Matt McIrvin's Freaky Alternate Reality

One of the Christmas gifts I just got was a Kindle Oasis, a member of a dying breed of handheld electronic device--the unitasker with a screen. It's just for reading (and buying, oh yes indeed buying) electronic books. It can't do anything your smartphone or tablet can't already do with Amazon's free Kindle app. And, unfortunately, in the way of such things, it's very much joined at the hip to the Kindle e-book store--the point is for Amazon to get even more of your money once you've got it set up. I haven't experimented with how easily it can be subverted to read other types of material, if at all.

Physical design

But since it only really does one thing, it can be physically specialized for that, and it is pretty exquisitely so. It has a touchscreen e-ink display, which is in some ways less capable than a phone screen in that it's slower and entirely monochrome--but it's lovely to look at, high-resolution enough that you don't notice the pixels, and unlike a regular backlit LCD or OLED display, it is reflective in nature and is perfectly readable and crisp even in direct sunlight.

It's cleverly designed to be used one-handed, more comfortably than a phone, except that it's a bit heavier. The device is nearly square, but its screen is the width and almost the height of a mass-market paperback page. The rest, to the left or right side (depending on which way up you're holding it), is a blank bezel area you can grab, with two physical buttons for going forward and back a page (which you can also do with swiping gestures, or by tapping the edges of the screen). The buttons are positioned so that you can operate them with your thumb if you're gripping the device with that hand. The device has an orientation sensor, and the display and the button controls will invert if you reverse it, so you can grip it equally well with your left or right hand. It doesn't rotate sideways--you read only in portrait orientation.

edit: There are a large number of reading options I hadn't discovered when I originally wrote this review, accessed through the "Aa" menu in the toolbar; it is actually possible to set the screen orientation to landscape, but it doesn't happen automatically as a result of rotating the device.

Most of the screen portion of the Oasis is very thin; the rest, undoubtedly containing the battery, is a little thicker, and the resulting ridge in back forms a kind of handgrip if you don't have a cover folded back over it. There's a cloth cover accessory, a little like Apple's iPad Smart Cover, which has a magnetic catch and will power down the device automatically if you close it. The cloth backing makes it easier to hold the device if you do have it folded back. There's also a power button on one end of this battery section, and a Micro-USB port on the other end. I'd have preferred USB-C, but I guess they haven't gotten with the times in this regard.

Light, text and power

The screen does light up, and to be comfortably usable, it pretty much has to light up. While the display is reflective, the "white" pixels are not actually a brilliant paper-white--they're more like a medium gray-tan, so the display looks great without self-illumination in bright daylight (in fact, it might be more readable than white paper under these conditions), but otherwise the reader's internal light is key. Fortunately, it's a really good light that illuminates the screen absolutely evenly with a comfortable brightness; you can adjust it so that it creates a convincing illusion of white paper. Most impressively, it's at least a two-color LED and the color temperature and brightness are independently adjustable. There are clever options to automatically throttle the brightness and make the light warm and yellow in the evening hours, much like modern phones have.

You can vary the font size over a fairly wide range--at largest, the letters are about a centimeter high; I've certainly known people with sufficiently low vision that this would not be adequate, but it will work for most. On the other hand, there are only a couple of sizes you can choose for the text of user-interface elements, which could be an accessibility barrier. You can apparently get it to do some screen-reading things over Bluetooth (see below). Out of the box, most books are readable only in Amazon's default font, a fairly good-looking serif face called Bookerly, but apparently it's possible to add fonts to the device over USB and use them instead.

edit: Again, there are options here I hadn't discovered in the "Aa" menu. You can actually choose any of nine built-in fonts, including sans and serif options and one designed for readers with dyslexia, and you can also vary the boldness, margin size, justification and spacing.

The e-ink display uses less power than a phone display, but Amazon's claim of a battery life of "weeks" is rather exaggerated--I think it assumes you'll be reading infrequently and won't be using the light, and, as stated, you will use it. A full charge on a new Oasis would probably last a couple of days of very heavy binge-reading, like I've been doing. The device, annoyingly, doesn't come with a charger (it does come with a USB-to-Micro cable). You can do some kind of syncing with your computer but I haven't experimented with that.

Sound, networking and software

The Oasis has no internal speaker, but can supposedly pair with a Bluetooth audio device to read text aloud or play Audible audiobooks (but I have not been able to get this to work with the one slightly janky pair of headphones I tried). There is no headphone jack; given the audiobook features I think they probably should have included one, but such is the way of the world now.

It downloads stuff via wi-fi, but I found that I could only get it to connect to the 2.4 GHz channel on our home router, not the 5 GHz. For downloading books, this is perfectly adequate. It's pretty easy to find the MAC address via the settings screens if you need it.

Amusingly, given that this is physically my first Kindle, Amazon identified it as "Matthew's 6th Kindle." I know why this is--I've actually had a lot of devices that ran the Kindle app, if you include not just phones but all the wonky old tablets I worked on a couple of jobs ago, and I think I at least set it up on several of them. Which is not to say they ran it particularly well. But, as I said, any one of them could technically do all the things this reader can.

It's pretty much just a book reader and Kindle store client; there's also a link to the Goodreads review/recommendation service. These functions are as snappy and responsive as they need to be. The device (or maybe just an Amazon account itself) comes with free Kindle access to a collection of dictionaries of a number of major languages, which you can download and then use to look up words while you're reading--this is surprisingly useful. It can also look up words in Wikipedia. There's some kind of "vocabulary builder" app that can quiz you with flashcards about words you looked up in the dictionaries, if you are so inclined.

There actually is a general web browser app, marked "experimental", but it is primitive and painfully sluggish, not good for much. The Oasis really isn't built to do this.

Reading books

This section was going to be at the top, but I didn't want it to be like those recipes that require you to scroll past a multi-page personal reminiscence to get to the ingredients list.

I find that I haven't been reading books a lot over the past couple of years, and I think it's partly that the attraction of "doomscrolling", as people call it now, is too great. There's this inexhaustible supply of outrage, horror and snark generated by the news, and everyone's two cents' worth of opinion on the news, that has great power to distract--and if you're waiting to get sleepy in a dark room, where someone else is also trying to sleep, that can be a terrible temptation that's actually easier to access than a paper book. I don't think it's been good for my mental health, which makes me think I should get back into reading more--reading books rather than the news or somebody's political blog.

Paper books are great, in some ways superior to any electronic device. But having a device that really doesn't do anything but read e-books, and is more convenient than a paper book in some specific contexts, might actually help. Even if it's a bit too closely tied to Amazon.

Well, I sprung for XPlane. MSFS2020 won't run on my Mac, though it might be interesting to check out the XBox version when it appears. But XPlane 11 does. It's a lot of fun, though I think anyone getting into it needs to realize what it is and isn't.



This is pretty much the opposite of the Google Earth flight sim. Which is to say, it treats the world scenery with some indifference, but this is the one for people who are really, really, really into the planes.



The cockpits are all modeled lovingly in three dimensions, and you can move about the cockpit and examine all the switches and gauges in detail; many of the doodads in there will actually be operative. By default the sim starts you on the runway with a full tank and the engine already running, but if you want, you can run through all the checklists and bring up a jetliner from completely powered-down (there's a version of this sim, not the standard retail version, that is certified for pilot simulator training--the key thing is apparently that it has to make sure it's always running at maximum FPS). You've got a mix of general-aviation aircraft, jetliners and fighter planes, an incredibly fun ultralight, a helicopter (haven't tried that yet), some exotic things like an X-15 and a Space Shuttle orbiter. For the latter things, there's support for planes carrying other planes (you can be the carrier, or be carried and have an AI take you up).



One problem if you're just using a single laptop monitor is that some of these control panels are physically gigantic, particularly in the big airliners, and in something like a 747 it can be hard to even look at the standard gauges while also keeping an eye out the window. You spend a lot of time panning around inside the cockpit with the view controls. The more modern planes have glass cockpits and you end up trying to interpret a little screen on your big screen. There's a HUD mode available with all planes (whether it would make realistic sense or not) that alleviates these issues, but I think one of my favorite planes in the program is the Stinson L5 Sentinel, a World War II-era military light plane, just because it's got a compact basic instrument cluster that is really easy to look at. People who are really into this stuff have elaborate multi-monitor setups that wrap around them almost like a professional simulator cockpit, sometimes with subsidiary computers just to control additional displays; XPlane supports a boggling variety of options, including VR headsets and using tablets to display control panels. I might try connecting an external monitor and using the main one just to display instruments.

Pilots seem to really like the flight model here; according to them, at least, the planes handle realistically. I wouldn't know, but the handling differences between the aircraft are palpable. I tried plugging in an XBox controller, and XPlane calibrates and handles it flawlessly*; a more serious user will doubtless prefer a realistic flight stick or yoke, but of what I've got, this is clearly the best way to control the simulation. The twin sticks give you independent rudder and aileron controls so you can practice coordinated flight; some buttons map to the throttle by default. When you have to push on the stick like a pilot would, instead of effectively controlling a virtual yoke with a keyboard or mouse, it does give you an appreciation of what trim controls are for.



What it's not is a sim for sightseeing of famous landmarks in the virtual world. It doesn't stream scenery like Google or Microsoft do, so the whole world has to be stored on disk, and the main way to scale back detail for lower-powered machines is to reduce the number of generated buildings. The scenery generally looks OK, but in a generic way. Cities will have more or less the correct street layouts, but be populated somewhat haphazardly; if you've got the scenery complexity set to low (which I think is the default), many towns will have streets but almost no buildings, and the world ends up with a barren, post-apocalyptic quality, almost Fallout 4-like. It looks better if you dial up the graphics settings (at risk of getting nagged that your system is too feeble to hit 60fps--at least you can turn off the nagging). But you'll still get a weird alternate-universe feeling if you try to go somewhere familiar to you, because almost everything is a generic object.

On the other hand, it's got a gigantic number of airports, basically every airport of significance in the world. Those are quite detailed--but with an eye to functional aspects a pilot would care about, not striking terminal architecture. Toodle around a major airport (the ultralight is useful for being a fly on the wall) and you'll see planes with all sorts of airline liveries, moving airport vehicles, etc. You can communicate with the tower, request and receive clearance if you want to play that way, get refueled and even pushed back from the gate by a tractor. (small correction: you can summon the snack truck and the pushback tractor, but not the fuel truck--you "refuel" just by editing your vehicle settings: in real life that takes a long time anyway.) All the navigation and ILS equipment is of course operative (even 1980s Flight Simulator had a lot of that). The night lighting of runways and taxiways is spectacular. But you won't find IAD's swooshing Saarinen terminal or LAX's Theme Building rendered recognizably. If you mostly want that, you might as well stick to Google Earth. Where this game gives near-pornographic aesthetic attention is the aircraft.

My impression is that MSFS 2020 is an attempt to bridge these worlds--it's got the accurately rendered planes in a mind-blowing streamed world. But even there, you may need fairly grunty hardware to get all the benefits.

* I also got around to trying the XBox controller with the Google flight sim, and it tried to handle the thing but there were hopeless calibration/centering problems that there seemed to be no way to address. Will need to stick to the mouse with that one.

I've been interested in the generally positive reaction to the new Microsoft Flight Simulator, though I don't actually own a computer that can run it--I'd probably have to bogart my daughter's souped-up Windows rig. Apparently some kind of XBox version is coming, though I can only imagine it'd be very stripped-down.

But the YouTube reactions to it are interesting: a lot of people just seem wowed by the idea of a flight simulator that can stream a whole world's worth of scenery over the net, in Google Maps fashion (it's actually based on Bing Maps, with a heavy overlay of procedural-generation trickery to fill in gaps in coverage).

The funny thing is that a much simpler version of that has existed for 13 years, for free!

You may have forgotten about it, but Google Earth used to be a standalone app (acquired from an outside company called Keyhole), and that actually still exists. It's called Google Earth Pro, but despite what that might seem to imply, it's a free download. The app seems kind of janky and archaic these days, but it uses the same data as Google Maps. And it has a flight-simulator mode, added in 2007. Originally, it was an Easter egg only accessible through a key combo (command-option-A on the Mac), but now there's a menu item for it.

No pilot or serious "simmer" (like the person who made the video below) would take it as more than a silly toy. There's no cockpit panel, just a simple HUD inspired by fighter planes; it seems to lack such luxuries as brakes of any kind or the ability to steer the plane on the ground (there must be some way to do the latter because they do it in the video, but I can't figure out how), your fuel tank is infinite, and it happily lets you taxi on water and fly through buildings. Sometimes you clip through the landscape and get a peculiar sliced-up view of the underside of the ground. You have two planes, a Cirrus SR22 or an F-16.

(edit: Shift + arrow keys seem to be the rudder pedals. There may not be any auto-coordination. Don't know if there's a way to control them with the mouse. I also found an ancient Google Earth Blog post that explains that comma and period are the left and right wheel brakes, which you can use to control or stop your motion on the ground--the documentation the app actually links to won't tell you that at all. They recommend using the brakes to steer, I guess because it's easier than explaining to their readers how to deal with the rudder.)



But on the basic "I can see my house from here" level, it's a terrific toy. I hadn't played with it much until recently, inspired by the hype for FS2020. I really like buzzing around my town and environs--it makes me realize that eastern Massachusetts is really kind of a lovely place, seen with fresh eyes. I'd always wondered what it would be like to take off from the little general-aviation airport down the street from here, and you can (in the SR22, at least).

Looking at Google's 3D scenery with the usual Maps interface, it feels like you're interacting with a very detailed miniature, like a model-train setup. Somehow, looking at the same scenery at the same level of detail with the flight sim gives you a much more visceral feel of moving through a full-sized world. It's fun to zoom around at Mach 2 in the F-16 at treetop level, presumably destroying the eardrums of anyone in the vicinity (the game itself is silent), though that will usually over-fly the streaming of the scenery so you only see low-resolution versions of the stuff; the slower plane is better for sightseeing.

You can take off from any level place you can click on in the map, but there are a bunch of presets. One of them, doubtless put there for old flight-sim fans, is Meigs Field in Chicago, the original starting location for Microsoft Flight Simulator and Flight Simulator II. But Meigs was abruptly ripped out years ago and the area converted to a park and sculpture garden. So the program will just let you take off right through the sculpture garden, which is surreal.

Once we were shut in I decided to revive my old bread-machine oatmeal-bread recipe, only to discover that (1) my bread machine was dead, and (2) my yeast was all dead. It can survive in the freezer for a long time, but evidently not most of a decade.

Instead of trying to repair the 20-year-old machine, I threw it out, and decided to get into doing it by hand. We managed to get this comically huge industrial-sized double package of Fleischmann's instant yeast from BJ's, so we have effectively infinite yeast at this point. Bread flour is hard to get, but I got some all-purpose and some whole-wheat flour. I don't have a bread pan so I'm making peasant loaves on a cookie sheet for now.

Instead of posting whatever recipe I used last, I'll be more useful and say that the recipes I'm converging on are all variants of this simple one:

The basic recipe

4 cups some kind of flour, plus more as needed (see below)
1/2 tsp instant yeast
2 tsp salt
2 tsp some kind of sugar
2 cups warm water

Mix all dry ingredients in a bowl. Add the warm water and knead it for about 10-15 minutes. Initially the mixture will probably stick more to your hands than to itself. Add small amounts of extra flour (a tablespoon or so at a time) and keep kneading until it doesn't. Eventually it will be an elastic dough ball with a nice doughy texture.

Cover that in the bowl and let it rise for a couple of hours or so. With basic white bread you'll see it double in size. Punch it down if you like, put it on the cookie sheet, shape it into a loaf, put some artistic slashes in the top and let it rise, I don't know, 30-60 minutes.

Preheat the oven to 450F. Bake the bread at 450F for 15 minutes, then lower the temperature to 375F and bake for 20-25 minutes, until the crust looks nice and browned. Take it out, let it cool for 15-30 minutes, then serve.

Yeast

I'm using instant yeast, which is the identical stuff also called "quick-rise yeast" or "bread machine yeast", because it's what I got. This stuff is convenient because you can just mix it with the dry ingredients and it will work.

You can likely use other kinds of yeast identically in the above recipe, except you can't just mix it in dry, you have to do the "proofing" thing that you can look up somewhere else, where you put it in water with something to feed it and make it bubble. I don't have experience with that.

You can shorten the rise time by bombing the mixture with far more yeast, but you get this obvious yeasty flavor that you've probably tasted in bread-machine recipes (you can see that I used way more in my old recipe). There's a point of diminishing returns for that.

You can use less yeast and wait longer, and it might do good things to the flavor--today I tried 1/4 tsp and a four-hour rise, but I'm not sure it made a big difference. I haven't tried retarding the rise in the fridge overnight, or any tricks like that (experiments will probably happen in the future).

Ingredient variations

This recipe works fine with all-purpose white flour to make a basic white bread--you don't need bread flour. You can substitute in wheat flour for at least one cup of the white and it works basically the same.

Today I tried adding about a quarter cup of oatmeal to the mix as well. That, or substituting in a lot of whole-wheat flour or other whole grains, is going to make the loaf rise less, I think because the bran breaks up the formation of gluten (everything I know about this comes from Harold McGee's On Food and Cooking). So the oatmeal bread will be a denser loaf; you might try waiting for it to rise longer, but in my experience with the oatmeal bread, even with my old bread machine, a lot of the rise you get is actually going to happen at the beginning of baking. With white bread, it gets more of its rising done before that and is fluffier.

The salt actually strengthens the gluten network, something I didn't realize until recently. According to McGee, this is because the ions neutralize electrically charged regions on the gluten protein that would otherwise make adjoining molecules repel each other. I'm not using a huge amount of salt, but putting in some does help the bread to rise and retain its shape.

You can use brown sugar in place of white; I usually do that--I think it tastes better but it doesn't make a huge difference. I'm using much, much less sugar than I used to with the bread machine. I suspect you don't actually need any in this recipe, but I haven't tried leaving it out entirely.

Adding other stuff

My old oatmeal-bread recipe was more complicated and involved milk and oil. I'm not sure any of that was doing anything good aside from affecting the flavor a bit. Putting in oil tends to have a similar effect to adding oatmeal, in that it weakens the gluten and therefore makes the dough rise less. It might increase the shelf life a bit and give it a moister texture. But I'm leaving it out now.

I actually have some wheat gluten that I bought ages ago. Adding a bit to the mixture can help with the oatmeal variant. Bread flour is just white flour with more gluten than all-purpose.

 In part 1 of this series of indeterminate length, I mentioned the Thue-Morse sequence, a certain sequence of zeros and ones, and the counterparts of it for more than two symbols. When I was messing around with Thue-Morse I came across a webpage by Zachary Abel that mentioned something remarkable about them: interpreting them as a turtle-graphics program gives you a thing that somehow converges to the Koch snowflake curve, the archetypical simple fractal. (Abel defines the Thue-Morse sequence as the complement of the way most other people define it, starting with 1 instead of 0, but for this purpose it hardly matters.) This paper by Jun Ma and Judy Holdener proves in some detail that this does in fact converge to the Koch snowflake:

This isn't the textbook way that you usually draw the Koch snowflake with turtle graphics; it's simpler. You don't even need a "right turn" rule. You don't need recursion (except inasmuch as recursion is built into the Thue-Morse sequence, which you could argue it is). You just have the turtle follow the sequence, and turn left 60 degrees when it has a 0, and move forward one unit if it has a 1. The turtle at first seems to be crawling around in drunken squiggles, but if you just let it go, the overall shape of those squiggles turns out to be the Koch snowflake.

And, actually, that's not even the only way to do it. It also works if you interpret the bits the other way around. Or if you tell it to turn 120 degrees instead of 60--the fine details of the path are different but it still converges to the snowflake. Or if you give it a bunch of other ways of interpreting the Thue-Morse sequence as turtle instructions. Abel's page explores all sorts of different cases that work.

Not every interpretation works. If you just put in an arbitrary number for the turn angle, most of them seem to just give a mess. Sometimes, that mess starts to resolve into a directional path on a very large scale and it's possible that you still get the Koch snowflake again, but I haven't been patient enough to tell. Maybe you get something else entirely! Here's 70 degrees:

On the other hand, some you wouldn't think work, do. 75 degrees is sort of a punk rock snowflake:



Sometimes symmetries get you. When I first tried this, it didn't work because I was using the same sequence interpretation I use for the Heighway dragon curve: move forward every time, and turn left if you have a 1 and right if you have a 0. That fails because of the symmetries of the Thue-Morse sequence: every four symbols, the turtle ends up moving in its original direction, and you just get a long line with aperiodic wobbles in it, not very interesting. That's true no matter what turn angle you try.

But if you tweak the rule a little to break the symmetry, so that it turns, say, 120 degrees left for a 1 but only 60 degrees right for a 0... then you get the Koch snowflake again. Only the small-scale squiggles are different.

The Ma and Holdener paper proves that this is indeed the Koch curve for the first rule I mentioned, and for the complementary case where the symbols are interpreted the other way around, and they speculate that there are many other related sequences that have the same property. Kennard, Zaremsky and Holdener then found a whole family of these sequences.

But I don't think there's been a comprehensive study of what happens when you change the turtle interpretation of the series, as Abel did and as I'm doing here. I suspect that in many cases, changing the turtle rule is equivalent on a larger scale to modifying the series instead, in some way that keeps it consistent with Ma and Holdener's conditions.


But there's more...

This is only the tribute.

I was fascinated by turtle graphics, as embodied in the computer language LOGO, which was intended for kids. This is a system where your drawing cursor, the "turtle", has a position and a movement direction associated with it, and takes relative angular rotation and forward movement instructions. It's good for drawing things like simple fractals and certain mathematical curves.

The first thing I implemented, I recall, was an interactive turtle graphics system where the user chose commands directly from a menu by number. But that was pretty limited for anything other than simple geometric doodling. What I really wanted was a turtle-graphics library I could call from BASIC. Well, in a language that doesn't even really have functions that is a kind of awkward thing to manage, but I did figure out that I could have a kind of named GOSUB by assigning a line number to a variable, so that was nice. I just had to put all my "arguments" in special global variables and then call GOSUB TURN or GOSUB DRAW, etc. Then I had a sort of BASIC program skeleton I could save on a floppy, load whenever I wanted and put my little application at the beginning. It was sort of a mess, but it worked. It's long-gone, but today I reconstructed something like it in a couple of hours.

So, having reawakened my long-dormant memories of how to make a usable Atari disk image for saving files, I spent some time actually messing with Pinball Construction Set, Bill Budge's 1983 game that introduced most of the world to the concept of a game editor. Is it possible to make a decent pinball with this thing, one that will be fun to play in 2019?

Amazingly, it is! The game's limitations become immediately apparent when you mess with it, and there are a few things I wish you could do that you can't, but I think the program's limits were actually a big part of what made it enjoyable for the general public. You can whip together a table that can do most of what Pinball Construction Set could manage in a couple of hours, and the rest is just subtly tweaking and decorating it. There are modern PC pinball editors designed to let you put together accurate simulations of complex modern tables, but to use them well you need to master half a dozen technical disciplines. There's not much of that here. It sticks close to the goal of being a fun toy. Everything in the game editor works through a simple joystick-and-fire-button GUI, maybe the first time many people ever encountered something like that. The beautifully written manual (a time capsule from the absolute high point of software marketing and packaging) is 12 pages of text and three cheat sheets, and it covers everything comprehensively, even giving design hints and subtly acknowledging some of the game's anomalies and quirks.



There's some graphical crud here that is probably from bugs in the application, which I haven't bothered to clean up. Also made no effort at all to decorate the table or the "backglass", which you can do; these are just the visual elements the game gives you. (Adding artwork to your pinball is one of the more cumbersome and unpleasant elements of the editor, unfortunately, and I didn't mess with it much. Editing the polygonal table geometry can also be a bit frustrating in this pre-mouse era.)

The first thing I did was to save a file with just my rendition of an "Italian bottom", that is, the more or less standard arrangement of slingshots, flippers and wire guides down at the bottom of the table. I figured I'd be going back to some variant of that a lot.

What I was going for here in general was nothing particularly original, but a homage to Ed Krynski's 1970s Gottlieb drop-target-fests like Big Shot or El Dorado. It seems like the first tables people make with PCS often clutter up the middle with a lot of bumpers and stuff, but I went for an open layout where the main goal is supposed to be bashing away at that bank of eight drop targets on the right (actually two adjacent banks of four). The tightly grouped pop bumpers in the upper left actually have a lovely Pat Lawloresque straight-through-the-bumpers shot off the right flipper, that also nails the spinner below them. That was an accident, and I ought to put some target in the path to make it worth more, because it's great. I put in a little diagonal barrier to keep the ball bouncing around in the pop bumper area once it gets in there, and then I put a couple of stand-up targets on the bottom to make it more than just a passive barrier, and because designers like Lawlor do that a lot.

The gizmo behind the small third flipper is a simple multiball lock. You can't do anything fancy like require some action to activate the lock; it's always active, it holds three balls and when it's full it dumps them all out for multiball. Simple to understand.

Lots of reviews of PCS complain about the weird and unbalanced scoring options: for some reason, the game emphasizes gigantic end-of-ball bonuses that dwarf the amount of points you can directly award for a target. Maybe the reason the bonuses get so big is that they are also the only really nontrivial table rules you can establish. There's a single type of logical element, an AND gate, and all it can do is AND up to three items together to do one of two things: add points to the bonus, or bump up the bonus multiplier. I did the common thing of making completion of those three rollover lanes at the top increase the bonus multiplier, and I have a bonus for getting the two standup targets (maybe too cheap a victory, I don't know), a bonus for getting multiball, and a relatively big bonus for completing all eight drop targets on the right. That last goal is very hard, and maybe it should be worth even more, but I didn't want to make it too much of a Golden Snitch. You can't award specials or extra balls or anything like that, as far as I can tell, or I'd definitely put one there.

But for purposes of a program like this, I'm not sure you really want much more complex table rules. I mean, I get it, I once had great fun wiring up a baroque state machine in Disney Infinity to play the old Merlin Magic Square game (with added killer robots) just to see if I could, but keeping it very simple probably feeds the player's sense of accomplishment and avoids getting people into really frustrating situations where they break their table by wiring up unexpected behavior they can't figure out.

What did I miss? The main thing, overwhelmingly, is that I wanted more control over the geometry of the flippers. You have big and small left flippers, big and small right flippers, and all of them have a resting state where they're canted at a weirdly steep angle, much steeper than a typical bottom flipper but not as steep as a sideways flipper. This makes it really hard to aim a shot with a trapped ball (though I appreciate that you can trap balls) or a ball coming down the inlane: the ball slides down the steep flipper in an instant, much faster than on a typical real pinball. It makes my target-shooter table design particularly challenging. I'd like to either have some control over the angle of the flipper, or at least have a better default.

Still, this pinball I made using this ancient toy editor from the dawn of personal computing isn't the greatest, but it is actually fun to play! I got it to the point where I wanted to play it more than tinker with it some more. You do get just a little taste of the pinball design process, where the designers make a plain prototype of the physical design called a "whitewood" and then spend some time "flipping the whitewood", that is, playing the naked table so that they can tell what works and what doesn't, and what shots are fun to make so that it's fun to reward the player for making them.

So, as the previous post mentioned, I was messing around a little with a few early personal-computer pinball games, specifically the Atari 400/800/etc. ports. How are they?

I think I'm a little spoiled for these tables, actually, because they're based on real pinball machines that I'm already overly familiar with, both in real life and in later simulations. The first big hit in computer pinball was Bill Budge's 1981 Raster Blaster, originally a program for the Apple II, ported before too long to everywhere else. Here's Mackinstyle's video of the Apple II version:



The Atari version looks pretty similar, except that there are purple As in place of the Apple logos on the bumpers, and the artifacting colors aren't quite as elaborate (I couldn't find a video that showed them properly). Anyway, a pinball fan will immediately recognize this table as an unlicensed imitation of Firepower, a pioneering 1980 table designed by Steve Ritchie.

(If you're primarily a video game fan, you probably know Ritchie better as a voice actor: "FINISH HIM!!!" His impressive voice is also often heard coming from his pinballs.)

It wasn't, by a long shot, the first multiball pinball, but it was the first multiball of the solid-state era, and the moment when it took, so to speak: after Firepower, almost every pinball had a multiball mode. It was also the first one where the user could rotate the lit rollover lanes by pressing the flipper button. It's a fast, simple, kind of brutal layout. The only real long-term table goal is multiball. You've got to hit those six little standup targets in the center repeatedly to light the locks for multiball, and every time you do, you risk draining the ball right down the middle. I find it addictive. I first encountered it in Farsight Software's Pinball Hall of Fame: The Williams Collection, played it more in their later sim The Pinball Arcade, and eventually got to play it for real at the late lamented Pinball Wizard Arcade.
Bill Budge, for his part, went on to make the epochal Pinball Construction Set for the nascent Electronic Arts (back when they were the cool, artsy, envelope-pushing video game company), one of the first games to offer the general public an editor that they could use to make their very own custom games! I didn't have this program back in the day and I've only scratched the surface of it--my impression from messing with it just a little is that its user interface is kind of touchy and unforgiving by modern standards, but a lot of time has passed since the early 1980s. One of the things it could do was allow you to save your pinball as a standalone executable, and apparently these user-created pinballs (most of them undoubtedly very bad) absolutely infested the BBS scene in those days.

Every port of Atari's 1979 arcade game Lunar Lander seems to map the controls differently, and it's an interesting exercise in translation.

Lunar Trails Art Project: http://moonlander.seb.ly/
IGN: https://my.ign.com/atari/lunar-lander
It's also on Atari Flashback Classics, volume 1.
(The Atari 2600 Space War Star Ship cartridge also had a mode called Lunar Lander, but it's a completely different game, not much of a game really; it actually predates the arcade game.)

It's interesting to think that Lunar Lander was developed concurrently with Asteroids (many of which ended up shipping in surplus Lunar Lander cabinets, because Asteroids was way more popular). It feels like a throwback to the pre-Space Invaders era of the early 70s, when a coin-op arcade game could be something other than frenetic action. It's just a quiet, tense battle between you and physics, and physics is definitely trying hard to kill you. The inspiration was old text-based mainframe "Lunar Lander" simulations that were, obviously, themselves inspired by current events. You're trying to land on the Moon, and you've got to fire your rocket in such a way that you don't either come down too hard or run out of fuel and end up as a smear on the landscape. Usually they would only simulate vertical motion.

Atari's was two-dimensional, for an added degree of challenge--it was a hard game, though still nothing compared to what real Apollo astronauts had to deal with. Your LM started out traveling horizontally over a craggy landscape with just a few good, flat places to land, some of them labeled with score multipliers; you'd be gradually accelerated downward by gravity, and there was also a tiny, fictitious friction force operating as a bit of mercy toward the player (at least in the easiest difficulty mode). Instead of some number of lives, you had a finite fuel reserve. When you either crashed or landed successfully, you'd get a text message and some number of points based on your performance; either way, if you had any unburned fuel, the scenario would just start over with the remaining fuel and a new landscape. When you hit the ground with no fuel, the game was over.

The controls were somewhat eccentric:
www.arcade-museum.com/images/118/118124212764.jpg
There were left and right rotate buttons like Asteroids, but instead of a button, the rocket-engine throttle control was this gigantic analog lever that offered variable engine thrust. I think it was bigger than the one in the real Apollo Lunar Module. And then there was an "abort" button that you could push if you were in deep trouble, that would immediately right the LM's orientation and push out a huge rocket blast to send you up into the sky (also spending a lot of fuel).

So how do you map these to a keyboard, or a game controller? The Lunar Trails project, which I think is a new reimplementation (albeit one that looks a lot like the original), just maps rotate left, rotate right and thrust to the left/right/up cursor keys, and I think it leaves off the abort button entirely (or at least I can't find it). It's actually possible to land nevertheless. I don't think it simulates the different difficulty levels.

The IGN implementation, which I think is an actual emulation, maps rotate just like the Lunar Trails version, but up and down now increase and decrease the rocket thrust, to give you fine control. The space bar is the abort.

What drove me to write this is that I initially thought the Atari Flashback Classics version was broken: I couldn't figure out how to control the lander. Turns out they mapped left, right and up on the D-pad or joystick, which would be left/right/thrust in an Asteroids-type game, to rotate left, rotate right and abort. (Update: I'd misremembered; it's the A button that is mapped to abort.) The throttle is the right trigger, which makes some sense because on the XBox One controller this is a one-dimensional analog control. You can vary the thrust by squeezing the trigger to varying degrees. It's probably not as fine control as that big old handle, though.

Eventually Atari elaborated the basic concept into Gravitar, which is the same basic idea of "spaceship in ruthless gravity field" only even harder and you're also fighting enemies. I suppose you could call it Lunar Lander crossed with Spacewar. Gravitar reverted the engine control to a simple thrust button.

https://my.ign.com/atari/gravitar

It is astoundingly difficult. Gravitar was, I think, not that popular.

I've been playing pinball sims again a lot lately, and I've noticed that recent software understandably leans toward emulating the complex and flashy tables of the 1990s and later (and fantasy elaborations of same). But sometimes I want something simple and quick, like an old electromechanical table. There are some modern emulations of those, but another way to dial it back is to play some of the earliest home pinball sims, the pinball games made for the Atari VCS/2600. I know of three, which I can illustrate with other people's YouTube videos...

Video Pinball, Atari, 1980

Atari made three completely distinct videogames called "Video Pinball", unrelated except for being pinball sims. The first was a slick and pretty fun coin-op arcade game that superimposed a 1-bit B/W video display on a physical, disco-themed playfield model that was reflected in a half-silvered mirror. The second was a visually primitive stand-alone home console in the late 1970s, and the third was an Atari 2600 cartridge. This is the only one of the games here that I actually played back in the heyday of the 2600, and I played it a lot. Here's espaciodejuegos playing it:



This game looks OK, if blocky, and it takes care to make use of the whole screen, but partly for that reason it's easily the least like real pinball of these cartridges. The playfield is kind of weird. The bottom is a vast horizontal expanse that the tiny ball can just bounce around to its heart's content; you spend a lot of time waiting, and with some skill a ball can last for a very long time. (The harder difficulty switch setting opens a couple of small extra drain holes in the bottom.) On the other hand, it's the only one of these games that actually allows nudging, with a fairly generous tilt setting, by holding the fire button down and moving the joystick.

There are 1- and 2-player games. Aside from the per-player difficulty switches that open the drain holes, the only other game variation controls whether accumulated bumper values are held between balls.


Midnight Magic, Atari, 1984/86

This was one of those games released after the 1983 videogame crash that few people saw in its cartridge heyday, though it's gained some notoriety since then as a remarkably good pinball sim considering the platform. It seems to have been intended as a port of the popular personal-computer game David's Midnight Magic, but that game had a layout shamelessly ripped off from the classic real pinball Black Knight, and this 2600 game definitely doesn't. As played by highretrogamelord:



It's symmetrical, probably because the 2600 hardware made symmetrical playfields much easier to produce. (They cheat graphically a little by not actually showing a full plunger lane.) It has some aesthetic similarities to David's Midnight Magic, though, and you can kind of see that they were trying to get some aspects of the Black Knight layout in there.

This game is really good, and much, much more pinball-like than Video Pinball, despite having no nudge. The playfield takes up much less of the screen, but it's the right shape and the bottom is like a real pinball machine, with inlanes and outlanes. With the default difficulty setting, the outlanes have always-active kickbacks in them and there's a pretty wide post in the center drain, which makes the game considerably easier; setting the difficulty to hard takes away the kickbacks and the center post, making it much more arcade-realistic. Like an old multi-player electromechanical, nothing about the table state is held between balls; that makes it a fairly challenging game in any event, and there's no shame in playing on the easier setting.

If I have a complaint about it, it's that the crucial drop targets at the top of the layout (which advance a multiplier if you get them all) are above the pop bumpers, so you can't really shoot for them from the bottom flippers, only from the side ones, which give you almost no ability to aim. It'd also be cool if, instead of the kickbacks being always on or off, you had to activate them by making some shot, like on many real pinballs. But trying to nail that spinner in the middle, or make the rollovers up in the corners, is a lot of fun.

Bumper Bash, Spectravideo, 1983

This one is really obscure, apparently very hard to find in actual cartridge form. I mostly remember Spectravideo for trying to introduce an MSX-based home computer in North America, but apparently they also made Atari cartridges. Theshadowsnose apparently has one, and has some entertaining commentary on it:

Again, there's no nudge--the copy on the box tried to paint that as a feature by joking about it being "cheating". But the game has enough to interest pinball fans anyway.