Shut the Box

There are lots of traditional games whose history is very little known. This is one of them. The best that I could find out is that it was popular among sailors of Normandy in the late 19th century (according to David Parlett, The Oxford History of Board Games), and that it was brought to England from the Channel Islands in 1958 (according to Timothy Finn, Pub Games of England). It is usually played with two dice and a special box that has tiles with the numbers from one to nine. Wikipedia describes the game thus:

During each round, a player repeatedly throws the dice to cover the tiles of the box. The round ends when no tile can be covered on a throw and the player counts each successfully covered number as a point for his score. If, for example you put down the numbers 3, 5, and 9, you get three points. The goal is to cover all numbers, that is, shut the box, which finishes that game. Play continues until each player has completed three rounds, at which time scores are compared and a winner is declared.

The old Star Trek mainframe game

I have now two Windows implementations of this old game on Download Central (WinTrek and Stellar Explorer) and have started researching its history a bit. It turns out to be rather complex, and the various accounts often contradict each other at least in the details. Here is one from David Ahl’s 1978 book BASIC Computer Games, where it was included as Super Star Trek:

Many versions of Star Trek have been kicking around various college campuses since the late sixties. I recall playing on at Carnegie-Mellon Univ. In 1967 or 68, and a very different one at Berkeley. However, these were a far cry from the one written by Mike Mayfield of Centerline Engineering and/or Custom Data. This was written for an HP2000C and completed in October 1972. It became the standard Star Trek in February 1973 when it was put in the HP contributed program library onto a number of HP Data Center machines.

In the summer of 1973, I converted the HP version to BASIC-PLUS for DEC’s RSTS-11 compiler and added a few bits and pieces while I was at it. Mary Cole at DEC contributed enormously to this task too. Later that year I published it under the name SPACWR (Space War—in retrospect, an incorrect name) in my book 101 BASIC Computer Games. It is difficult today to find a computer installation that does not have one of these versions of Start Trek available.

The next one is from the readme of a 2002 (according to the time stamps on the files) Windows version of Super Star Trek with the credits by David Matuszek and Paul Reynolds, with Modifications by Don Smith, Resurrected by Tom Almy. Most likely the author of these lines is Don Smith, since the account is found under the header modifications:

Back in (about) 1977 I got a copy of this Super Star Trek game for the CDC 6600 mainframe computer. Someone had converted it to PDP-11 Fortran but couldn’t get it to run because of its size. I modified the program to use overlays and managed to shoehorn it in on the 56k byte machine.

I liked the game so much I put some time into fixing bugs, mainly what could be called continuity errors and loopholes in the game’s logic. We even played a couple tournaments.

In 1979, I lost access to that PDP-11. I did save the source code listing. In 1995, missing that old friend, I started converting the program into portable ANSI C. It’s been slow, tedious work that took over a year to accomplish.

In early 1997, I got the bright idea to look for references to Super Star Trek on the World Wide Web. There weren’t many hits, but there was one that came up with 1979 Fortran sources! This version had a few additional features that mine didn’t have, however mine had some feature it didn’t have. So I merged its features that I liked. I also took a peek at the DECUS version (a port, less sources, to the PDP-10), and some other variations.

Windows 3.0 games, 1990

I just updated the list of Windows 3.0 games, adding, for now, all those from the year 1990 now on Download Central. I added only the native 3.0 programs, not the hybrids like Klotz 2.11a, Checkers or Backgammon, but I did add those where an older 2.0 version exists. In total, there are now 24. This is probably a significant portion of the Windows 3.0 games from that year that exist at all. Now, of these 24,

1. at least seven were written by Microsoft employees,
2. seven are definitely arcade games,
3. six are Tetris clones of some sort,
4. not a single one is a card game of any sort,
5. four are of unknown origin, and
6. only four are from some distinct location outside the USA.

Point one just shows that Windows development, or at least Windows game development, was still mainly practiced mainly within the walls of the Microsoft headquarters. I write at least because they do not usually identify themselves as such, so these are just the ones I know by name or that put a Bogus Software or Duff Software copyright notice into their games. The completely anonymous Hex Tetris seems to be a sample program of a Microsoft application and may as such fall into this category as well.

Point two to six are interesting because they run contrary to what is true or at least said about Windows games in general. It is also interesting that one out of four Tetris clones for 16-bit Windows are from 1990 or older! As for card games, I checked those that I uploaded so far, two are 1991, nearly every other 1992, and the rest later. I guess it has to do with Visual Basic.

Point five is just a curiosity. As for point six, these four games are from the UK, Ireland, Australia, and Germany. It just shows that Windows wasn’t used or at least programmed for outside the USA a lot at that point.

Anyway my curiosity is now piqued. I’ll do a systematic search for Windows 3.0 games from 1990 and post updated statistics when I’m done.

Targ, Attack Force, Crossfire

Some time ago I uploaded a Windows arcade game, Alien Force. It was among the first Windows 3.0 games and very popular in its time, you found it on most shareware CDs. Today I uploaded a similar, but far less known game, ZoneGame from Outer-Space. In spite of the obvious similarities, the two games claim to be remakes of two different 8-bit games: Attack Force on the TRS-80 resp. Crossfire on the Apple ][. That got me interested, and I researched a bit.

Nebula Fighter

About two years ago I came across the shareware version of Nebula Fighter, a 1997 shoot ‘em up by Fabrizio Farenga and Alfredo Siragusa. The closing credits announced the full version for November of the same year, to be distributed exclusively by a Texas publisher, One Reality. Later I uploaded the shareware to Download Central, and it became quite popular (currently DOS game #9, new download #5). Here’s a screenshot from the Nebula Fighter shareware:

I have often wondered if that full version ever manifested. I still don’t know, but probably not. The game was however published two years later on CD-ROM, but as a Windows/DirectX game. Holodream Software, the label under which Nebula Fighter was developed, still maintains a website, and still offers a Windows demo for download, which I have now put up as well, untested and untried. Here’s a screenshot, from the Holodream website:

Nebula Fighter is interesting because like Katharsis or T-zer0 it represents an attempt to salvage a dying genre into a new era. None of these attempts was particularly successful, and the games remained mostly obscure (currently Nebula Fighter does not even have a MobyGames entry). For a while, scrolling shooters just like the once so dominant platform games were simply considered quaint. As the interest in classic gaming rises, these latecomers may be discovered again.

Castle of the Winds

A couple of days ago I uploaded Castle of the Winds, which in a way renewed my interest in roguelike games and led to all that discussion about hack121 and so on. Here are some additional tidbits for this game.

First, you’ll often read that Castle of the Winds was originally released in 1989. That’s complete nonsense. There are two releses, 1.0 from 1992 and 1.1 from 1993. That’s all. 1989 was the year Rick Saada started working on it, and he reflected that in the copyright notice (1989–92). That’s not an uncommon thing to do. Castle of the Winds was quite certainly a Windows 3.0 game from the beginning. Windows 3.0 was released in May 1990, but the SDKs tend to be available several months earlier, especially to Microsoft employees.

A question that is less easy to answer is whether it is correct to see Castle of the Winds as a roguelike game. Rick Saada never used that term, in the help file he refers to his game as a “graphical adventure game, loosely based on fantasy role playing games, and drawing much inspiration from Norse mythology.” Some dungeons may be random (this is usually seen as a main criterion), the first one definitely isn’t. The game also puts a stronger emphasis on story than roguelike games usually do, which due to their randomness take more of a sandbox approach.

However you want to answer this question, Castle of the Winds is definitely a remarkable game. With 13500 registrations, it may have been the greatest shareware success on 16-bit Windows. It still has its fans and fan sites, and Rick Saada still gets mail about it. It was one of the first RPGs for Windows, it remained one of few, and better than most. While back than it was often said that there were so few arcade games for Windows, this was never really true. The type of game that was really rare on Windows was exactly this kind, RPGs and adventures, games that demand a sort of commitment from the player and offer a sort of immersion (the big exception here, as in so many aspects, is Balance of Power). They got more numerous as SVGA cards with Windows drivers became more widespread, but Castle of the Winds remains one of the few 16 color games with these qualities.

More about hack121

So I’ve uploaded this mysterious game now to Download Central. You can download it, and look at some screenshots.

This game is really very mysterious. It has no title screen, no copyright notice, no release year, no name of the author hidden somewhere in the code. You can get the message Hack version 1.21 (Slak was here!) by pressing STRG+V at any time in the game. That is all. The only documentation is a file named moves.txt which gives only the barest gameplay information and surprises a bit with incorrect English (weild, veiw, hungary) not mirrored by the in-game messages. Maybe it was written by someone else. The files in the archive all have the time stamp 2000-04-26. Alone the fact that the time stamps are exactly identical is an indicator that they have no relevance for the creation date of the game.

Hack 1.21 has a few interesting gameplay details. Right at the start it puts the player in a shop with a random sum of gold. A light source is absolutely necessary, or the character will be blind in the dark caverns. There is no pet, though this was a feature at least since Andries Brouwer’s Hack 1.0. It also seems to be a lot tougher than Don Kneller’s PC Hack, the probability that you’ll die a few seconds into the game is high. It is interesting that it shares this feature with Mike J. Teixeira’s MAG, which according to the author was based on Jay Fenlason’s Hack. This is just a guess, but maybe Andries Brouwer toned down the difficulty a bit.

Another feature it shares with MAG is the use of color. Don Kneller’s PC Hack used the extended IBM character set pretty much the same way, but refrained from the use of color (at least in the default setting, I never really explored the options). And that’s really all I can say about it. If you want to know more, check out the already quoted NetHackWiki entry on Jay Fenlason’s Hack, it has detailed comparisons between hack121, Hack 1.0 and PDP-11 Hack.

Hack update

Okay, so maybe I should work on my reading skills. Somehow this sentence in Don Kneller’s Hack docs escaped my notice:

PC HACK is the MSDOS version of UNIX HACK which was originally written by several people at the Stichting Mathematisch Centrum in Amsterdam.

A few more questions were answered, and details added, by the NetHackWiki entry on Jay Fenlason’s Hack. 1982 was the year he started working on it, not the year of release. The Hack timeline looks about like this:

• 1982: Michael Toy and Ken Arnold speak at the USENIX conference in Boston. Jay Fenlason, who would later write the GNU profiler gprof, starts working on Hack.
• 1984: Jay Fenlason’s Hack is published on the first USENIX software distribution tape, which seems not to have survived. In December, Andries Brouwer posts Hack 1.0 to net.sources. Not long afterwards, net.game.hack is created.
• 1985: Andries Brouwer posts Hack 1.0.1, 1.0.2, and 1.0.3. Michiel Huisjes posts PDP-11 Hack and later in the year a PC/IX port. (PC/IXwas a Unix clone for IBM PCs, there is very little about it on the web.) Don Kneller starts porting Hack to PC/DOS.
• 1986: Don Kneller releases the last version of PC Hack, 3.6.
• 1987: Mike Stephenson releases NetHack 1.3d. The Hack era is over.

As for Hack121, it remains mysterious. It is the only Hack version that starts in a shop where the player can buy equipment. It does not have pets, something that was present at least since Andries Brouwer’s Hack 1.0 (nobody really knows what was in Jay Fenlason’s Hack). It is thus a somewhat independent project. Unless its creator steps out into the light at some point, we’ll probably never know who made it, where, and when.

Some Hack mysteries

Hack is somewhat the missing link between Rogue and NetHack, and it has always got less attention than these two. It seems that its history is a lot more complicated than I thought.

The first Hack was written by Jay Fenlason in 1982, with help from Kenny Woodland, Mike Thome and Jon Payne. It was a clone of Rogue, but with a lot more monsters, and distributed under the BSD license. It seems that this was only ever a Unix program, and that it has been lost.

In 1984, Andries Brouwer posted a Hack 1.0 to net.sources. He developed it up to 1.03, it was distributed under a copyright notice of the Stichting Mathematisch Centrum, Amsterdam.

In 1985, Michiel Huisjes posted a Hack for PDP-11 to net.sources in five messages (1, 2, 3, 4, 5). Apparently this version first introduced pets, in the form of a little dog.

Then there is a sourceless anonymous PC executable known as Hack121, since it is found in an archive with the name hack121.zip. It starts out in a shop where the player can buy equipment, uses the extended IBM character set and colors.

What I don’t know is how and if all these versions relate to Don Kneller’s Hack, which dates back to 1985 and which I had always assumed to be the basis for NetHack. In the doc to his Hack 1.0.3, no other Hack is mentioned, only Rogue.

As a final note, the origin of the name Hack seems unclear too. I had always understood it to be derived from the fact that Rogue had been hacked. But it may also be derived from a lesser known usage of hack at MIT (New Hacker’s Dictionary):

To explore the basements, roof ledges, and steam tunnels of a large, institutional building, to the dismay of Physical Plant workers and (since this is usually performed at educational institutions) the Campus Police. This activity has been found to be eerily similar to playing adventure games such as Dungeons and Dragons and Zork.

Note: Shortly after posting this, I noted a few inaccuracies. They are corrected in the next entry.

Heaven, Hell, and Tetris

Just some fun stuff from XKCD: Heaven and Hell in terms of Tetris. As for the comparison with sex (in the alt text of Heaven), well, there’s always this.