I have a huge Minecraft world I’ve been working on for over a year… but I got kind of bored with it. I played Skyblock again, and then decided to start yet another world. Since the idea was to occupy the gap before version 1.18, I called it Interimland.

Bases are hard to show off in still pics, so I decided to make a video about it:

Oops– I just noticed it’s only 360px, though I recorded it in 1920×1080. Probably I rendered it wrong, though since it took an hour to upload, I’m not going to correct it. (Edit: it’s fine.) Also, I really don’t like hearing my own voice, but I can’t really change that, so you’re stuck with it.

The most interesting bit, perhaps, is the ravine build I start out with. I turned it into the kind of cité-puits that Moebius used to love drawing. Or the beginnings of one; it could use multiple levels.

I find that I’ve been doing more and more automation. My base currently includes these things:

  • AFK mob farm
  • regular old mob farm
  • XP blaster
  • cow & pig farm
  • wheat/carrot/potato farm
  • chicken farm
  • kelp farm
  • sugar cane farm
  • cactus farm
  • dripstone lava farm
  • concrete maker
  • apiary

All this produces so many emeralds I don’t know what to do with them all. I used to buy glass and arrows, but the mob farm produces too many arrows and excavating desert for the ravine city gave me an excess of sand (for glass). So I mostly buy XP bottles for the XP blaster.

Really, at this point villages kind of make Minecraft too easy. There is always something to do, but you can get just about everything there (diamond gear, colored terracotta, arrows, glass, quartz, spell books, blank maps). When 1.18 comes out in a few days, I may try avoiding villages, at least for awhile.

Bester: Virtual Unrealities

Alfred Bester is my favorite classical sf writer– see my review of The Demolished Man here. Mostly this is because he’s a great storyteller and an energetic worldbuilder… his novels are some of the few which don’t contain monocultures, which show off the same sort of diversity and weirdness we find in the real world. This book collects some of his best short fiction.

It’s also because, to be frank, a lot of classical sf was distorted and deadened by a reactionary streak… mostly due, it appears, to John Campbell, editor of Astounding. Robert Silverberg, in his introduction, describes how Bester ran into this. After a break from sf, Bester returned in 1950, sold one story to Campbell, and no more. Campbell, as Silverberg says, was then “obsessed with Dianetics”, the precursor to Scientology, and Bester hated this. But really, once you’re aware of it, it’s hard not to see that reactionary streak. Mostly it’s expressed as a belief that an educated technocratic class, or maybe supersmart mutants, should control the world: Heinlein’s superscientists, Asimov’s Foundation and robots, Poul Anderson’s time police, Van Vogt’s Weapons Shops and Slans. Bester is one of the few to resist the idea. The Stars My Destination is explicitly populist, while Demolished Man is simply too chaotic a world for anyone to control.

However, this collection highlights another side to Bester– a darker side. Many of the stories are more horror than sf, and end unhappily. Two are about the end of the world; quite a few are about some superhuman ability that causes more trouble than it’s worth.

Bester decorated his works with the usual high-tech spaceships and such, but it’s clear that technology as such doesn’t interest him. The science he’s most interested in is psychology. (Which fits in with the novels: Stars is about teleportation, Demolished Man about telepathy. And both are particularly interested in how society should handle disturbed individuals.)

A glance over some of the stories:

  • “Disappearing Act”: a pretty sharp satire of 1950s conformity and the Cold War. The particular sf idea is cute but not deep, but it allows Bester to show how a society can lose the very things it claims it values.
  • “5,271,009”: An extremely surreal story, told with Bester’s characteristic stylistic gusto. In brief: Bester warns about indulging one’s childish fantasies– such as being a superhero or the last man at the apocalypse. The moral gets a little more bite if you imagine it aimed at his fellow sf writers.
  • “Fondly Fahrenheit”: the story of a murderous android. This one has been widely anthologized, but it’s rather typical of the short stories, which often take their idea inexorably to a horrible conclusion.
  • “Time is the Traitor”: here the idea is that one man, named Strapp, is a Decider. Things in his multistellar world are too complex for individuals or even AIs to figure out; but a man with abnormal intuition can decide things anyway. A neat idea presented with satirical verve (the rituals of Decision, and the nature of Strapp’s entourage, are lovingly described). As usual, things end badly.
  • “The Men Who Murdered Mohammed”: also widely anthologized. Bester uses time travel in several stories– mostly as a way to parody human foolishness and the cussedness of the universe.
  • “They Don’t Make Life Like They Used to”: one of the end-of-the-world stories. Probably more interesting for its slice-of-post-apocalyptic-life portion than for its final revelations.
  • “The Flowered Thundermug”: I always like a heist story, and this one features two very accomplished heisters. The satirical world– it’s another post-apocalypse– doesn’t really connect.
  • “Galatea Galante”: About the creation of an artificial, perfect woman. Bester realizes and underlines how creepy the idea is, but still can’t really get past a very 1950s-male idea of femininity.
  • “The Devil Without Glasses”: another very surreal story– the basic idea is that two opposing forces from dreams are affecting reality, one pushing for human liberation, one opposing it. This might have made an interesting novel, but in short story form all Bester can do is turn it into another pessimistic fable.

I’m probably not selling the book well. I enjoyed it a lot while reading it. Bester can write, and he can be scintillating without losing the thread of the story. And if you like dark fables, he’s got plenty of them, and they’re a bracing corrective to the techno-optimism of some of his peers.

Still, the novels are masterful; the stories are merely fun. If there’s a common flaw, it may be that Bester gets too caught up in the surface details: the dialogs, the little swerves of plot. So some (not all) of the stories feel like they go on longer than they have to, or take a little too long to make their point.

Another biographical detail: Bester took another long break from sf in the late 1950s– instead writing for a travel magazine, Holiday. It must have paid better than sf, since he stuck with it for years. Also, before he got into sf he wrote for comic books… I’d love to see an anthology of his comics work!

Middle East Construction Kit

It’s here!! Here’s my descriptive page; here’s the Amazon page for the softcover. The Kindle will be out soon— sometimes this is quick and straightforward, sometimes it’s a mess. There will also be a hardcover option.

Edit: Kindle version is here. Hardcover too– see the description page.

The cover is a little different— I corrected the centering and changed the font for “Middle East” to Sanvito— but I’m too lazy to fix the pic. I can’t add much to the description on, so I’ll just talk about process for a bit.

I got the proof copy last week, and made a bunch of corrections, reaching the point any author will recognize: thinking that this bunch of corrections is the last. But I’m also at the point where I’ve almost memorized the text, meaning that errors are likely to escape my eye. (My wife found a bunch of typos my eyes just blooped over for months.)

The back cover of the proof had a horrible typo: the snatch of Hebrew I used as an illustration was backwards. It turns out that Photoshop Elements cannot handle Hebrew (or Arabic). I had to use a bitmap instead. I hope the Kindle doesn’t have this problem…

This book took about three years. This is longer than the India and China books. You can mostly blame Covid for that— the library was closed, for one thing. Anyway, I think it turned out pretty well! (The book, not the pandemic.) There’s a lot of fascination in these sandy empires, and it was fun to get to know Semitic and Sumerian.

What’s next? An alert reader gave me a great idea: a book devoted to creating religions. I’m also still tempted to do a follow-up to this book, starting in the 600s or so, and covering Arabic, Turkish, and Persian. But for the immediate future, I want to concentrate on my Almea+400 project.

Why do food replicators suck?

If you like mega-nerdy discussions about impossible things, this Tumblr and Metafilter discussion delivers. I found myself convinced by entirely opposite conclusions several times. The basic question: is Star Trek replicator food as good as real food?

The originator of the discussion, Beatrice-Otter, thinks it obviously is:

In Star Trek, replicator technology is part of the same tech tree as transporters. Replicated food would be identical to the food it was based on, down to the subatomic level.

The discussion makes many good points.

  • Working from a fixed recipe makes every instance of a dish taste the same. That could get boring fast.
  • If you’re a foodie, the number of options for a dish just explodes. Precise ingredients, multitudes of obscure varieties for each one, their terroir, different cooking methods, seasonal changes, the taste imparted by the pan you use. Maybe even which microorganisms live in the local tap water…
  • Maybe the replicator is good with pizza, but does it do Ethiopian food well? To say nothing of all the hundreds of planets and dozens of species the Federation has run into. You can have a database that’s huge, yet really only works for a subset of people, probably the developers.
  • Everyone has their own individual associations and preferences. You could store the ratatouille recipe that brings Anton Ego back to his childhood— but visiting Anton Ego’s childhood isn’t the same for you and for him.
  • The ship’s replicators are military grade and just don’t have the options the crew had back home.
  • Humans are humans, and will find a way to complain even in a post-capitalist utopia.
  • Our senses are affected by ambience. Anton Ego’s ratatouille might be spectacular in a restaurant in Paris, bland and disappointing in a cold starship galley.
  • In ST canon, replicators can’t produce live animals. What about that Klingon dish that includes live ingredients? But you don’t have to invent other species to get this problem: I once saw a video about a Mexican street dish that included live insects. For that matter, yogurt is supposed to have live bacteria cultures…

Someone pointed out that once replication tech exists, people would deliberately mess with it— e.g. make a steak where some of the proteins are replaced by raspberry syrup. Or you deliberately play with glitches that would be impossible in any other cooking method. (E.g. imagine layering hot and cold levels within a pastry— each 0.01 inch wide.) Or a similar layering effect with chocolate and fruit.

To me, the killer consideration is data space. How much physical space does a recipe take up? We’re used to thinking of data as weightless: an e-book takes up way, way less space than a book. Thus we can easily imagine a replicator filled to bursting with variations on the pizza recipe.

But an e-book is negligible because printed letters are enormously huge at the molecular level. It really doesn’t take much to fool the eye. The stomach is another matter. You have to replicate the food way below the cellular level. E.g. you’re making a steak: you have to make all the proteins correctly down to the molecular level— if you make the proteins wrong, the food isn’t even nutritious!

A pound of steak contains on the order of 1025 atoms. We need to store the location and velocity of each one.

  • Yes, Heisenberg would have something to say about that; take it up with Star Trek.
  • Velocity? If you think “but the atoms aren’t moving much,” I remind you that electrons moving is what we normally call “temperature”.
  • Just in case you have some bright idea about storing info about multiple atoms in one atom using floating point numbers… atoms are quantized. E.g. the state of any electron can be given by two integers (spin and energy level). And these are not numeric registers: if you set an electron to a high energy state, it won’t stay that way, it’ll send off the energy as a photon.)

Anyway: to store the state of each atom, we have nothing smaller than an atom. The recording of the state of 1025 atoms will itself be 1025 atoms— a pound of computronium. Maybe more, since any data format has some overhead; plus of course you need the mechanisms to actually read the data. Now, do you still think that the replicator can store thousands of recipes, and thousands of variants for each one? The ship will end up being made of recipe blocks.

Fine, you say: we’ll compress the data. And I’ll say: there is no lossless compression. Every method you use will have some effect— and could change the taste. Make enough compromises and yeah, replicator food sucks.

Now, I’m happy to grant that compression can be done. But a lot of thinking about this will be due to analogies that don’t apply here— namely, our experience compressing visual and auditory data. It’s pretty easy to fool the eye and the ear with low-res data. How much of that applies to taste and smell is another matter. (A lot of biological processes, including these, depend on recognizing large molecules; you can’t depend on macro-level fakery. That’s one reason making fake meat is so hard.) And again, satisfying the digestive system is even harder.

Edit: Compression can be lossless if the data is strongly patterned. E.g. an entry in the Library of Babel that consists entirely of the letter “e” can be precisely described in a few words (I just did it). But complex biological material (a.k.a. “food”) isn’t strongly patterned in that way.

Suppose you can compress the recipe a thousandfold: you can store the recipe for a pound of beef in 1022 atoms. Great! Now you can store a thousand variants of the recipe in a pound of computronium, a thousand recipes still weigh half a ton. How many replicators did you want in your starship, by the way…?

OK, we compress it a millionfold! Just 1018 atoms per recipe. That’s just a pound of computronium for a thousand dishes with a thousand variants each. Good enough for government work! Only… come on. You can’t just say you’ll keep 1 bit in 1,000,000 and claim that the loss is undetectable. Some people will insist on this, but I call this an instance of engineer’s disease: a certain type of nerd thinks that their tech solution cannot fail, an optimism belied by fifty years of engineering experience.

(Also, please don’t quibble over the portion size. You could store 1 cm3 of beef instead and save a couple of orders of magnitude. Only, one, building food out of 1 cm3 cubes will definitely be noticeable. And two, I could easily increase thousand recipes by several orders of magnitude too— that was just a simplification for ease of calculation. The Joy of Cooking contains 4500 recipes, and it’s hardly the last word on food. And Star Trek ships include people not only from every human culture, but from other biospheres and other species.)

All this shed some light, I think, on why Star Trek transporters can’t be used to store multiple backups per person, create endless clones, etc. A person takes 1027 atoms to store— the same weight as a person. (You want that data compressed? And then you’re gonna cheerfully step into that machine?) This would explain why (as one Metafilter user points out) during one ST transporter malfunction, storing all the person-data completely overwhelmed the ship’s computers. Transportation cannot be a matter of reading the data, storing it in some mega-compressed form, beaming it somewhere, and 3D-printing it. The best way of storing those 1027 atoms is to use those very atoms. They’re transformed into plasma beams or whatever tech-gobbledygook you want, but they continue to be 1027 atoms, and they’re physically moved to the destination.

Oh, one more thought experiment for you. You go into the transporter carrying a portable food replicator— whose core is a pound of computronium. As we’ve established, every atom in a recipe is important: this data has been compressed as far as it can be. Does the replicator still work after you’ve been transported? Does it work well? Analogies based on JPGs won’t help you here!

Finally: this is all speculative and no answer, including mine, is the last word. But there is a conworlding lesson here, and that’s that though you can rescue a magical sf idea with more magic, you shouldn’t. A certain type of nerd just wants those replicating machines to work perfectly and keeps adding more impossibilities to keep the idea pure. But to get an interesting sf idea, you want limitations and tradeoffs and possible flaws.