samiam

Since I ended up writing a quick and dirty JSON parser yesterday (There wasn’t, until yesterday, any public domain Lua JSON parser, until I wrote one), I gave the JSON parser the name “Black Castle” in honor of the almost impossible “Black Castle” puzzle this game has. Why do I bring up JSON? Because, this game, as mentioned above, was written by the same guy who later on invented JSON. Warning: Spoilers ahead The game has a very difficult puzzle where a secret passage was hidden in the black castle; the passage was so difficult to find, I had to run the program through a 6502 monitor to find it back in the 1980s. My dad (rest in peace) had written a 2 kilobyte 6502 monitor which could hide itself where Atari put their floating point library (making it immune from copy protection which looked for ROMs in the $COOO-$CFFF range); I knew how to program 6502 and use his monitor. While playing Galahad and the Holy Grail, I used the monitor to flip the page used to view which room we were looking at without changing the colors for the room; this let me see all of the invisible mazes and secret passages the game has. When looking at the black castle (which yes, I fairly extensively explored without assistance, but failed to find the passage), I finally saw the secret passage to the hidden black castle maze. While I never finished the game back then — that rabbit is just too much of a pain to deal with without the ability to save the machine state the way modern emulators can — I knew the solution was to use the grenade to blow up the ship room multiple times until it was possible to go from the fairly easy to reach left entrance of that room to the almost impossible to reach top entrance of the ship room. I am not the only one who could find this passage in the black castle without assistance; a 1984 reviewer of "Galahad and the Holy Grail" had the same issue: In "The book of Atari Software 1984", the reviewer could not find this secret passage until Crockford sent them a map and hint sheet. Indeed, in the 1985 (?) Antic revision of this game, there is an arrow in the black castle pointing to the secret passage. In addition, Crockford (I believe he’s the one who revised the game for the Antic bugfix release) had the gray castle (whose entrance kills you in the original 1982 APX version of the game) teleport you to the same place as the black castle’s secret passage does. After writing my “Black Castle” JSON parser, I felt inspired to finally finish this game. As suspected, once I got past the rabbit (with the sword and loading the emulator’s state just before entering the rabbit room until it approached me in a direction I could kill it with), it was relatively straightforward going through the mazes until one is in the dragon room. The dragon can be killed with multiple loads using a modern emulator and saving the state before entering that room; once the dragon is killed, I had to go through some more mazes, including frustrating invisible mazes, until I was in the grail room. Once I picked up the grail, since I was playing the original APX version of the game, I can not blow up the “right arrow room” (to the left of the Pac Man room) and use its top exit to get to the gray castle which, in the Antic version of the game, leads to the black castle secret and eventually back to the starting location where one needs to place the grail to win the game. Instead, I had to blow up the ship room twice, which is a cycle of killing the two spiders, the rabbit (ugh), then getting the grenade from the Pac Man room to the ship room, dying, going all of the way back to the Pac Man room, killing the spiders and rabbit again, then blowing up the ship room a second time. As it turns out, while the usual passage to the Pac Man room consists of getting both keys and then opening the gates in the white castle (the first gate is opened with the yellow key, which is located in the teleport maze just before the entrance to the left side of the ship room; the second gate is opened with the green key, which is located just before the maze of 2600-style rooms in the duck forest), it’s actually possible to get to the Pac Man room without any keys: In the room north of the Excalibur room (in the maze one gets to from the black castle secret passage), there is a hidden teleport to a room with some light pink parts in the wall. Touching any of these light pink parts is a hidden shortcut from the post-Excalibur maze to the Pac Man room. Indeed, it’s possible to win the game without hitting “Select” to disable the locked doors and without getting any keys: After blowing up the ship room twice, one does not need the pink key (also in the area around the Excalibur room) to get to the grail; one can simply enter the ship room from on its left side (in the room just after the “blinking spider” room) then find a passage to a remaining piece of the ship which teleports the player to the difficult maze just before the grail room. Once one has the grail, one goes through the destroyed ship room again and exits on the left, then goes north and up the mountain to get back to the starting area and finally the starting room to win the game. And that’s how the original APX version of Galahad and the Holy Grail is won. To Crockford’s credit, writing a JSON parser was actually easier than winning this old game of his. Just as the Antic version of Galahad fixes annoyances like the hidden black castle passage and the need to blow up the ship room multiple times before winning, my JSON parser fixes JSON annoyances such as adding support for comments (via standard Python style # comments), allowing bare words for object keys (we need to parse bare words with JSON anyway since “null”, “true”, and “false” are bare words in the spec), and, unlike other excessively pedantic strict JSON parsers, allowing a comma to be placed after the final item in an array or JSON object. I once had a job where I had to use a particular limited domain specific language to translate our internal data in to JSON format; the language was so limited, I tried to add a final comma to a list of items in an object, which Python’s JSON parser doesn’t accept. So, I was forced to put "__ignoreme": "dummy" } at the end of every JSON list; people complained in meetings multiple times about that dummy JSON item, and I had to keep explaining why I had to add it, and how Python’s JSON parser doesn’t allow a final comma at the end of a list. Which is why, now that I have finally written my own JSON parser, I made very sure a final comma is allowed.

I’m going to side with Dan on this issue. Yes, Open source programs tend to have more silly usability issues that closed-source programs have. Let me be more direct: My open-source program has silly usability issues that it would not have if it were a closed-sourced program and I was able to have it bring the groceries home and the rent paid by working on it all day every day. There is no such thing as a free lunch! When open source developers aren’t (with certain fairly rare exceptions like Firefox and parts of the Linux kernel) getting paid to develop a program, they will tend to have less time to work on the program to fix issues like an install path limitation that’s too short. - Sam

I think what you’re seeing is a variant on the “the idiot won’t RTFM” problem. People will download and begin to use software while reading the absolute minimum amount of documentation needed to do so. That’s how I ended up using VMWare player for commercial use for so long; I simply was ignorant about the player’s commercial use restrictions. For someone with a license which only allows non-commercial use, there has to be prominently visible message reminding them of this fact everywhere. I mean everywhere. NO COMMERCIAL USE WITH MYSQL needs to be the first and last thing in every document. Every time the program is run, its output needs this message as the first and last line of output (if command line) or prominently placed in the window’s title and elsewhere in the GUI (if GUI/Window based). It has to be impossible to miss the message about the commercial use restrictions, or the freeloaders will not see it. For legal purposes, to have a contract like this hold up in court, a user has to click on “I agree to the terms of use”. It has to be an active action on the user’s part. Even if not going the legal route (which is a huge can of worms; nothing pisses off the hacker/open source/freeloader communities faster than a lawsuit [1]), having a checkbox saying “I agree to not use this commercially with MySQL without buying a license first” will help too. - Sam [1] I do not have the hatred others have towards Prenda Law. If people were following the law and not downloading and seeding copyright infringing torrents, they would not have been hit with a Prenda Law suit. It’s that simple.

Personally, I prefer the word “freeloader”. The issue with all of the “tard” words is that, by using “tard” in a negative context, there is, yes, a subtext that says “a developmentally disabled person has less value as a human being” In addition, there is the issue that “freetard” is a recent neologism and has two distinctly different meanings (“Annoying and/or militant Linux advocate” and “freeloader and/or advocate of freeloading”)

I think it’s important to always and constantly remind users of the “no commercial use before paying me” nature of such software. I used the “free beer”/“gratis” VMWare player for commercial purposes for years because I never bothered to read the license’s fine print, which, indeed, explains that VMWare Player can not be used commercially without being upgraded to the for-pay VMWare Workstation product. VMWare finally got sick and tired of people doing that, and put the message “non-commercial use only” everywhere a user may glance when using the software (on the Windows’ title bar, with the note “for commercial use, purchase a license by clicking here” in help->about, etc.) I finally got the message ... and now use the inferior [1] free-for-commercial-use open-source VBox program instead, since I really am not in a financial position to spend $500 or whatever on a software license if there’s some reasonably equivalent open source software out there. - Sam [1] Inferior because, unlike VMWare, VBox doesn’t allow a virtual machine to have multiple IPs in a virtual NAT environment.

We briefly got distracted there, but I sometimes get really mad about silly things myself; I understand and forgive anyone who got a bit heated up. I think we can continue to have a mature worthwhile discussion. Likewise, users of open source sometimes have expectations for help and support that stem from their experiences using commercial software. With, if you will, “pure” open source, the answer to any question asking how to use the software is a combination of RTFM (read the f***ing manual) and UTSL (Use the source, Luke), and the answer to any feature request is “UTSL and implement it yourself”. And, yes, I do give my users the RTFM bomb if they ask me about how to do something in my software that is already well documented, especially if the mailing list has a lot of traffic and/or the same user has asked me a bunch of questions already. I’m also not averse to telling users “patch it yourself” (like the guy who wanted better OpenBSD support. Well, I don’t have OpenBSD, they have the source code, so time for them to get cracking) I already mentioned that open source still allows money to be made. However, compare the size of RedHat with the size of Sun during the 1990s. Sun used to be one of the hugest companies in Silicon valley, with multiple campuses in different locations. These days, those campuses have been either closed, rented out to other companies, or are big empty offices with only a few workers in a big building. RedHat, who sells essentially the same product Sun does (UNIX servers) is tiny compared to Sun at its peak. Some solid numbers: In 2001, Sun had a market cap of over $37 billion. RedHat today has a market cap of just under $9 billion dollars. The consumer price index has gone up by 33% since 2001, so Sun was over five times the size of RedHat today at their dot-com peak. Oh, yes, I remember the RIAA killing digital audio tape recorders for home stereos in the 1990s. Indeed, I ended up finally opposing SOPA because of the RIAA’s historical shenanigans: http://samiam.org/blog/20120119.html I really wish I could come up with a business model for me to get rich, or even putting food on the table making open source software. I’m not sure I have the needed people skills to start up my own business. I like that.

I really have no problem with the GPL (but prefer either public domain or simple BSD for my own open-source projects); I’ve met and indeed like Richard M. Stallman. I think the reason the GPL is so heavily derided in some circles is because a lot of people see the GPL as condoning a “handout”. Or likewise, they see the GPL as saying “you don’t deserve to get paid for your hard work making software”. I’ve also heard horror stories of people using GPL software in a commercial product without fully understanding the license and its consequences, and then having to reimplement all of the GPL code in a hurry so their entire product is not bound by the GPL license.

Lots of ideas here and lots of emotions running high. Here are my thoughts to the ideas being presented: The work for some classes of software. The flourish when a company has another way of making money with free software. Apple can make a free and open-source browser because they sell iDevices, not browser software. Google makes free browsers because they sell ad space on web pages, not browser software, and having more control over the browser allows Google to more effectively make money with their business model. They work for software where people are willing to pay real cash for real technical support and timely security updates. This is why RedHat is so successful. The reason why I use the “unsavory” word “freetard” is because it allows me to quickly describe a certain class of annoying user. It’s really no different than the words “troll” or “spammer”. The annoying behavior will exist whether we assign a word to it. Someone who says “don’t use the word freetard” is in effect saying “There is nothing wrong with someone demanding all kinds of things from software developers without ever paying them” (“freetard” can also mean “annoying Linux user/fanatic”, but I try and not use it in that manner) It’s like the pirates who say “don’t use the word pirate!” What they are really saying is “Don’t use a word which says it’s morally wrong to copy intellectual property (IP) without compensating the people who worked hard making the IP” The reason why Paula Deen has been so savagely attacked for saying the N-word is because, by saying that word, Paula Deen was using a word that says “It’s OK to have a world where someone has fewer opportunities because they were born with the wrong skin color”. Yep. Linux on the desktop might never had happened if there wasn’t this non-open-source QT toolkit around that allowed people to make professional-quality application software in Linux. To summarize Linux on the desktop in the 1990s, the first real Windows/Macintosh clone desktop environment for Linux was something called “KDE”. KDE was developed using this non-open-source toolkit because there plain simply wasn’t an open source toolkit suitable for making a desktop environment. This made a lot of Linux purists upset, since all this code didn’t have a “pure” open source license. What ended up happening was that KDE’s existence motivated other developers to make a toolkit (actually, extend an already existing toolkit made for a single application to work as the basis for an entire desktop system) and then develop an entire different desktop environment, called Gnome, around that toolkit. Eventually, QT was finally made open source. This is one problem with a lot of open source software: It is often so hard to set up, it takes more time to set up an existing open source program that makes more sense just to roll one’s own code. For example, the standard blogging software suite (MySQL + Wordpress) didn’t work for me because I wanted to have something easier to set up, more reliable, and with fewer moving parts. So, I created my own scripts for managing my blog (which create static HTML pages). Fewer moving parts, more reliable, and I can easily replicate my blog across multiple servers (something non trivial with MySQL) OK, I won’t use the word “thief”. What word should I use? This word needs to connotate that piracy is morally questionable. Or perhaps you’re offended that we’re calling a spade a spade. I don’t feel piracy is always bad in all circumstances (it’s OK for me to remove DRM to, for example, play an old game on a netbook without a CD drive; it’s also OK with my set of ethical values to use piracy to preserve art and code which may otherwise become lost in the sands of time). If we did not have strong laws in place to protect intellectual property (IP), movies, music, and software would not have the same quality it has today. Instead of multi-million dollar big budget movies, we would have to pass the time watching amateur quality YouTube videos. Instead of well-produced well-recorded music written by hard working talented musicians, we would have to pass the time listening to amateur low-quality garage bands. It is perfectly legal to protest IP laws by deciding to not consume anything created by a movie studio associated with the MPAA, to not listen to any music without a free-for-commercial-use Creative Commons or equivalent license, and to only use open source software on one’s computer. However, when someone first bemoans IP laws, then downloads content in violation of said IP laws, they become a hypocrite. They become, well, a freetard, because they are now getting something for nothing. Maybe there’s a better economic model for making content than protecting it with strong IP. However, the pirates that bemoan IP laws never come up with an alternative working economic model. Can Kickstarter give us a 200 million dollar blockbuster movie? I don’t think it can.

Thanks for your support guys. We’re seeing is what I call the post-Napster Internet. I blogged about that here: http://samiam.org/blog/20120907.html#slashdot and here’s the relevant part: Napster changed the face of the Internet and how people looked at copyright. Before, the Internet community understood that making meaningful intellectual property was hard work and should be respected. Once Napster allowed anyone with an Internet connection to have all the music they wanted for free, with only a small nagging detail that it was violating copyright, thousands of self-centered young college students and others did just that. Instead of taking the high road, Slashdot decided to (by and large) side with the masses of people coming up with convoluted rationalizations and justifications for downloading copyrighted content without permission. Now, I’ll be honest: I really appreciate all of the people who have made recent homebrews of 2600 games and have freely given them today. And, yes, at those Atari user groups many moons ago, everyone was pirating games and there was even a group of 6502 programmers who worked together to crack copy protected games. One time, being a little kid who couldn’t keep his mouth shut, I made the mistake of letting one of the game developers know who it was who cracked his game (to be fair, the cracker was talking to the developer about how there were illegal/undocumented op codes in the game’s boot loader), but the developer was really cool about it. [1] So, yes, piracy is nothing new. But none of the crackers pretended what they were doing was morally OK. They knew they were being naughty—the cracker later on told me in private that I shouldn’t have told the developer that his game was cracked. And, to be fair, there was no attempt from the game developers to sue or prosecute the pirates back then, which probably has recently spurred people on the fence about piracy to fully condone copyright infringement. It’s unfortunate that so many on the Internet think everything should be free and I think one of the reasons it’s a lot harder for talented programmers to make a living is because end users no longer consider it normal to pay for software. - Sam [1] I actually think I can still contact the developer and the cracker in question, so I may be able to name names once I get permission from both. Or just name the developer if the cracker wants to keep his naughty behavior from 30 years ago under wraps.

That’s what I did after getting bored with all those 2600 games I got for $5 a pop. By 1985 a new Atari 800XL was to be had for $100, which cost far less than a 2600 console did just a couple of years before, and was fully programmable to boot. The XE series and inexpensive ram expansions, as well as Turbo Basic kept them going for quite a few years—what little gaming I did was either on that 800XL (I was playing and programming it as recently as 1993) or on the family Amiga. PC gaming just wasn’t there until the Internet and FPS (first-person shooter) revolution a few years later. I knew one guy who owned one. The Intellivision’s problem was that, while better than the 2600, with hacks like the 6-digit score hack, it was possible to make games almost as good on the Atari. Atari’s better ability to secure home gaming licenses for coin-op games was the nail in Mattel Intellivision’s coffin. By 1982, Mattel was selling “M-Network” 2600 games. Also, I get the sense that the 6502 is a lot cleaner and simpler to code in than the CP1610 was. It took a David Crane level of programmer to want to program one (he made the Intellivision port of Pitfall, which arguably was the best game for the Intellivision—but I’m happy to be refuted by any lurking Blue Sky Rangers out there).

Wise move. I think open source caught on big time during the dot-com boom because, with the way Microsoft was taking all of the oxygen away from Netscape [1] (or anyone else who dared to stand in their way), the only strategy for software developers that could at all withstand Microsoft was to give it all away. This is why Google spends to this day millions of dollars both developing both their own Chrome and financing Firefox; it is to ensure that the Internet never becomes a close standard defined by a single company. I don’t regret my project; it kept my skills current while I lived and worked in Mexico for a few years, and gave me two job offers during one of the worst recessions when I came back. This happened late 1984, early 1985, right around the time of the Tramiel takeover. They may even have been stolen in a “they missed payroll” kind of way, but the story we were hearing from the guys selling us the stuff was that their boss said “We can’t make payroll. Get a truck and we’ll load it with a palette of 800XLs.” The guys were too visible and high-profile for it to have been downright theft. My sense with the post-crash 2600 scene is that, come 1984, it was impossible to sell a new 2600 game. Once all of the overstocked 2600 inventory was liquidated in late 1984 or early 1985, the stores started to stock, on a very limited basis, low-cost 2600 consoles and games again. I saw 2600s for sale as late as 1987 or 1988, with the very occasional new 2600 game, but, by then, everyone was playing Nintendo instead. Atari only held on to the 2600 for as long as they did because they didn’t have another hit to replace it with. The tiny trickle of money they were getting in the late 1980s from $50 cheap 2600 consoles and the very few new games released for it was better than nothing, and probably the reason why they held on to it until 1992. [1] I remember when Netscape had a hugs campus in Mountain View. Today, the buildings have been repainted and the big “N” logo has been replaced with Symantec’s logo. I remember visiting Be’s headquarters; they were bought out by Palm, who could not get any traction post-iPhone and recently died after being bought out by HP and renamed webOS (anyone else remember the 2600-style clearance of webOS devices from just a couple of years ago?). SGI ended up competing with Windows NT, and their buildings are now all offices for Google. These are just three companies whose tombstones are best captioned “Here lie companies that dared to compete with Microsoft in the 90s”

Naturally, it wasn’t. That said, as someone who has his own open-source project—a “long tail” DNS server [1]—I have become really jaded by the hordes of “freetards” on the Internet who expect something for nothing. I finished that program and put the project on the back burner (bug fixes only) nearly three years ago, but I still get in the occasional flame war with someone because I no longer implement features “for fun and for free”. It wasn’t a total loss—the project has its own Wikipedia page and gave me the job I have today—but dealing with freetards has changed my entire attitude towards Linux and the entire Open Source movement. I think David Crane is on to something there. The sense I got (again, as a young kid) is that, come 1983, Atari 2600 games started to get the perception as being “lame”, in a large part because the most marketed 2600 games out there (Pac-Man and ET) were quite lousy. Seeing 2600 games being sold at the local KB Toy Store for $5 a pop strengthened the perception that 2600 games were not worth anything any more. By 1984, it made no sense whatsoever to make a 2600 game. Compare Montezuma’s Revenge on the 2600 and on the then almost as inexpensive Commodore 64 or Atari 800XL [2]. I didn’t even see the financial point in making a port for the 2600. It was a cheap system for cheap games which did not hold a candle to all of the low-cost 8-bit computers and consoles out there. By the time Super Mario Brothers revived the video game industry, I had lost almost all interest in video games and programming, and fancied myself an electronic musician. [1] MaraDNS. I call it “long tail” because it has never been one of the “big 2” DNS servers. When I started MaraDNS, the “big 2” were BIND and DJBdns (and the flame wars were really ugly); these days the “big 2” are BIND and Unbound/NSD, and the discussions about which DNS server one should use are much more dignified and quiet. [2] Here’s another story: I remember Atari employees coming out and selling us 800XLs for pennies on the dollar. The story I heard—and this is probably as apocryphal as the story of Atari filling a truck with unsold ET cartridges and burying them in the desert—is that Atari was giving employees pallets of Atari 800XLs and other merchandise to sell themselves instead of paychecks.

I’m a little reluctant to reveal the company’s name because I’m a bit worried about protecting the privacy of the people involved, since most of them are still alive today. While there is no scandalous gossip for me to share, I have over the years lost touch with those guys, and I am not 100% sure if the people involved in that startup want the story of their rapid rise and fall as remembered by a then young child made public on the Internet. I will Google around and see how much of the story of the people involved is already a matter of public record.

The thing I remember seeing was a one-page photocopy-of-a-photocopy description of the Stella registers at the top of page 0; having access to even that was considered very hot. The manual was very terse, and I don’t think it had the official names like “GRP0” and what not, but it let people make very compelling 2600 games. Even though I could, in a fashion, program in 6502 at the time, I was too young to really be a part of the original video game tech bubble, but I remember watching the grown-ups (some friends of my family) sitting around for hours after a local computer users group talking about their 2600 video game startup, discussing the market, how millions could be made, about getting financing and office space to make the start up happen. It was very exciting to listen to. I watched as that company making 2600 games was getting formed with the dreams of hitting it big and retiring young, only to have, just a year later, 2600 game companies dying like flies left and right. The startup I saw up close and personal was very lucky—they were bought out for pennies on the dollar; the management structure was retained and most workers kept their jobs for years after the 2600 no longer was a cash cow. [1] I know there has been a lot of speculation about what went wrong and why the bubble burst so quickly (the 1990s “dot-com” bubble, as a point of comparison, lasted half a decade before popping). My personal theory is that Pac-Man and ET did the most damage; both games got the most publicity but were plain simply not very good games. Because of the extensive marketing on Atari’s part, instead of something really strong like Pitfall, it was 2600’s Pac-Man which made the evening news and was the most visible in the local toy store. While Atari was able to later on make a compelling Pac-Man clone for the 2600 with Ms. Pac Man [2], the damage had already been done. Along with all of the lousy third party games (lousy not because the people making them were bad programmers; lousy because a lot of the companies just didn’t know how the 2600 fully worked and because, like Pac-Man and ET, the games were often times rush jobs) people perceived the 2600 as only being able to play “lame” games, even though it had really fantastic games like Demon Attack, River Raid, Pitfall/Pitfall II [3], and the Supercharger games such as Phaser Patrol. When the Nintendo Entertainment System first came out, in the later days of the crash, I looked at it and said to myself “They’re crazy to try and release a new video game console”. This was when dozens of different Atari 2600 games were available in the bargain bin for $5 a game, and arcade coin-ops were shutting down left and right because everyone got bored with playing those video games; many though that video games would go the way of Bell Bottoms, Disco, and the Hula Loop. [1] The Regan-era greed was only starting to take hold, and ideas that became normal during the dot-com era, such as buying out a company and then firing most or all of its employees, had not become mainstream. Good workers back then were valued a lot more in Silicon Valley than they are today. [2] Personally, I think what killed Pac-Man was the flickering. Ms. Pac-Man still has some pretty bad flickering problems. This probably could have been minimized by having only three ghosts, and using missile instead of player graphics for the floating fruit (giving it an abstract form). [3] I looked at any new 2600 game that came out after 1983 and said to myself “what’s the point?”. For example: Montezuma’s Revenge and, later on, California Games. Both of these games are considered examples of the best games ever made for the 2600. Both games do not hold a candle to the home 8-bit computer or NES versions of the same game. Anyone with money to buy a new video game had money to get a better system than the 2600 to play the game on.

I was just looking at some of the 2600 Hacks for Wizard of Wor, and I must say the 2600 “Arcade” version is impressive. After looking at the rom (which is, alas, not readably playable in Stella, because of the “joystick in the wrong port” issue), I started reading what other information on Wizard of Wor is out there, and while there’s some interesting information on its history and what not, there’s not any information on how the game was actually played. Let me take you down memory lane... During the peak of the coinop and Atari 2600 era, I was really too young to go in to arcades. This didn’t, however, stop me from going, much to my mother’s annoyance. Since my parents weren’t exactly condoning me being at the arcade, I never had any quarters in my pocket. So, I would go and just watch everyone else play games there. Every time I saw people playing Wizard of Wor, they did the same trick to win the levels without really trying: Play the game as a 2-player game. Take both characters and put them in the same corner location, ideally in a corner where both directions were facing a nearby wall. Have the players facing two different directions at right angles from each other to cover both approaches to the corner location. Do nothing but hit “fire” over and over and over until the level is cleared. Later, rinse, and repeat for the next level. Like patterns in Pac-Man or shooting UFOs when there was only one or two asteroids were left in Asteroids, this was a strategy not anticipated by the original game creators, and allowed people to win far more easily and play for a much longer time with relatively little effort. Indeed, one of those Arcade game strategy guides out there described this trick for winning Wizard of Wor. If Wizard of Wor had been successful enough to merit a sequel (while a reasonably well-known game, and popular enough to port to the 2600 and home computers of the era, it did not dominate the arcade scene the way Asteroids and later Pac-Man did), I am sure the sequel would have changed the game mechanics to punish players who did not move and just shot anything that approached them until the level was finished.