The Brogue procedural level generation described here is a more sophisticated algorithm than the ones I've studied. The algorithms I'm familiar with do only one top-down pass to generate the level, with only minor adjustment afterwards. Brogue on the other hand (mostly via the "machine" concept) is iteratively rewriting the level after the initial pass, and in highly procedural ways by defining special rooms using conditions instead of premade templates, with level validity ensured after each step by algorithmic checking. The tricky thing about doing generation this way is tuning the machine requirements so that, on one hand, they actually are satisfied often enough for the feature to show up commonly, but, on the other hand, are strict enough that coding the rewriting part isn't too difficult.
This guy is awesome. He made one of the best games ever, totally free. I tried to donate to him once and he turned me down. Clearly he made the game because he wanted to make something fun for people and nothing more.
if he said 20,000 at 20k views someone would have freaked out if he said 10,000 at 10k views someone would have freaked out if he said 50,000 you wouldn't have just freaked out right now starting to get it?
I've been dying to find out what Brian Walker has been working on since Brogue, and I'm thrilled to finally know. Applying the kind of brilliant level generation that Brogue had to a platformer is a great idea. I'm looking forward to it!
I've been binging all roguelike/procedural generation vids I could find on youtube and this one is hands down the most informative/inspiring/entertaining. Cheers!
if he said 20,000 at 20k views someone would have freaked out if he said 10,000 at 10k views someone would have freaked out if he said 50,000 you wouldn't have just freaked out when it was 30k (which btw its now 48k) starting to get it?
Very informative talk as per usual! I do hope the platformer game works out and doesn't crash and burn like the recently released Chasm. Its looking great so far!
They have the advantage of engaging player's imagination. If you used the same white-walled cubic room with props in it multiple times in a 3D game, it would be jarring. But the same rectangle of letters and walls can be a gloomy holding cell, a cave, a small shrine - just by virtue of changing few ASCII characters and using different colors / flavor texts. Player's brain replaces all the fidelity the games might lack, drawing from a huge database of fantasy / medieval tropes that they learned throughout their lives
@@IDontReadReplies42069 I will be probably get flamed or hated for this but rogue type games are terrible looking - yes theres interesting code behind it but graphically pleasing they are not. Ironically you dont even need much graphical prowess to turn a rogue like game which are lets face it niche into world beating games Minecraft and Terraria are examples of this
@@skyblazeeterno "I will be probably get flamed or hated for this but rogue type games are terrible looking" not flamed or hated, but definitely laughed at for talking about an opinion like its a fact.
"Hi everyone, uuh, I'm Brian Walker, uuuh, I made, uh, Brogue!" Best introduction ever
"and..uuhh...I was trying to think what I should talk about today"
The Brogue procedural level generation described here is a more sophisticated algorithm than the ones I've studied. The algorithms I'm familiar with do only one top-down pass to generate the level, with only minor adjustment afterwards. Brogue on the other hand (mostly via the "machine" concept) is iteratively rewriting the level after the initial pass, and in highly procedural ways by defining special rooms using conditions instead of premade templates, with level validity ensured after each step by algorithmic checking.
The tricky thing about doing generation this way is tuning the machine requirements so that, on one hand, they actually are satisfied often enough for the feature to show up commonly, but, on the other hand, are strict enough that coding the rewriting part isn't too difficult.
This guy is awesome. He made one of the best games ever, totally free. I tried to donate to him once and he turned me down. Clearly he made the game because he wanted to make something fun for people and nothing more.
In case you heard it around 27:50 and were wondering what the "minimal cut" algorithm was - en.wikipedia.org/wiki/Karger%27s_algorithm
19:20 "And a 30 000 who watch on youtube"
HOW DID HE KNOW
I was wondering the same thing
if he said 20,000 at 20k views someone would have freaked out
if he said 10,000 at 10k views someone would have freaked out
if he said 50,000 you wouldn't have just freaked out right now
starting to get it?
I've been dying to find out what Brian Walker has been working on since Brogue, and I'm thrilled to finally know. Applying the kind of brilliant level generation that Brogue had to a platformer is a great idea. I'm looking forward to it!
I've been binging all roguelike/procedural generation vids I could find on youtube and this one is hands down the most informative/inspiring/entertaining. Cheers!
Wonderful talk! I'm just getting going on programming my first roguelike, and I found this both accessible and inspiring.
What happened to the platformer?
how did he know it would have literally 30,000 views?
if he said 20,000 at 20k views someone would have freaked out
if he said 10,000 at 10k views someone would have freaked out
if he said 50,000 you wouldn't have just freaked out when it was 30k (which btw its now 48k)
starting to get it?
Great talk!!!
Very informative talk as per usual! I do hope the platformer game works out and doesn't crash and burn like the recently released Chasm. Its looking great so far!
I fuggin love Brian
I'm trying to find more information about the roguelike platformer but with no success - anyone have a link?
not gonna lie, roguelikes are surprisingly varied given the limitations they work in
They have the advantage of engaging player's imagination. If you used the same white-walled cubic room with props in it multiple times in a 3D game, it would be jarring. But the same rectangle of letters and walls can be a gloomy holding cell, a cave, a small shrine - just by virtue of changing few ASCII characters and using different colors / flavor texts. Player's brain replaces all the fidelity the games might lack, drawing from a huge database of fantasy / medieval tropes that they learned throughout their lives
limitations? bro, having to work on graphics and in general having graphics is the real limitation.
@@IDontReadReplies42069 I will be probably get flamed or hated for this but rogue type games are terrible looking - yes theres interesting code behind it but graphically pleasing they are not. Ironically you dont even need much graphical prowess to turn a rogue like game which are lets face it niche into world beating games Minecraft and Terraria are examples of this
@@skyblazeeterno "I will be probably get flamed or hated for this but rogue type games are terrible looking" not flamed or hated, but definitely laughed at for talking about an opinion like its a fact.
the platformer part of this was interesting and at least looks graphically appealing but its kind of Terraria territory not that is a complaint
such a cutie, smart too!