves and flash-forward seven months later and we did a little work with them by modifying the Wolfenstein 3d engine and licensing it to them for a game called shadow caster if anybody remembers this this game it was released in 93 shadow casters tech improvements were sloping floors lighting and fog this engine looked slightly better than Wolfenstein 3d but it just wasn't good enough for our next game so John Carmack he spent some months thinking about how more advanced the new engine should be for the game that we decided to call doom based on the rapid development of our previous games we came up with another important principle we are our own best testing team and we should never allow anyone else to experience bugs or see the game crash don't waste other people's time test thoroughly before checking into your code no throwing it over the fence for testers to find and put in a bug put a bug in a database and then fix it later it's a wasteful cycle so after 1992 its software second stage of company development was complete along with another principle as soon as you see a bug you fix it do not continue on if you don't fix your bugs your new code will be built on a buggy code base and ensure an unstable foundation if you check in buggy code some will be someone will be writing code based on your bad code and you can just imagine how much waste that creates ok so the ideas for doom came from our DMD campaign which is full of demons we also love the movies Evil Dead and aliens so dooms design was a just a mash-up of these ideas and at the beginning of Doom's development we created a new core principle use a superior development system than your target to develop your game so before doom we were making games for das while developing on das computers and we knew that we could do better if we used more powerful computers in a more advanced operating system to develop our games so we developed a doom on next step workstations so they are far superior to das next step eventually turned into OS X OS 10 this also meant that one of our core principles was upheld great tools helped make great games we could make far better tools on next step so doom ed and quake ed were two of the most important tools that we used they both really helped us create levels and test them very quickly you might not have known this but we had 5 people on our team creating doom after Tom Hall left we hired sandy Peterson Dave Taylor which brought us up to six people we also do the Super Nintendo port of Wolfenstein 3d right in the middle of Doom's development it took us three weeks to make this port because we had to learn the Super Nintendo hardware so we uploaded the shareware version of doom to the University of Wisconsin server on December 10th of 1993 the excitement for the game was unprecedented people were sitting on the server in the upload directory creating files with sentences like when will we see doom and we got random calls in the office asking when it would be out so time for a quick story during the final day of Doom's creation we work 30 hours that day so we had that we had the game running on all the computers in the office to ensure that the game was was really solid so on a couple computers the game froze the menu could be brought up at the gameplay like just it just stopped John Carmack thought about what could possibly be happening in the game and he figured out the solution without actually doing any debugging so he was correcting his solution and we finally uploaded it after this five minute fix and more and more testing so at the beginning of 1994 I started working with Raven software and developed heretic using the doom engine I wanted to see how an inventory system in a medieval version of Doom would play it turned out great I think does anyone remember heretic ok made it yeah we also made doom to 1990 for over eight months and it was released October 10th in addition to this we did the Jaguar port ourselves and again we were multitasking to making multiple games so we made two games in 1993 three games in 1994 1995 we started working on quake and we got up to nine developers there were four there's three coders two artists and I was the only one to actually do both coding and design I wrote quake ed I experimented with level design in full 3d again we started with a clean code base we had no code from doom using quake which was another one of our core principles of development up to that time write your code for this game only not for some future game you're going to be writing new code later because you're going to be smarter also you're not tying yourself down to the limitations of your past code always try new things so quakes engine was being developed by John Carmack and the rasterization was being developed by a Michael Abrash John cash worked on the network code and he went on to become the lead programmer of World of Warcraft time for another quick story so this relates back to our belief that developing in the superior operating system can result in a better game so while making quake we actually had a deal with crazy per computers to buy a Cray Y MP for Half Price so our plan was to have our development team connected to this Cray to be SP and light our maps at supercomputer speeds as well as crunch whatever kinds of new data that we thought that we might need for our next games engine so the computer room in quakes DM 3 level was going to be full of Krag computers as a reference to this new hardware that we're going to acquire unfortunately Cray was bought by Silicon Graphics and before quake was done and that deal fell apart so the computer room in quake is filled with the usual rectangular mainframes instead of c-shaped Cray supercomputers after publishing heretic I started working with Raven on hexan I wanted to see how an FPS would play with a hub level system and character classes on it hexam launched on October 30th of 1995 during the deathmatch 95 event that was happening at Microsoft headquarters in Redmond so month later I got Raven started on my next design which was hecatomb it would be the third game in the series heretic Hexen and hecatomb hecatomb was never finished it was turned into hexen too because I left it during this time we noticed a small game company making some nice games like Raptor called the shadows and we brought them down from Illinois to make a game that we would publish they called themselves rogue entertainment and about 14 months later they released strife which used the doom engine it was an FPS RPG and it was really fun and it showed the combining genres could actually make a fun FPS has anyone played strife here alright it's on Steam by the way also during 1995 we made ultimate doom which was a retail for a retail version of the full version of doom with an extra episode and we made the master levels of doom that year as well so it's softer was still 9 developers in size and we released two games in 1995 while we're working on quake so we're continued on quake and 14 months after we started we released Q test on February 24th of 1996 for the world to test our first internet game play during the next four months we worked very hard to complete quake we also released final doom which was created by team TNT and the Kasauli brothers and we also released death Kings of the dark Citadel which was an additional set of levels for hexan so one important principle that helped us get quake done faster was this one encapsulate functionality to ensure design consistency examples of this in quake would be like just the torches on the wall we could have had us level designers place a torch model then a fire model then animated then a torch sound entity all at the same location in a map but then if we needed to move things around something could have gotten left behind so it was far easier to just create a torch entity that we dropped in the game with all the functionality built in and spawned by the engine so also the water in the game needed sound effect entities all over the place to fully cover the water areas so you'd hear it everywhere so if the water got modified in the level moving all those endings around and deleting some it would have been a huge mess so it was easier to make the game just play water sound when the when the water was being rendered so it was a renderer level feature and it was out of the designers hands and that ensured consistency and it saved memory and we did the same thing for the skies in quake sky audio so I released quake shareware on June 22nd 5:30 central time on the University of Wisconsin at Madison site so quick story while Michael Abrash was programming the renderer he was interleaving CPU instructions with FPU instructions to calculate prospective correct texture mapping so sometimes while he was playing the game for one frame the game would show a completely different part of the level he isolated the only instruction where that result could actually happen and determined that it was impossible for it to be an invalid value so he had a friend from Intel come over and go through his analysis and his friend agreed with him and told him that there was a known error with the floating-point divide instruction on the Pentium and it was a hardware error so there was nothing that we could do about it so he left it alone but this bug is known as the pentium fdiv bug which you can look up online so quake is the game that introduced the world to Mouse look to high speed true 3d world's desert texture mapped and internet multiplayer so clans sprung up immediately as soon as the game came out in eSports and tournaments we're starting all over the place quake World launched five months later to smooth out the internet gameplay so making games was and still is our life we love it more than anything else as you can tell by our release of 28 games in five and a half years by less than 10 people many other games were released that use our licensed technology over the years and here are some more core principles that we learn from doing all of this work try to code transparently tell your lead lead and peers exactly how you're going to solve your current tasks and get feedback and advice don't treat game programming like each coder is a black box project could just totally go off the rails and cause delays programming is a creative art form based in logic every programmer is different and will code differently don't waste time focusing on a rigid coding style it's the output that matters so we keep working to make great games it's in our blood as it's in your blood and it's why you're here at GDC so you want to be the best go be the best thank you for your time and I'm ready for any kind of questions you may have thanks so yeah I don't know if someone's handing a mic around or anyone has any questions about this I feel ok going alright who wants to ask some questions first or doesn't nobody want to ask a question okay got one right over here hey great talk thank you and one of the first principles is show you mentioned that don't prototype and today one of the main principles in our game production as we learn as we work we are prototyping what's your opinion on that well we we because of the time pressures we we had no time to do prototypes and we were a small company that only had four people at the time so we could actually not we didn't need the prototype because we knew from our ten years of experience that the games that we're making were small in scope they were very easy to visualize them are in our minds and that when we started it and we finished it say one or two months later it was what needed to be done in that amount of time so you know there may be twenty levels in the game and it might tend to end up being ten levels you know it's like whatever would fit in that time but all the gameplay and most of the characters and Nai and all that stuff could totally get done in that in that space but I think nowadays obviously with any really big project you're definitely going to be prototyping if you're still like a lot of mobile games they might not need prototyping if you already know exactly how the game is going to play but if you don't know if you're developing your game kind of organically and you want to see where the game goes I don't know if I would call it prototyping I mean it's like if you start making your game and at some point you might discover something about it that changes what you're doing in the game but that's your game it's a prototype if you're throwing it all the way and starting over but if you're still going forward I don't know if that's a prototype I think that's just the beginning of your game thank you all right yeah hi I um I always felt especially looking back on doom and quake that that the music in your games was always a big part of the games I for example I often listened to the to the doom soundtrack and I was just wondering for example the the collaboration with Trent Reznor on quake how how did that happen maybe you got another quick story that you could share yeah the the story with how we got Trent Reznor involved in quake was kind of funny after doom it was it was a hit everybody duma's was such a big game that hollywood started to get into they tried to get into games so they knew like there had been an earlier Hollywood invasion of the game industry and they were rejected and they decided to go in in a different way so we were contacted by Hollywood Asians ICM and we had a person handling our our account that our name bill block in bill block at that time had a lot of other clients and one of which was was Nine Inch Nails well Nine Inch Nails they played doom all the time on the tour bus they had four computers all networked on the tour bus when they were done with the concerts they were on the tour bus death matching and they did it all the time I it's funny because I just spoke to someone like two weeks ago about that like they they were interviewing Chris varenna and getting all this like oh we played doom all the time stories and so Bill found out about that and said hey the clients of mine and we heard that they were clients and we said wow it would be really cool if he would to the soundtrack for the next game instead of having MIDI kind of metal soundtrack we could actually get some industrial real music in here so so basically they agreed they agreed to meet up with us and they came over to our office in Mesquite you know Trent would pull up in a limo and I don't know how but there'd be a bunch of women outside the building waiting I mailed they don't tell them anybody that they're going there but they find out where he's going and so they just wait outside while he's up there and we're talking about stuff you know about the games and the kind of music that we're thinking about doing and and we hung out several times and just kind of kept in touch through email in American McGee would flight into New Orleans where Trent lived and and just basically kind of coordinate the music creation and it was kind of funny because quake was delayed because of the music because if you involve like someone like Trent who's like out on tour all the time and has a ton of a ton of Rights around the music there were 12 signatures we had to get from Interscope to actually get that music released to us so it took us extra time and finally as soon as it was signed off then we could actually release the game so just watch out for that thank you he's nice okay so you had a rule stating that you didn't want to reuse code because you would not get the results that you wanted at what level did you decide between using engine code and specific game code because you use Commander Keen the engine code several times over at what level do you make a position and how well the Commander Keen code let's say for Commander Keen one through three it was the same engine so it ran the data that we created so we created levels and characters and audio and all that and that was just data the engine worked on and the engine itself had if deaths basically in it for each of the command routines back then so if this is keen one then these actors would be you know would would their code would be it would be running and we would load these files so there were if deaths in the engine just to handle the specific sets of data that or sets of code that were different for the different we just didn't want the executable to be full of three games worth of specialized information so we if deaths some of it but um but in to keep things really small on a disk we had to take like generic data files and turn them into object files and then link them into our executable and then compress the executable because we had to fit on a you know 140 hero3 what was it back then it was a one point four four three and a half inch disk so we had to we we did need to do that where we did we didn't keep all of our code and one big xq ball but it was it was if theft out so in the code it was there but not in the executable but when what I was talking about was when he went between major jumps like keen keen one through three to keen four through six those engines have no shared code in them or Wolfenstein or doom or quake it's basically start with the new code base and that was actually back in the 80s that was that was pretty typical like if you look at the Ultima games you know Lord brittish's Ultima games every one of them was just brand new code just start writing the new game and maybe there's a faster way to render tiles on the screen or you know use code paging reader your overwriting code in memory because the game is too big etc so yeah we we didn't we didn't reuse code unless it was something that didn't it was very small and it didn't matter like loading a file but usually it's so small that we probably just type it ourselves again the runner all right now obviously you were working on porno clothes if my bros free during the 80s joke early nineties why do you think side-scrollers even though they were really popular during that era even though they fall in our favor now why do you think as a pioneer of the first parachute person shooter genre why do you think the show has gone from strength to strength to strength to strength and has no there are no signs of it like going down in terms of popularity you mean the shooter genre yeah the first-person shooter genre that ultimate ID pioneered back in the day well the the great thing about the shooter genre is that is started to fragment its design across different styles so when we were making our games like we just made this this one type of game which when I mean I'd say Wolfenstein and doom and quake are very similar in that you go through levels and you just like kill monsters and get to the exit and it just looked better each time and things acted differently and sound better at all and look better but um you know as soon as other people started to get into making shooters then we started to see people needing to be like more creative because they're not we're not the they're not the only game in town you know those now there's a bunch of people making them so if you look at say the Red Storm Games when they started coming out with Ghost Recon Rainbow six and Splinter Cell those word that was basically saying here's a whole tactical section of the shooter genre where you're creeping along you're getting to do really long distance sniper sniper rifle shots and and in in so call of duty where you have checkpoints and everything advances between checkpoints so people started to really try and innovate within the genre and now you have you know overwatch you know tf2 and and and uh lawbreakers in in in this whole competitive way of playing shooters and I know that that we have let's say 'la MMO shooter or destiny so everybody's taking this viewpoint and this way of basically shooting stuff and applying it to a lot of different ways of getting through you know or getting through objectives basically and wrapping some kind of narrative around it that's really different and sometimes you get something that's you know very different like Splinter Cell is is pretty different from those other games but you're still you know you're still shooting in some parts so I think it's just because there's so much competition people are trying to do something new with it if you if it has anyone played the Hitman sniper game on mobile because that's another really cool way of making a shooter where you just have a vantage point somewhere on your mobile phone and you're just zooming into a scene and scrolling across the scene and tracking people and it's totally a first-person shooter but you don't move and it's just another way this is another example of how people are trying to innovate and push forward with this first-person shooter perspective but but how you play it is different and I know that the innovation is just going to continue you know people are just going to continue trying to innovate with that viewpoint and and what else can you do with it alright so yeah so your story has been very inspirational to many many people in the industry and what do you think was unique to its offer in those days that enable this to happen and do you think it can happen again to two teams these days and what would be your advice yeah I mean I think it's it's a super common and you're really everybody can do the same in what it comes down to is people that match pretty well having a pretty good amount of experience but really having focus and determination to make your game and finish your game and do that over and over again so when when Wolfenstein came out that was game number 87 for me and doom was like nine game number 90 and we had made so many games and we'd iterated and we have shipped over and over it that we were we knew how to do it and in we didn't have to really worry about how to finish a game or even have a starter game we were just worried about how to design a really good game and to create new technology and kind of make those things work together nowadays you don't really have to worry so much about the tech and it really is all about no design is law it's all about what is that design that you're doing that you're creating that no one has seen before because the tech is already there for you to use you know except for when you're in VR and AR and there's a lot of new stuff that you're going to create there but basic traditionally on console and PC what is the design and can you focus and spend a lot of time on that you know innovative unique design to make something people haven't seen before the end is already out there if you look at Minecraft it's a perfect example of here's an amazing design that one person created you know in and it was all about what you do in this simple world people you know it's easy to complexify a design that takes nothing to make something really complex it's hard to simplify everything if you just sit there and spend time simplifying your your design taking taking away things from your UI and things that would just complicate what you're doing and try and make it contextual within the game then you're getting better and better at what you do and and so I say everybody has the same chance to do it it's just it takes you know it's not really team but it's helpful if you have a team and it takes focus and determination and I think for us it was us being in the same room and being co-located over the Internet it's a lot harder given that you didn't make prototypes and you made so many games were there any concepts that you had to walk away from a throwaway um let me see if we yeah we did so after we made our first Commander Keen games we were going to make a game called quake the fight for justice and it was gonna be an RPG and everything so we started working on it and we worked on for about two weeks and then we decided we knew that quake was supposed to be a really great game it's supposed to look great supposed to be a powerful game and that what we were making right after commander Cain was not matching that idea of who quake was in our minds that you know from D&D so we basically said forget it we're we're still making these simple kind of cartoony games let's just keep on getting better at it but we're gonna put quake off and so that was like early 1991 so we waited four years before I actually got the tech together and what we knew we could deliver on the original you know promise or design of quake in 1995 when we started working like quake other than that we we didn't really start on any game that we didn't finish it was we talked about games and talked about ideas such as when we finished catacomb 3d and we were planning on making the next set of commander Keynes keen seven through nine we decided that you know what let's not do this side-scrolling thing again I mean we can do maybe a keen someday in the future but maybe it'll be in 3d but let's start doing 3d stuff like we just made catacombs 3d which was really like one of the first texture map games even though it was in EGA so Tom Hall thought of this idea where you know what if there's a research facility and there's all the mutants all over the place because they created these monsters and are destroying level scientists you're going to go in and rescue them all and I just told them that's an awful idea I don't don't want to play that game already and so I said why don't we just remake Wolfenstein Castle Wolfenstein and we all just immediately went oh my god let's do that so that original idea that that was just another idea that we had that we didn't do but that's like pretty much the only extent that we had back then of not making something it was we kind of knew what we were going to make and we just made them so there wasn't a let's start doom and then reboot or something like that we we just made the games we just came up with the designs and made them yeah hello my question is you managed to make games so quickly so very very quickly how did you do it and what happened afterwards because I think it became a little bit slower yeah it's kind of funny a lot of new programmers don't fully understand what we had to do back then because there's so many API is nowadays that do everything for you so let's pretend in a world where there are no apap is because that's what we lived in there were no API so making our stuff there was BIOS you know the thing that when you boot your PC and you see some of that text that you should never see ever again like that's your BIOS and that's the only thing that we could actually talk to through this simple interrupt 21 assembly language API that was it so putting pixels on the screen for us was we knew where in the computer's memory if we put a value in memory a dot would show up and so that's how we put our graphics on the screen we actually had to write all the code in assembly to put the dots on the screen quickly so you'd see a shape in backgrounds and everything there was no API to you know send a display list of triangles to a GP or anything because it didn't even exist so while making quake if you just imagine a full 3d world everything that those video cards were doing we actually had to do that in our own code and we knew that we when we were developing those ap is that when we're developing the code to rasterize quake we are also working with Nvidia on the pipeline that we wanted the cards to be using which are the API is that into the on card api's that directx and opengl and stuff talk to but we had to put every pixel on the screen ourselves we had to read the keyboard ourselves there was no nice functions to call to do stuff like that and that is just to get the tech working not to mention drawing characters you know in 2d so 2d keyframe animation everyone knows how long that takes there was no tweening there was no 3d stuff so it took us a long it took us so it sounds like it did it did take us a long time but we spent all of our hours working so especially when we're in Madison Wisconsin why would we go outside there's no reason like to even go out we could order pizza and live inside of a one room so that's what we did and and so we basically slept for eight hours and all the rest of time we worked for you know half a year while we were doing that which got you know several games done at that point but really it was all about like we didn't have anything else going on there was no Twitter no Facebook no internet no nothing so nothing is popping up in bugging us so we could just sit there in code all day long listening to heavy metal and stuff so that's what we did there were no interruptions and it was all about time focus and determination okay since you have id Software Dhoom 3 came out and doing 2016 came out only recently doom 3 work was criticized by some people for the fact that focused too much on some other horrors doom 2016 has been criticized even though love early doom fans have praised it for the fact they to come back to a back to basics formula a lot of people have also stated that it hasn't taken into account the joy innovations that have occurred since doing - if you were to be say like in an alternative universe if you could have had a hand in the making of doom 3 and 2016 how do you have Don Dean what what I have done for Doom three are doing four well I would have made a sequel I think I would have made a sequel to doom - because Doom 3 and doom 4 were reboots of the original doom in different in different ways so it would have been a sequel I can't say what that sequel would have been because I never thought about like making a sequel to doom - if you look at the games that we had made back then we made a game and sometimes we made a sequel to the game like Wolfenstein we did a sequel kind of prequel but then we did spear destiny we did doom doom - quake and I started designing quake 2 before we left but it was always main shareware game retail game main shareware game retail game invent new IP invent new IP invent new IP so we probably wouldn't have done that if I was there because we needed to create new franchises that was more interesting than living inside of our franchise for a long period of time since if it was a game that was a service based game that was operational every day yes I could live inside of that for a long time but but we weren't making those back then and probably wouldn't I wouldn't have personally want to make a doom 3 back then I would have that's why we made quake and so after quake 2 if I was still there we would have been done a totally different game so I guess the the answer is I would have done it all right 6 more minutes and excuse me for my a bit stupid question but I'll play in games today now and Thomas your favorite maybe favorite title or general video games now if it's not a secret I will say again I can couldn't understand that well what's your favorite general video games now today what's my favorite game or genre yes oh wow and David genre well I kind of like FPS is this it's kind of obvious I'm also loved mmo's people you know people don't know this but I worked on an MMO for four years has anyone heard of gazillion entertain so gazillion is an MMO company that I founded with two other people in 2005 and we worked on I worked on an MMO that I designed for four years before we had to stop working on the game but during that time I played World of Warcraft every single day for 5 years 6 hours minimum 12 hours minimum on weekends so I have thousands of hours and Wow so I love mmo's so we get 5 minutes um so when you did your original games we didn't really have the Internet as we know it today and we didn't have defeat back cycles on for instance bugs as we have today we've also gotten more advanced in our engines so it's easier to create books how has that how did that now how has that influenced game design and the creative process and do you think that's do you think it's a benefit or do you think it as a bad thing well yeah back then the only feedback cycles that I had back then was when I tried to get something published so I would have to send my games through the mail like paper mail with a disc in the envelope and then the editor at a magazine back when people for some reason typed games in from magazines in the computer so I published a bunch of games like that and the editors would get the manuscript they get the disc and the big huge and the big huge listing you know some of the language listing or whatever and I would I would they would see that they would review it and they would send back you know pages of little tweaks here and there you some of some of them were in the game most of were in the text of the manuscript and that was helpful because that actually they were usually really good suggestions and and as we as we kind of went got into the age where people could just like get online and start posting feedback and we actually had beta testers we would send our disks in the mail to in the early 90s and really good feedback because they would find some bugs in our games that we couldn't that that were not obviously crash bugs or anything just weird obscure little things and and so that was helpful and then when we actually get to the point where we're on the Internet people are just coming up with really good ideas and you know there's there's the off you have to like obviously take everything with a grain of salt you can't read the comments and all that you have to as a game developer you trust yourself you trust your design and and you just kind of have to believe that that you're making something that you like and but you do have to listen criticism because sometimes you know people have a lot of really good stuff to say so I think nowadays it's it's gotten better games games are getting better there's a lot of good feedback out there if you listen to it and sometimes the players can be a really good part of the process of defining where the games are going is as you can kind of see with the way Kickstarter works you know there's a lot of feedback going into Kickstarter campaigns and in people going forward I think we'll just do this one more was it okay so yeah I'm just starting out with programming well like two years and I'm starting older than you so this is kind of like a philosophical question more how did you when you started learning to program avoid burnouts or did you did you just program and you just had fun at it there was no burnout like I I couldn't wait to wait like in the summer times when I was going to school or when I wasn't going to school like get out of you know high school ends and and then it's like had eight o'clock I could wake up and I could code until 2:00 or 3:00 in the morning I couldn't wait to do that I actually used to get mad that I had to go to the bathroom so there was no burnout because I was so excited that every time my the kind of programming that I like to do is I like to see results all the time I don't like to build a giant chunk of something that hopefully will be cool at some point and I'll see it I like to see stuff as much as possible so my a duration time is just usually a couple lines of code before you look at it again I run my code over and over like thousands and thousands of times I'll write a couple lines I want to see how it works so I know that what I wrote is not going to turn into this big huge problem later and and so I'm always working on getting cool stuff up on the screen so I'm always excited the whole time I'm making something so burnout is maybe you know you need to get outside and eaten and see something other than computer screen but anyway I really I don't burn out that much okay yeah that was last question all right that was that was the last question I believe so so thanks everybody for listening to the story

Channel: GDC

Views: 151,045

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design, doom, id software, level design, game design, game programming

Id: E2MIpi8pIvY

Length: 56min 3sec (3363 seconds)

Published: Wed Oct 05 2016