An infinite procedurally generated city where every building has rooms and stuff is kinda the holy grail of sandbox games for me. If I had like 10 years free I'd probably end up making CDDA but pretty, kinda like what Rimworld is to Dwarf Fortress.
I don't know how apt that comparison is. Even the largest Rimworld colonies only have a dozen colonists, compared to the hundreds possible in Dwarf Fortress. I'm skeptical that you can even make a game that deep that is also "pretty". I guess it's theoretically possible, but I don't think it's feasible. Hell, even Dwarf Fortress is basically Tarn Adams' life work.
@@Inglonias Maybe, I don't know. I can definitely see how not letting graphics hold you back could help with the simulation complexity (on top of just having more time), but I'd like to think you can have both. Any thoughts on DF's Steam version?
@@Huntracony Not released yet, Steam DF will likely take a few years, a bit longer for QoL mods to kick in. There's a triangle between QoL, beauty, and complexity, where you have to pick and choose. QoL and complexity, in particular, have great difficulty coexisting. You have to pick a spot on the spectrum, and accept the drawbacks. The strength of DF is the complexity, particular in dwarf "free will", but this comes at the cost of QoL, where you can't take control of a single dwarf and micro them effectively. (Instead, you have to create allowed and disallowed zones, draft and undraft, create complex machinery, and so forth). The strength of Rimworld is the QoL, you can control individual colonists, but in exchange colonists have very little free will to generate stories with. Modding can help bridge the gap, but there is always give and take. I believe the sweet spot would be having an "overlord" that can be micro'd, low class minions would have no free will, but middle class minions would be full colonists with significant free will. Example: A dragon/dungeon core/lich might have an army of loyal but dumb skeletons/homunculi/kobolds/goblins, but any specialist henchmen/humans/princesses/rogues it kidnaps/recruits would be powerful wildcard.
Aside from being a fun series with a good cause behind it, it’s been fantastic to see an oral history of Introversion software recorded and documenting the development of modern indie games. I wish more folks would do things like this and would love to see a ‘Success MasterClass’ series for your more popular games!
It also is nice to see that even Indie devs who have made a very sucessful game and reached every other indie devs dream still sometimes have to scrap projects cuz they just couldn't do it.
There's some great stuff from GDC in this sort of vein; the classic game postmortem series in particular comes to mind, you might want to check those out!
I fondly remember seeing the blog posts showing off the city and building generator and how it made my young self dream of the games in the futurue and how awesome it's going to be.
Oh yeah, I was so excited when I saw all the various bits, like the logic systems they were implementing! At the very least it's finally nice to see this crop up again to where so many people can see it and hopefully learn from what happened!
Yeah, I remember being so bummed when this was cancelled. Wrong Wire (That bomb defusal game you prototyped) reminded me heavily of it. Personally, I still think, even today, Subversion deserves another chance with your improved skills and hindsight with what worked, what didn't work, etc. because the concept had serious potential and the market for stealth-action games is a big black hole now that Metal Gear and Splinter Cell are both dead. Call it "Inversion" or "Aversion".
"That's just...that's just madness" 100% Subversion was so subversive it subverted itself into a completely different game. It's glorious to see all the signs on previous games on what later became Prison Architect and even The Last Starship.
There's an interesting looking game in development called Shadows of Doubt - it looks like it takes on the city, people and system simulations, and somehow makes it work
Yes, but with slightly different goals. Introversion wanted subversion's levels to feel real and have people going about their days. Invisible Inc was okay with trading realism for gameplay. The locations don't feel like real buildings. But that's okay because that wasn't the priority. Definitely one of my favorite games though
Mark is the kind of friend EVERY artist needs - he's very to-the-point, but also very honest and helpful. He's like a contract skill worker you bring in to diagnose what's wrong with some big system and can just immediately diagnose the issue, and I haven't seen that in game design. He's annoying good at it XD
Here we are, the elephant in the room.... ok, now go back to this and finish it. There are SO many ways to make this game fun. Feel free to hire me free of charge to help you develop it. Since there is fog of war, imagine researching where does the bank have the blueprints of the building, and thats like a side mission, getting the blueprints, so it would make the heist easier, it would show the inside of the building but without showing the people. Also, something many games fail at, game credits/money being useful for something, being able to improve the gadgets with the money you get playing. For example the wall sensor being weak and gets stronger and more detailed information as you improve it. Gadgets for aggressive playstyle as well, like an EMP grenade And since the first time i saw this game, i always liked the idea of having your own customizable hideout, adding your own security to it, and having to leave as little proofs as possible on your heists, because after each heist, it would simulate a detective system that tries to find proofs to know who and your crew are, trying to find where your hideout is. And of course, the cherry on top, make everything in this game interactable with the already polished Uplink hacking system, having a crew member in the van with a laptop, and allowing you to hack all the building as you would on Uplink. Don't get me started on online co-op possibilities
it's a big shame this game never came to existence .. i was very much waiting for it! I really loved `They stole a million` back those days and there was a version of `Rainbow Six` where there was a planning phase before the mission and I thought this game would take it a step further, but maybe there is some hope :)
I encourage anyone who is interested in this game (especially what they show starting around the 20:30 mark) to give Invisible, Inc. a try. Perhaps they mentioned it in the video but I haven't watched it fully yet.
It is really a shame that all this beautiful work amounted to nothing. I'm sure that you have heard this before, but still: consider this a request for releasing the source code somewhere so people can look at it and maybe make something out of it.
I had no idea that you guys got so far with this. I only ever saw the procedural city generation which I already thought was pretty mind-blowing. Would love if you guys picked this up again. Definitely feels like the Frozen Synapse 2 devs may have also gotten some inspiration from you guys! ;)
When it comes to procedurally generating a "dungeon" with layers of locked doors/challenges leading to a goal and ensuring it is always winnable/connected up correctly etc, I think Joris Dormans of Ludomotion (Unexplored) is a good person to look to for how to make that work. It's hard for sure if you want anything complex (just use a 2d grid :)), but doable. Defiantly a major investment to develop the tools for that, so understandable that you didn't after working so long on this already. I've been super inspired by you guys, ever since Uplink. I'm messing around with ideas for a hacking game myself and all the lessons you have learnt and shared in this series have been super insightful.
I tried implementing that paper way back in the day as well. If I remember correctly, the original implementation used L-Systems with actual strings that were later converted into a street graph. To make it more efficient, I made my version work on the actual object instances, so that I could later interpret them as the street graph without having to use strings as an intermediate representation. I never implemented anything past the street level, though. During that time, I stumbled upon Subversion and due to the shared interest in procedural city generation, have been a follower of Introversion ever since. Seeing this canceled was a huge bummer for me.
It's a great ambition, and perhaps PA was so good because of the lessons learnt along the way. I really love how this series has celebrated failure - it's a vital feature of success. And with 9 fantastic failures under your belts, this new project could be one hell of a success.
This series is criminally underrated. It's so insightful and wholesome, and gives a real window into the mindset of an interesting gaming company. Looking forward to see more details of the new game.
Simulating things down to their individual parts is something that excites me as the player a lot. Both as a "Wow, that's really cool how it's designed" and "I wonder how many different ways I can interact with this" Like the camera, for example. You could disable the motor to create blind spots, mess with its lens and narrow its FOV, disable or reroute its connecting wire, etc. People can get really creative when it's a simulated system of individual parts.
What about...slow down the movement of all cameras by 20% so now you have to move in a specific way to make sure no camera sees you? :) this kind of gamedesign is amazing to me, I wish there were more games like that
I wish this would be given another chance. I was so hyped for it. I don't care if the ambitions are greatly reduced, I just really like the aesthetic and concept of an expansion of Uplink in this direction.
I cant believe its been 5 months, I often come back to this video. Those early tech demos are/were a huge reason I got into game design and my love and respect of coding. Its still so impressive all these years later. I know you guys listed the reasons this idea is totally broken and would never work, but I implore you, of all the demos so far, this is the one to go back to. You can totally do this, I know it in my bones this can be done
This was the game that put Introversion on my radar :P Still sad nothing ever came of it. This sort of stealthy-messing-with-the-systems-of-the-game is by far my favorite shit to play around with. The Hitman series is after all my favorite series just for that reason. Hitman on a larger scale with crappier graphics to be able to have it all running... I'd be down with that :P
You never _really_ left =) *But*, a humble request: Please make it somewhat early access and let us enjoy your monthly videos again! It's ... idk, I said it probably before - but this format just works for me. Idk, Introversion games (even the "failed" ones here) are unique and special somehow. They all are more than the sum of their parts, really. Anyway, _take care guys_
Shows a lot of technology that has still not been implemented in other games or is just now getting partial implementation. So much potential. Maybe some day there will be a sandbox game with full building interiors.
22:08 the way to "square that peg hole of procedural vs authored", in my opinion, is making the level generator "scriptable". You need to be able to get to a point where you can tell it "make an office level with three levels of interconnected rooms with doors. Inside the last room there should be a hardened laptop of value X" with code. It isn't impossible, but it requires a lot of "solving problems that sound easy for humans but are very hard for computers".
This video was very enjoyable. Thank you very much for this series and all the insight into your process and for the prototypes you've provided through out the series. There is a wealth of game dev information in those prototypes and I even as a non game dev can appreciate that. I never knew about Subversion and only found out by following an obsolete link to your website which showed me the old design and at the bottom I saw Subversion in between the rest of your works and I had to click on it. And so I found my way here and found out there was going to be a sequel to Uplink, my favorite hacking game. Your 3 original IPs had a very unique and intrinsic charm to them that are still unforgettable to this day and I bet Subversion would have been the greatest of the bunch. I still hold out hope that someday Subversion will see the light of day as a fully realized game because it's a great concept. The hacking gameplay of Uplink + what you showed in the demo would make a great game.
Big thank you to Chris and Mark for a fantastic series. I was hooked on the first one. Introversion has always been and interesting company and having this insight into your mindsets has given me a window as to why that was. Looking forward to this new project.
This series was a really fun and interesting journey in the mind of game developers> Honestly enjoyed all of these videos ! Thank you for the great content guys.
I hope y'all are proud of yourselves! Even though none of these games may not have "qualified" to become full games in their own right, each one demonstrated how much talent, skill, and passion you guys have for making top tier videogames. It's incredibly exciting and inspiring to watch! Looking forward to all your future successes 👍
Excellent and interesting video. Even though you're explaining how and why it died, I felt the rush I had 10 years ago after discovering uplink and then finding out you're working on a heist game! It looked cool, the mechanics were amazing as they were complex, and there is a part of me hoping you find that "Eureka" moment to pull it all together and jump back on Subversion (doubtful but I live in hope). Thanks for the hour of entertainment today, and for years spent enjoying amazing concepts like Uplink and Subversion
I'm glad you guys shared this in earnest. Subversion was among the first games I really got into as a kid, and I followed you guys all through Darwinia and Defcon. I was awesomely excited for Subversion, and sad to see it go. It's very cool to see the seed of Prison Architect form and become something to. Good show.
I love that simulation of things that you guys go with. Its the sort of thing I'd love to do if I could program. When stuff is systemic rather than arbitrary it means you can interact with it. Often indirectly. Which is fantastic.
This is bloody brilliant! Even for now adays. Truly impressive on every level. I landed on this video be a use I'm trying to figure out how to generate cities, and stuck around just consistently being more impressed. You should release the game open source and let the massive game dev community take a crack at it with the technology that's available now. Oh, the video ended with yall saying you're going to release the build... Well, that's simply charming. That's incredible. Brilliant!
Excellent video series guys, thanks very much for putting it together. The city generation for subversion was so beautiful it would be great to think that there is still some potential there for it to be used in a future game of yours. When you come to your next game adventure, I hope too that you'll give us more development videos because they are really interesting to watch even when i wasnt always keeping up to date with trying out each of the PA alphas myself.
This is mindblowingly cool. Everything about it. I've been surprised multiple times while watching this. I feel that you have shown too many impressive things in this video to call it quits on Subversion. It's almost like you're doing a disservice to gaming by giving up on developing this game. I understand what you mean when you say that it's hard to merge the hand-crafted with the proc genned. But you should really take note of what Tarn Adams has said about procedural generation: when making proc genned systems, don't try to make huge complex systems. Make multiple small units that work independently and then connect them together to make a whole system. That's how he managed to do such a good job on his own game.
I've got a box set of Darwinia on my shelf here --- the green DVD case (although it's actually a CD inside). I've even still got the printed manual with the fake newspaper articles in it and the slip-in note explaining where to go for help if it didn't install! Which, IIRC, it didn't due to early game bugs. I must play it again; it's still a fantastic game, it runs really well on modern systems, and you don't need to use gestures to command units any more.
So you mentioned in the previous video that you like designing games with interactive systems in them (like in this one - the security wiring and all that) and were looking for game ideas. With your experience with Prison Architect, it wouldn't be too hard to imagine a game like Space Station 13 (maybe the Colonial Marines variant) would be something up your alley. Half of the fun of that game is the systems that interact with one another and weird shenanigans you can do (like killing people with doors because you rewire their logic). It would be lovely to see a modern take on it that doesn't rely on a 20 year old single thread engine...
I just started watching Money Heist on Netflix and remembered about this game, I just checked the introversion website to see an update today, what a coincidence!
You guys make great games, many of these fail masterclass games have been amazing concepts. Until prison architect I thought I’d never heard of you but I loved defcon as a kid and enjoyed darwinia. Prison architect is one of my most played games and I can’t wait for your next game, I’m willing to wait tho
Reminds me of "The Sting!" from way back, just procedurally generated. Which kinda describes the ideal game for me. When are you guys coming back? We've been all waiting for the next big thing from Introversion.
Oh man here it is. The one I dared not wish for. What a shame this one didn't come together, I do remember the impression the Dev blogs left on me after playing Uplink for the first time. Excited and nervous to watch this one lol
Everything you guys talked about here sounds to me like stuff that could absolutely be done now, it's just that you were making it in 2007, with the resources of a indie developer. It's still a game I would absolutely love to see you folks take another crack at.
I still would want to see this, just because of the amount of interconnected systems. Games like Space Station 13 or even Rimworld are still very popular games because of the huge amount of small, interconnected systems. I'd still love to play Subversion fully developed, I was so hyped back then and I am still today. I miss Subversion, even though I never played it...
Chris: well we built this city generation algorithm Me: oh that look really cool C: and we wrote a paper on it Me: oh that's pretty cool C: this was before unity and unreal Me: WHAT THE FUCK C: and this is our own 3D engine with quarter of a million polygons in 2006 Me: WHAT THE FUCK C: and the buildings would procedurally generate offices and stairwells and people with a day/night cycle Me: OMG CHRIS YOU ARE A GOD Honestly I agree with Mark too in that it was way too complex to limit it to a proper well designed game but omg this is god-level programming. Too bad he didn't get to the people and day/night simulation
3 роки тому+3
It wasn't his paper; the paper was written by someone else and Chris implemented the algorithm.
I love the procedurally generated cities and the procedurally generated buildings. It’s amazing technology but I agree that the environments will all end up feeling too similar. A bank is a bank is a bank. And procedurally generated puzzles is at odds with the realism goal. I got a great kick out of your story of how you abandoned procedurally generating a prison and instead crowd sourced it and found a new, unexpected game in Prison Architect. I wonder if you could crowd source the generator rules to your players somehow.
I remember this! I think some version of the city generator was in a humble bundle at some point? I remember hearing about it more fully and thinking the idea was very cool!
Great series, thanks so much for sharing I've always wanted to have a "fable" or "Skyrim" game which was properly simulated. they put in a bit of stuff to make it feel like a real place but it never really feels totally real. you could totally adapt your simulated behaviours to create a working village. then your actions would always have direct consequences to the world.
Also, kind of hungry for you to do one of these videos but with Josh Parnell and Limit Theory. Would love to hear the three of you comparing and contrastng.
I'm an aspiring game dev and your series has been incredibly interesting. I recall you guys saying something like you want to get your videos out in the algorithm more, maybe try a video like "indie game Devs react" to get eyes on the videos and I'm turn watching your other videos. Just an idea :) Ps I love your released games!
You could have used opposing AI to determine the levels themselves. For example, for a bank, you would have specific objects that would have worth (cash) and you would pit AI vs AI where one would try and create a layout that would protect it and another would try to get it, while a third is a "neutral" AI that must be able to perform business within the space provided. Systems are added and deleted as necessary to optimize the space, but leaves remnants to make it more realistic (like maybe fake cameras or extra wires in the server room). Objects themselves would have built-in use cases to take care of actions...for example, a teller desk would require a teller and small amount of cash on one side while forming a line on the other side, requiring room and the ability for the employee to get to it. Just an example....but everything in the world has use cases that can be given reasons for existing to AI.....maybe the protective AI needs to also maximize decor of the neutral AI as part of the optimization pattern.
Ah Subversion, fond memories - those screenshots and tech videos still haunt my dreams…
An infinite procedurally generated city where every building has rooms and stuff is kinda the holy grail of sandbox games for me. If I had like 10 years free I'd probably end up making CDDA but pretty, kinda like what Rimworld is to Dwarf Fortress.
I don't know how apt that comparison is. Even the largest Rimworld colonies only have a dozen colonists, compared to the hundreds possible in Dwarf Fortress. I'm skeptical that you can even make a game that deep that is also "pretty". I guess it's theoretically possible, but I don't think it's feasible. Hell, even Dwarf Fortress is basically Tarn Adams' life work.
@@Inglonias Maybe, I don't know. I can definitely see how not letting graphics hold you back could help with the simulation complexity (on top of just having more time), but I'd like to think you can have both. Any thoughts on DF's Steam version?
@@Huntracony Not released yet, Steam DF will likely take a few years, a bit longer for QoL mods to kick in.
There's a triangle between QoL, beauty, and complexity, where you have to pick and choose. QoL and complexity, in particular, have great difficulty coexisting. You have to pick a spot on the spectrum, and accept the drawbacks.
The strength of DF is the complexity, particular in dwarf "free will", but this comes at the cost of QoL, where you can't take control of a single dwarf and micro them effectively. (Instead, you have to create allowed and disallowed zones, draft and undraft, create complex machinery, and so forth). The strength of Rimworld is the QoL, you can control individual colonists, but in exchange colonists have very little free will to generate stories with.
Modding can help bridge the gap, but there is always give and take.
I believe the sweet spot would be having an "overlord" that can be micro'd, low class minions would have no free will, but middle class minions would be full colonists with significant free will. Example: A dragon/dungeon core/lich might have an army of loyal but dumb skeletons/homunculi/kobolds/goblins, but any specialist henchmen/humans/princesses/rogues it kidnaps/recruits would be powerful wildcard.
DF is far more complex a simulation than Rimworld, both great games in their own way.
Aside from being a fun series with a good cause behind it, it’s been fantastic to see an oral history of Introversion software recorded and documenting the development of modern indie games. I wish more folks would do things like this and would love to see a ‘Success MasterClass’ series for your more popular games!
It also is nice to see that even Indie devs who have made a very sucessful game and reached every other indie devs dream still sometimes have to scrap projects cuz they just couldn't do it.
There's some great stuff from GDC in this sort of vein; the classic game postmortem series in particular comes to mind, you might want to check those out!
I fondly remember seeing the blog posts showing off the city and building generator and how it made my young self dream of the games in the futurue and how awesome it's going to be.
Oh yeah, I was so excited when I saw all the various bits, like the logic systems they were implementing! At the very least it's finally nice to see this crop up again to where so many people can see it and hopefully learn from what happened!
Yeah, I remember as well, I was so excited about the development. Can't believe how much time has passed.
Yeah, I remember being so bummed when this was cancelled. Wrong Wire (That bomb defusal game you prototyped) reminded me heavily of it.
Personally, I still think, even today, Subversion deserves another chance with your improved skills and hindsight with what worked, what didn't work, etc. because the concept had serious potential and the market for stealth-action games is a big black hole now that Metal Gear and Splinter Cell are both dead. Call it "Inversion" or "Aversion".
"That's just...that's just madness"
100%
Subversion was so subversive it subverted itself into a completely different game. It's glorious to see all the signs on previous games on what later became Prison Architect and even The Last Starship.
There's an interesting looking game in development called Shadows of Doubt - it looks like it takes on the city, people and system simulations, and somehow makes it work
Invisible inc. is a good example of procedurally generated heist levels.
invisible inc was inspired by subversion
Yes, but with slightly different goals. Introversion wanted subversion's levels to feel real and have people going about their days. Invisible Inc was okay with trading realism for gameplay. The locations don't feel like real buildings. But that's okay because that wasn't the priority. Definitely one of my favorite games though
This is a day I didn't know I was waiting for. Always thought Subversion sounded awesome!
This dude was soo ahead of his time
Mark is the kind of friend EVERY artist needs - he's very to-the-point, but also very honest and helpful. He's like a contract skill worker you bring in to diagnose what's wrong with some big system and can just immediately diagnose the issue, and I haven't seen that in game design. He's annoying good at it XD
Here we are, the elephant in the room.... ok, now go back to this and finish it. There are SO many ways to make this game fun.
Feel free to hire me free of charge to help you develop it.
Since there is fog of war, imagine researching where does the bank have the blueprints of the building, and thats like a side mission, getting the blueprints, so it would make the heist easier, it would show the inside of the building but without showing the people.
Also, something many games fail at, game credits/money being useful for something, being able to improve the gadgets with the money you get playing. For example the wall sensor being weak and gets stronger and more detailed information as you improve it.
Gadgets for aggressive playstyle as well, like an EMP grenade
And since the first time i saw this game, i always liked the idea of having your own customizable hideout, adding your own security to it, and having to leave as little proofs as possible on your heists, because after each heist, it would simulate a detective system that tries to find proofs to know who and your crew are, trying to find where your hideout is.
And of course, the cherry on top, make everything in this game interactable with the already polished Uplink hacking system, having a crew member in the van with a laptop, and allowing you to hack all the building as you would on Uplink.
Don't get me started on online co-op possibilities
it's a big shame this game never came to existence .. i was very much waiting for it! I really loved `They stole a million` back those days and there was a version of `Rainbow Six` where there was a planning phase before the mission and I thought this game would take it a step further, but maybe there is some hope :)
I encourage anyone who is interested in this game (especially what they show starting around the 20:30 mark) to give Invisible, Inc. a try. Perhaps they mentioned it in the video but I haven't watched it fully yet.
It is really a shame that all this beautiful work amounted to nothing. I'm sure that you have heard this before, but still: consider this a request for releasing the source code somewhere so people can look at it and maybe make something out of it.
I had no idea that you guys got so far with this. I only ever saw the procedural city generation which I already thought was pretty mind-blowing. Would love if you guys picked this up again. Definitely feels like the Frozen Synapse 2 devs may have also gotten some inspiration from you guys! ;)
Sad to see this series come to an end, excited to see your next project.
When it comes to procedurally generating a "dungeon" with layers of locked doors/challenges leading to a goal and ensuring it is always winnable/connected up correctly etc, I think Joris Dormans of Ludomotion (Unexplored) is a good person to look to for how to make that work. It's hard for sure if you want anything complex (just use a 2d grid :)), but doable. Defiantly a major investment to develop the tools for that, so understandable that you didn't after working so long on this already.
I've been super inspired by you guys, ever since Uplink. I'm messing around with ideas for a hacking game myself and all the lessons you have learnt and shared in this series have been super insightful.
I tried implementing that paper way back in the day as well. If I remember correctly, the original implementation used L-Systems with actual strings that were later converted into a street graph. To make it more efficient, I made my version work on the actual object instances, so that I could later interpret them as the street graph without having to use strings as an intermediate representation. I never implemented anything past the street level, though. During that time, I stumbled upon Subversion and due to the shared interest in procedural city generation, have been a follower of Introversion ever since. Seeing this canceled was a huge bummer for me.
It's a great ambition, and perhaps PA was so good because of the lessons learnt along the way. I really love how this series has celebrated failure - it's a vital feature of success. And with 9 fantastic failures under your belts, this new project could be one hell of a success.
This series is criminally underrated. It's so insightful and wholesome, and gives a real window into the mindset of an interesting gaming company. Looking forward to see more details of the new game.
I've been waiting for this day for years.
Simulating things down to their individual parts is something that excites me as the player a lot. Both as a "Wow, that's really cool how it's designed" and "I wonder how many different ways I can interact with this"
Like the camera, for example. You could disable the motor to create blind spots, mess with its lens and narrow its FOV, disable or reroute its connecting wire, etc. People can get really creative when it's a simulated system of individual parts.
What about...slow down the movement of all cameras by 20% so now you have to move in a specific way to make sure no camera sees you? :) this kind of gamedesign is amazing to me, I wish there were more games like that
You mean I can finally play a version of Subversion!? I have waited for this since the very 1st video was released on it. I love this concept!
I loved following the development of this one as it was happening. This brings back fond memories.
This plus Uplink and Watch Dogs could make for one really great game.
I wish this would be given another chance. I was so hyped for it. I don't care if the ambitions are greatly reduced, I just really like the aesthetic and concept of an expansion of Uplink in this direction.
Chris, you need to take your random city generator and run to the Cities Skylines developers. They'd love it
I cant believe its been 5 months, I often come back to this video. Those early tech demos are/were a huge reason I got into game design and my love and respect of coding. Its still so impressive all these years later.
I know you guys listed the reasons this idea is totally broken and would never work, but I implore you, of all the demos so far, this is the one to go back to.
You can totally do this, I know it in my bones this can be done
This was the game that put Introversion on my radar :P
Still sad nothing ever came of it. This sort of stealthy-messing-with-the-systems-of-the-game is by far my favorite shit to play around with. The Hitman series is after all my favorite series just for that reason. Hitman on a larger scale with crappier graphics to be able to have it all running... I'd be down with that :P
Such a shame that this was abandoned, this was looking really promising.
You never _really_ left =)
*But*, a humble request: Please make it somewhat early access and let us enjoy your monthly videos again! It's ... idk, I said it probably before - but this format just works for me. Idk, Introversion games (even the "failed" ones here) are unique and special somehow. They all are more than the sum of their parts, really. Anyway, _take care guys_
Shows a lot of technology that has still not been implemented in other games or is just now getting partial implementation. So much potential. Maybe some day there will be a sandbox game with full building interiors.
22:08 the way to "square that peg hole of procedural vs authored", in my opinion, is making the level generator "scriptable". You need to be able to get to a point where you can tell it "make an office level with three levels of interconnected rooms with doors. Inside the last room there should be a hardened laptop of value X" with code. It isn't impossible, but it requires a lot of "solving problems that sound easy for humans but are very hard for computers".
This is a fantastic concept that would make a great game. Looking forward to your new project. Keep at it team.
This video was very enjoyable. Thank you very much for this series and all the insight into your process and for the prototypes you've provided through out the series. There is a wealth of game dev information in those prototypes and I even as a non game dev can appreciate that.
I never knew about Subversion and only found out by following an obsolete link to your website which showed me the old design and at the bottom I saw Subversion in between the rest of your works and I had to click on it. And so I found my way here and found out there was going to be a sequel to Uplink, my favorite hacking game.
Your 3 original IPs had a very unique and intrinsic charm to them that are still unforgettable to this day and I bet Subversion would have been the greatest of the bunch.
I still hold out hope that someday Subversion will see the light of day as a fully realized game because it's a great concept. The hacking gameplay of Uplink + what you showed in the demo would make a great game.
I feel like nearly 10 years of me wanting to play this has past already, and there is going to be 10 more years of that ahead of me.
I don't know what this "CrystalA" drug you keep talking about is, but I definitely want more.
Big thank you to Chris and Mark for a fantastic series. I was hooked on the first one. Introversion has always been and interesting company and having this insight into your mindsets has given me a window as to why that was. Looking forward to this new project.
Chris you've had enough Delay with this game,🤣 please finish it, it's a great game!💪
Seeing the growth of prison architect I know you all can do anything you put your minds to it.
This series was a really fun and interesting journey in the mind of game developers>
Honestly enjoyed all of these videos !
Thank you for the great content guys.
I hope y'all are proud of yourselves! Even though none of these games may not have "qualified" to become full games in their own right, each one demonstrated how much talent, skill, and passion you guys have for making top tier videogames. It's incredibly exciting and inspiring to watch! Looking forward to all your future successes 👍
Excellent and interesting video. Even though you're explaining how and why it died, I felt the rush I had 10 years ago after discovering uplink and then finding out you're working on a heist game!
It looked cool, the mechanics were amazing as they were complex, and there is a part of me hoping you find that "Eureka" moment to pull it all together and jump back on Subversion (doubtful but I live in hope). Thanks for the hour of entertainment today, and for years spent enjoying amazing concepts like Uplink and Subversion
This seems like an incredible game, would have loved to play this (and still do).
Getting serious hitman vibes from the bank heist! Very cool.
Looking at this, I think Subversion could be described as a level generator for a Hitman game.
Every time I see that city generator I'm amazed by how cool it looks.
That's simple unbelievable and fantastic, what you accompished!!!!
Loved the series! What a long trip through memory lane. Cant wait for your next upcoming game!
I'm glad you guys shared this in earnest. Subversion was among the first games I really got into as a kid, and I followed you guys all through Darwinia and Defcon. I was awesomely excited for Subversion, and sad to see it go. It's very cool to see the seed of Prison Architect form and become something to. Good show.
These demos are awesome! It's exactly the kind of game that I'd love to play
I love that simulation of things that you guys go with. Its the sort of thing I'd love to do if I could program. When stuff is systemic rather than arbitrary it means you can interact with it. Often indirectly. Which is fantastic.
Whatever is your next game, I hope there will be a devblog on it at some point, these are hilarious. Perfect ending for this series ! Cheers !
This is bloody brilliant! Even for now adays. Truly impressive on every level.
I landed on this video be a use I'm trying to figure out how to generate cities, and stuck around just consistently being more impressed.
You should release the game open source and let the massive game dev community take a crack at it with the technology that's available now.
Oh, the video ended with yall saying you're going to release the build... Well, that's simply charming. That's incredible. Brilliant!
Excellent video series guys, thanks very much for putting it together. The city generation for subversion was so beautiful it would be great to think that there is still some potential there for it to be used in a future game of yours. When you come to your next game adventure, I hope too that you'll give us more development videos because they are really interesting to watch even when i wasnt always keeping up to date with trying out each of the PA alphas myself.
This is mindblowingly cool. Everything about it. I've been surprised multiple times while watching this. I feel that you have shown too many impressive things in this video to call it quits on Subversion. It's almost like you're doing a disservice to gaming by giving up on developing this game.
I understand what you mean when you say that it's hard to merge the hand-crafted with the proc genned. But you should really take note of what Tarn Adams has said about procedural generation: when making proc genned systems, don't try to make huge complex systems. Make multiple small units that work independently and then connect them together to make a whole system. That's how he managed to do such a good job on his own game.
This is really really incredibly impressive given the times in which all of these demos were created. It's impressive even in 2021, to be honest..
I've got a box set of Darwinia on my shelf here --- the green DVD case (although it's actually a CD inside). I've even still got the printed manual with the fake newspaper articles in it and the slip-in note explaining where to go for help if it didn't install! Which, IIRC, it didn't due to early game bugs. I must play it again; it's still a fantastic game, it runs really well on modern systems, and you don't need to use gestures to command units any more.
So you mentioned in the previous video that you like designing games with interactive systems in them (like in this one - the security wiring and all that) and were looking for game ideas. With your experience with Prison Architect, it wouldn't be too hard to imagine a game like Space Station 13 (maybe the Colonial Marines variant) would be something up your alley. Half of the fun of that game is the systems that interact with one another and weird shenanigans you can do (like killing people with doors because you rewire their logic). It would be lovely to see a modern take on it that doesn't rely on a 20 year old single thread engine...
Oh my.. i'm amazed ! Your work was years in the future ! Shame you never manage to finish it so we could enjoy it !
I just started watching Money Heist on Netflix and remembered about this game, I just checked the introversion website to see an update today, what a coincidence!
Amazing series! Thanks for bringing us joy from failure, and great insight into game design. :-)
Looking forward to your upcoming title.
I really enjoyed this series. Thanks guys!
You guys make great games, many of these fail masterclass games have been amazing concepts. Until prison architect I thought I’d never heard of you but I loved defcon as a kid and enjoyed darwinia. Prison architect is one of my most played games and I can’t wait for your next game, I’m willing to wait tho
It's a great pity that Subversion was never released. The concept is sooo cool.
This is insane, I love it, and I so wish this was a developed game! Y’all are amazing!
You can tell how old these demos are by the fact that offices where modeled with individual rooms and dividers and not just a giant open office
I disagree - The open office layout standard predates even the earliest prototypes of this game by a decade or more. The 90s had them everywhere.
Yeah in the UK I've never worked in an office that wasn't open.
Reminds me of "The Sting!" from way back, just procedurally generated. Which kinda describes the ideal game for me. When are you guys coming back? We've been all waiting for the next big thing from Introversion.
"Prisons are lovely" (Chris Delay, 2021)
Oh man here it is. The one I dared not wish for. What a shame this one didn't come together, I do remember the impression the Dev blogs left on me after playing Uplink for the first time.
Excited and nervous to watch this one lol
Thief games would be amazing with a procedurally generated open world!
Guard is peacefully sleeping while gushing out gallons of blood. Let's hide the body while the floor is painted red in blood.
thanks for the amazing series! was great to watch
Damn, I wish there would be more of such games. I loved to play games like "The Clue!" back in the 90s!
As someone who has worked 10+ years in CIT, and visited about half the banks, cashpoints etc in this country.. this is 1:1 realism
Never thought I'd see the day! Never clicked a video so fast!
Its the final one you've PREPARED
I enjoyed the series a lot and wish you all the best for your new project! :)
Great series, thank you for sharing all that guys.
I still replay your first video on Subversion while playing the Uplink music
Everything you guys talked about here sounds to me like stuff that could absolutely be done now, it's just that you were making it in 2007, with the resources of a indie developer. It's still a game I would absolutely love to see you folks take another crack at.
Amazing. Thanks so much for doing this series.
Very interesting, very inspiring, very funny. :D
I still would want to see this, just because of the amount of interconnected systems.
Games like Space Station 13 or even Rimworld are still very popular games because of the huge amount of small, interconnected systems.
I'd still love to play Subversion fully developed, I was so hyped back then and I am still today.
I miss Subversion, even though I never played it...
the city generator is so beautiful
Chris: well we built this city generation algorithm
Me: oh that look really cool
C: and we wrote a paper on it
Me: oh that's pretty cool
C: this was before unity and unreal
Me: WHAT THE FUCK
C: and this is our own 3D engine with quarter of a million polygons in 2006
Me: WHAT THE FUCK
C: and the buildings would procedurally generate offices and stairwells and people with a day/night cycle
Me: OMG CHRIS YOU ARE A GOD
Honestly I agree with Mark too in that it was way too complex to limit it to a proper well designed game but omg this is god-level programming. Too bad he didn't get to the people and day/night simulation
It wasn't his paper; the paper was written by someone else and Chris implemented the algorithm.
Don't forget the physics engine!
Thank you for this series! This was a ton of fun!
Loved and love the proc-gen cities. There's no sense of homogeneity to the cities it creates.
If each floor was a Frozen Synapse level .. well, you'd maybe finish one skyscraper, but it would fit the style.
I love the procedurally generated cities and the procedurally generated buildings. It’s amazing technology but I agree that the environments will all end up feeling too similar. A bank is a bank is a bank. And procedurally generated puzzles is at odds with the realism goal. I got a great kick out of your story of how you abandoned procedurally generating a prison and instead crowd sourced it and found a new, unexpected game in Prison Architect. I wonder if you could crowd source the generator rules to your players somehow.
I remember this! I think some version of the city generator was in a humble bundle at some point? I remember hearing about it more fully and thinking the idea was very cool!
50:50 this game would be perfect if you opened up to the players to design and upload their own levels maybe
Great series, thanks so much for sharing
I've always wanted to have a "fable" or "Skyrim" game which was properly simulated. they put in a bit of stuff to make it feel like a real place but it never really feels totally real. you could totally adapt your simulated behaviours to create a working village. then your actions would always have direct consequences to the world.
Also, kind of hungry for you to do one of these videos but with Josh Parnell and Limit Theory. Would love to hear the three of you comparing and contrastng.
I remember this city generator, I believe it was distributed as part of a really, really early humble bundle or indie royale bundle as a tech demo.
I'm an aspiring game dev and your series has been incredibly interesting. I recall you guys saying something like you want to get your videos out in the algorithm more, maybe try a video like "indie game Devs react" to get eyes on the videos and I'm turn watching your other videos. Just an idea :) Ps I love your released games!
I remember these tech demos but I didn't know they were intended to be anything more than just that- tech demos
You could have used opposing AI to determine the levels themselves. For example, for a bank, you would have specific objects that would have worth (cash) and you would pit AI vs AI where one would try and create a layout that would protect it and another would try to get it, while a third is a "neutral" AI that must be able to perform business within the space provided. Systems are added and deleted as necessary to optimize the space, but leaves remnants to make it more realistic (like maybe fake cameras or extra wires in the server room). Objects themselves would have built-in use cases to take care of actions...for example, a teller desk would require a teller and small amount of cash on one side while forming a line on the other side, requiring room and the ability for the employee to get to it.
Just an example....but everything in the world has use cases that can be given reasons for existing to AI.....maybe the protective AI needs to also maximize decor of the neutral AI as part of the optimization pattern.
the 4 agents style reminds me of good old Syndicate
Whoa, awesome, I've been waiting for this
I was so hoping this will end up being a game...