Professor Jocular explains jokes.
This would ordinarily ruin the joke, but @ProfJocular is a bot. As such, its ability to understand humor jumps between the sublime and the absurd, sometimes making it even funnier than the original tweet it is trying to explain.
Dungeons & Dragons inherited random tables from wargaming, but took it much further. As noted in Playing at the World–a history of the development of D&D–wandering monster tables, such as the one above from B2: The Keep on the Borderlands, greatly expanded on the mechanic from Outdoor Survival. Every turn in the dungeon, there’s a 1-in-6 chance of meeting a wandering monster. Further checks determine which kind of monster, if it’s surprised, and so on.
And procedural generation wasn’t restricted to the monsters you might meet. Random dungeon generation has been a part of D&D for a very long time. In 1975, the first issue of Strategic Review (TSR’s precursor to the Dragon magazine) features a random dungeon generator for solo adventures. In Playing at the World, author Jon Peterson draws a connection between these random tables and the early dungeon crawlers on the PLATO system.
“The moral of this story is that random numbers should not be generated with a method chosen at random.”
–Donald Knuth, The Art of Computer Programming, Volume II
We use a lot of random numbers in procedural generation. You’d think that it’d be pretty easy to get some good ones. Just grab whatever random generator and get to work on making fun stuff. But some random generators are better than others, and how the random generator is seeded and used makes a big difference in the results you get out of it.
Rune Skovbo Johansen has written an article on generating repeatable random numbers, with a visual look at how to get good randomness. Check it out:
http://blog.runevision.com/2015/01/primer-on-repeatable-random-numbers.html
Sometimes bots get in trouble.
Olivia Taters is an accidentally-created teenage girl who lives on the internet, in the form of a Twitter bot and occasional political pundit. She’s also recently got her Twitter account suspended, which prompted a #saveolivia hashtag campaign, complete with protest songs .
A look behind the scenes of the procedural dungeon generator in Galak-Z.
Bots aren’t just about curation. They can also create their own content. The Generate A Cat bot pairs the cat generation of G. P. Lackey (who has a Tumblr) with the bot-ness of Bronson Zgeb (who also has a Tumblr blog).
There are many procedural generator projects that exist on their own without a context like a game. Putting one in a bot gives it a space to exist where it can find an audience. A weakness of procedural generation is that if you consume the content quickly (such as if this was a desktop toy that generates cats) you’ll rapidly see the patterns and lose interest. By placing the generation function beyond your control, so that the bot chooses when to display a cat and not you, it maintains that interest longer.
I found a great startup you should know about! They have a really slick website that looks just like every other startup website you’ve ever seen. They’re really redefining the paradigm with their cloud-based approach to delivering undeniable, resilient, beautiful platforms. Not to mention their partnerships with Societystr, Unpackify, and Changeme.
Humans are patterned creatures. We follow fashions, we form tribes. Creating a new pattern is hard, but following an existing pattern is something we teach machines. Distilling a pattern to a repeatable algorithm is an effective way to create a new procedural generator. It can also be an artistic statement: writing down the pattern can help us to understand the pattern and comment on it.
And next time you hear about a new company, maybe you’ll stop to wonder if it was procedurally generated…
The Startup Generator is a collaboration between Tiffany Zhang and Mike Bradley.
Peregrin (2014)
An autumn pilgrim
Speaks the essence of procgen:
A song just for me.
Created by Aeryne Wright for PROCJAM, Peregrin is a pilgrimage through an infinite forest.
Dungeon Crawl Stone Soup (2015)
Like the other major Usenet roguelikes, Dungeon Crawl has a ridiculously dense feature list. Beginning as Linley’s Dungeon Crawl (first released by Linley Henzell in 1997) Crawl has remained popular enough to continue to be updated by the Crawl devteam.
Crawl has lots of procedurally generated systems, but the interesting one we’re going to talk about today is its map generator. Today, Dungeon Crawl’s generator use a mix of several general layout generators combined with vaults, which are partially prefabbed sections of content that connect to the rest of the dungeon. At this point, there are thousands of vaults, ranging from tiny staircases to the massive end-game Zot:5 vaults.
This flexibility lets Dungeon Crawl get a lot of texture into various parts of the game. The game is about tactical control of the terrain, so a level of tight corridors feels very different from one that has wide open spaces. Compare the screenshots about that showcase some of the variability that the main dungeon maps can have. By mixing in pre-made content in ways that mesh with the rest of the level, the game can take advantage of the best of both approaches. One side-level is a portal to a labyrinth that changes as you move through it, highlighting the versatility of the game.
Creative AI: Procedural generation takes game development to new worlds
An article by Richard Moss about the contemporary use of procedural generation in games. Features interviews about ANGELINA, PROCJAM, GameForge, and other fascinating projects.
http://www.gizmag.com/creative-ai-procedural-game-development-angelina/35874/