Game 281: Galactic Trader

Today I'm doing an unplanned post at the recommendation/challenge of fellow blogger The Wargaming Scribe, who often comments here as Narwhal.

Back when I played through Doug Carlston's Galactic Empire on the TRS-80 as a prelude to the early games of Broderbund, I hadn't intended to continue the saga. The initial game was interesting, ambitious given the limitations of 16KB BASIC, and weirdly atmospheric, but Carlston would go on to bigger and better things, and Galactic Trader isn't especially well-regarded. However, after WGS covered it, and recommended I check out the third game, I decided to cover both.

 

I played this one very briefly after covering Galactic Empire and quickly decided it wasn't worth covering, but the chance to compare performances might make this more interesting, and the data-gathering aspect and trade route optimization does speak to me.

Ousted from your role as a military commander in Galactic Empire and marked by imperial assassins, you now roam the galaxy as a freelance trader in pursuit of money, which during times of peace means power - power that you'll want when the revolution happens. You make money buying goods low and selling them high, but you'll need to visit planets to learn what the rates on the goods are before you can do this effectively. It's much like DrugWars, which was, according to Wikipedia, inspired by a 1974 BASIC game called Star Trader, which also influenced Elite.

I don't fancy these odds.

Some of dangers you'll face are:

• On Galactica, the only world that deals in cash, you risk assassination with each visit.
• Cargo weighs down your ship and increases fuel expenditure, leaving you stranded in space if you run out.
• Traders may kill you if you offer them goods that you don't have.

 

I imposed these rules on myself:

• Difficulty was maximized at level ten.
• I created the save state as soon as I decided to play in earnest.
• I get one chance to finish the game. No restarting allowed.
• My only acceptable uses of save states are so that I can take breaks and resume later, and to undo  serious input errors and accidental breaks.
• I will log my trips and trades so that I can re-trace my steps should I need to load a save state.
  
• The game must be left running and unpaused when entering data or performing related number crunching.
• I may pause to take blogging-related notes.
  

 

Because Galactica is the only planet that deals in cash, and since going there can kill you, it makes more sense to me to think of fuel as the galaxy's de facto currency. This viewpoint is further supported by the fact that commodity traders outside of Galactica are slightly cagey about their own exchange rates, but the fuel cartels are up front about their goods-for-fuel rates, and as it turns out, are perfectly accurate predictors for what goods are worth on any given world.

The only time a commodity trader will ever tell you what the goods are worth.

 

I won't comment too much on the interface. It isn't great, and WGS said enough on the subject already, but I still find it an improvement over Galactic Empire. The biggest issue, one carried over from the previous game, is that if you start typing a number (e.g. an offer on goods), the game will automatically submit your input when you stop typing it. This could cause you to really screw yourself by submitting an offer to trade your entire stock of Rhiolans for 173 Hempites when they were actually worth 1,731 of them, and the last "1" didn't register. Less catastrophically but no less annoyingly, It could also mean that when you press Enter to manually commit an input, the game may automatically commit first, and then buffer your Enter keypress and process it on the next query.

Surveying the planet locations

 

My galactic chart:

	X	Y	Z	Distance	Gravity	Fuel/kg
Current	40	48	8	0	3
Galactica	40	48	8	0	5	3
Bok	29	50	8	11.18033989	3	8.590169944
Novena	27	49	8	13.03840481	3	9.519202405
Harkon	26	44	7	14.59451952	3	10.29725976
Zoe	30	35	9	16.43167673	3	11.21583836
Utopia	23	50	6	17.23368794	3	11.61684397
Farside	23	50	14	18.13835715	3	12.06917857
Proyc	34	30	6	19.07878403	3	12.53939201
Twryx	43	29	10	19.33907961	3	12.6695398
Javiny	34	30	15	20.22374842	3	13.11187421
Drassa 2	24	36	5	20.22374842	3	13.11187421
Sparta	52	27	7	24.20743687	3	15.10371844
Moonsweep	20	33	1	25.96150997	3	15.98075499
Alhambra	56	26	4	27.49545417	3	16.74772708
Kgotla	35	19	2	30.03331484	3	18.01665742
Viejo	57	18	7	34.49637662	3	20.24818831
Eventide	6	34	1	37.42993454	3	21.71496727
Ootsi	4	23	6	43.87482194	3	24.93741097
Yang-Tzu	52	5	6	44.68780594	3	25.34390297
Llythll	2	4	10	58.17215829	3	32.08607914

 

WGS noted that the farther away two planets are, the bigger the difference in price will be, and Llythll was the most distant from Galactica. In fact, nothing is particularly close to it. The farthest planet from Llythll is Alhambra.

I visited the local fuel cartel to check the exchange rates here. The interface, as noted already, is unintuitive and inefficient for a methodical playstyle, but I learned that, contrary to WGS's experience on the Apple II version, here you do not need to possess any cargo to merely offer it in trade, which will save me plenty of fuel during my intel-gathering phase. Just offer the goods, take note of how much fuel is offered to you, and be sure to reject the offer, or you'll face the cartel's wrath.

I discovered that the traders will give you fuel at 0.25 units per dollar of whatever the value of the goods you offer are worth, rounded down. Therefore, I'd always offer 4 goods to discern the true fuel value. Traders will sell you fuel at $2 per unit, so it's a better deal to sell your goods at Galactica and buy fuel with cash, but as noted, this isn't really a sustainable gameplay option.

Although I had the Galactican prices, this wasn't enough to tell me which items were expensive and which were cheap. Tenibles were the cheapest item at $13/each, and Feelies the most expensive at $382/each, but I'd really need to visit another world to compare prices.

I bought 309 units of fuel, leaving me with $382, just enough to buy a single Feelie later. It would be embarrassing if I spent all my money on fuel now, leaving me unable to start trading later! A trip to Llythll and back would cost just over half of my fuel, certainly leaving me with enough to make a return trip with some cargo.

I went to Llythll, carrying no cargo, and queried the fuel exchange rates of the goods available. Everything was more expensive here. Especially the rhiolans, which had gone up from $277 to $2582! If Galactican prices were typical of that corner of the galaxy, then the optimal strategy seemed to be to buy rhiolans there, trade them at Llythll for something that might bring a profit at home, trade several of them back for fuel (more on this later), and return home to trade for rhiolans.

Before returning to Galactica, I took a detour to Alhambra to check their prices. Rhiolans were even cheaper here.

My price chart:

	Galactica	Alhambra	Llythll
Tenibles	13	42	38
Silicates	69	4	645
Hempires	16	51	211
Feelies	382	37	702
Microbes	37	22	717
Dopae	24	282	1166
Pylates	53	72	494
Rhiolans	277	19	2582

Lacking any goods to trade here, and unable to spend cash, I returned to Galactica to buy a single rhiolan. Here, the computer informed me that on my next visit, I'd face a 66% chance of assassination. I'd have to make this trip to Galactica my last. From here on, I'd have to barter, so I spent my remaining cash on fuel.

The trip to Llythll left me with only 157 fuel, but a trader here, a Mr. Kanabe, offered me 67 tenibles for my rhiolan. The formula for what deals they'll accept is very simple; divide the total value of what you're offering by the value of a unit of what you want to get and round down to get the highest quantity you can ask for. The deal that they propose to you are always the fairest ones you'd get anyway.

A single tenible could be traded for two rhiolans back on Alhambra, so I traded all of my tenibles but one for fuel. Optimization could come later - for now I needed some reserves to ensure I could make a return trip.

The return with 2 rhiolans left me with only 50 fuel, and a big problem - the "intermediary consortium" declared exclusive trading rights on the Alhambra-Llythll route, and I could not trade them for tenibles as planned. Thankfully, I could still trade with the fuel cartel - I'd be very screwed otherwise - so I traded 1 rhiolan for 645 fuel, and embarked to Ootsi, the closest planet.

Here, my remaining rhiolan became 272 tenibles, and once again I traded all but one back for fuel. I took it to Viejo, the next-farthest planet from Llythll, where I was able to get 7 rhiolans! Bringing those back to Llythll, they fetched 475 tenibles.

Now, I had a problem that required math. Taking 475 tenibles back to Viejo would need 15,035 units of fuel, and I only had 158, which wasn't even enough to return without the cargo. I'd have to trade some of this for fuel, but how much? The more you sell, the less fuel you need.

Fuel consumption for a trip is determined by this formula:

(GravityFactor + Distance/2)*(CargoMass + ShipMass)

GravityFactor is "3" on every planet I had visited except for Galactica, which I would never visit again, and ShipMass is 10.

 

Substituting these values, and adding a variable "CargoSold," the formula becomes:

(3+ Distance/2) * (CargoMass - CargoSold + 10)

 

The amount of fuel available could be determined by this formula:

Fuel + CargoSold*CargoValue

As this amount must be greater than the fuel consumed, we get the inequality:

Fuel + CargoSold*CargoValue > (3 + Distance/2) * (CargoMass - CargoSold + 10)

 

Expressed in terms of CargoSold:

CargoSold > (3*CargoMass + 30 + (Distance*CargoMass/2) + 5*Distance - Fuel)/(CargoValue + 3 + (Distance/2))

 

After plugging in the values, and checking very carefully to ensure my math was right and rounding up distance values to err on the side of caution, I found I'd have to trade 369 of my tenibles for 3,505 fuel, for a total of 3,663. The trip to Viejo with the remaining 106 units would take 3,596. I sold 372, rounding up to a multiple of 4, to get the most fuel per tenible possible, and embarked to Viejo.

Using this formula again at Viejo, I converted 103 tenibles to 43 rhiolans plus the needed 1,536 fuel, and brought those back to Llythll, where I was once again truck-blocked by the consortium. So I traded 4 for enough fuel to get me to Ootsi, where the 39 rhiolans got me 693 tenibles plus  over 17,000 fuel. From there I went to Sparta, the next-furthest planet from Llythll, where I traded for 148 rhiolans plus 4,488 fuel back to Llythll.

One more round trip to Sparta and back, and I was hauling 515 rhiolans and 15,144 fuel to Llythll. And the consortium decided that was enough and blocked me once again, forcing another detour. Trading 12 for fuel, I went back to Ootsi, where the consortium had yet another exclusivity deal, and I had to trade another 8 to make it to Eventide. There I converted what I had left to 7624 tenibles plus over 228,000 fuel, took those to Viejo where I got 3,863 rhiolans, plus over 112,000 fuel to carry them all. Just how is my ship, which weighs the same as ten cargo units, supposed to be holding all this?

I wanted to take these back to Llythll, but stopped at Eventide on the way to avoid trespassing on the consortium's Viejo-Llylthll route. There, I reached my time limit, and the game suddenly ended.

The asset value here is exactly equal to 8 credits per rhiolan, which is what they were worth on Viejo where I got them. At Galactica rates, they'd be worth a million.

 

I replayed my original save game, using save states liberally this time. I got all of the prices ahead of time, planned my routes, and quickloaded to avoid time-wasting embargoes. I eventually found that this version of the game is unwinnable.

Say yes and the game crashes, unable to inventory 3 million silicates. Say no and it freezes, unable to calculate how many tenibles your cargo is worth.

 

Whenever you own more than a million of any commodity, it crashes, and even if you avoid this, the game starts behaving very strangely when attempting to sell quantities close to the limit, sometimes freezing, sometimes refusing to process input, and very frequently displaying numbers in E notation. Even if you could somehow acquire 999,999 of every asset, it would be worth only about 6.5 billion on this galaxy's Llythll, and you can only get cash on Galactica so many times (and we're talking hundreds of millions per trip at most).

GAB rating: Bad. Galactic Trader is a lot simpler than Galactic Empire, less interesting for it, and overall feels slapped together. The interface is a confusing mess, the mechanics unintuitive, and playing the way it seems to encourage will get you nowhere. If you don't like poring over spreadsheets like an accountant working overtime then there's no joy to be had here at all, and even I didn't have that much fun with it. The game-ruining bugs were the final nails in the coffin.

 

Here's my activity log from the "real" game, played without aforehand knowledge of prices or embargoes. Pink rows represent exploration, green rows commodity trade, and yellow rows are detours forced by the consortium.

Arrival year	Planet	Fuel	Sold	Bought	Fuel	Cargo	Galactican worth	Departure year	Fuel spent
1	Galactica	1,000	$618	309 fuel	1,309	$382	$382	15	335
73	Llythll	974			974	$382	$382	76	312
134	Alhambra	662			662	$382	$382	137	166
164	Galactica	496	$382	49 fuel 1 rhiolan	545	1 rhiolan	$277	168	388
227	Llythll	157	1 rhiolan	627 fuel 1 tenible	784	1 tenible	$13	228	357
287	Alhambra	427	1 tenible	2 rhiolans	427	2 rhiolans	$554	287	377
345	Llythll	50	1 rhiolan	645 fuel	695	1 rhiolan	$277	347	149
366	Ootsi	546	1 rhiolan	474 fuel 1 tenible	1,020	1 tenible	$13	386	333
421	Viejo	687	1 tenible	7 rhiolans	687	7 rhiolans	$1,939	424	529
480	Llythll	158	7 rhiolans	3.5k fuel 103 tenibles	3,692	103 tenibles	$1,339	490	3,506
546	Viejo	186	103 tenibles	1.5k fuel 43 rhiolans	1,722	43 rhiolans	$11,911	548	1,637
604	Llythll	85	4 rhiolans	2.5k fuel	2,667	39 rhiolans	$10,804	605	625
624	Ootsi	2,042	39 rhiolans	17k fuel 693 tenibles	19,402	693 tenibles	$9,009	628	18,979
676	Sparta	423	693 tenibles	4.4k fuel 148 rhiolans	4,911	148 rhiolans	$40,996	680	4,805
735	Llythll	106	148 rhiolans	73k fuel 2.3k tenibles	73,294	2,352 tenibles	$30,576	736	72,026
792	Sparta	1,268	2.3k tenibles	15k fuel 515 rhiolans	16,412	515 rhiolans	$142,655	793	16,023
848	Llythll	389	12 rhiolans	7.7k fuel	8,135	503 rhiolans	$139,331	850	6,418
869	Ootsi	1,717	8 rhiolans	3.8k fuel	5,527	495 rhiolans	$137,115	871	4,552
883	Eventide	975	495 rhiolans	228k fuel 7.6k tenibles	229,095	7,624 tenibles	$99,112	888	225,206
941	Viejo	3,889	7.6k tenibles	112k fuel 3.8k rhiolans	116,241	3,863 rhiolans	$1,070,051	945	114,257
998	Eventide	1,984				3,863 rhiolans	$1,070,051

Game 280: One on One - Dr. J vs. Larry Bird

The first sports game I ever covered was Atari's Basketball from 1978. A simplistic one-on-one basketball game, it nevertheless anticipates Electronic Art's One-On-One, which debuted on the Apple II in 1983 before being ported to Atari, Commodore, and IBM computers.

Though still simple by today's standards, One-on-One offers a lot of refinement and bells and whistles, and at least symbolically marks the beginning of the EA Sports dynasty.

Progress?

 

The manual has EA's typically high production values of the day, featuring photos, biographies, and interviews with Julius Erving and Larry Bird, who were consulted during the game's design sessions. 

Notably for the time, the players are not identical. Dr. J has better speed and dexterity up close, and superior hang time and dunking ability, while Bird brings stronger defense, better rebound game, and more accurate outside shooting. On the face of it, the advantage seems lopsided in Bird's favor. Standard play rules give the ball over to whoever loses the point, and offense scores more often than not, so victory seems to go to whoever can successfully defend more often than the other guy.

There are a few gameplay options:

• Game type - Four options from easiest to hardest: Park and rec, varsity, college, and pro. The more hardcore the game type, the more strictly the rules are enforced by the referee, and in singleplayer games, the better the AI plays. Default is pro.
  
• Play mode - Two player, singleplayer vs. Dr. J, singleplayer vs. Larry Bird
• End of game - Timed or to a set score (1 to 99, default 21). Default is timed, with four-minute quarters.
  
• Possession -  Point winner gets offense, or point loser gets offense. Defaults to point loser.
  

 

The Apple II version also supports Mockingboard music, making it the first such game I've ever seen apart from the Ultima series, playing Maple Leaf Rag during the title screen and options menu, but this is the only time it plays any music at all.

I played a round with "B" following standard rules. He played Bird, I played Dr. J.

 

We both had some trouble at first figuring out how to pull off an effective defense, but "B" pulled way ahead during the fourth quarter once he did, and neither of us felt like a replay.

Controls, just like in Atari's basketball, are based around a single button joystick, but you can do more things thanks to context sensitivity. On offense, a quick tap spins you 180 degrees, and you can execute shots or dunks by holding and releasing with the right timing. On defense, the button steals or jump-blocks, depending on where the ball is, though the graphics make it hard to tell if a steal was successful or not.

Some other features here are:

• Fouls
• Free shots
• Three point shots
• Fatigue bars
  
• Hot streaks (alluded to in the manual as an invisible game mechanic - were there "hot streak bars" in previous games?)
• Instant replays
  
• Breaking the backboard, complete with an irate janitor's glower as he sweeps up the shards
• A horrible sounding crowd cheer effect every time anyone scores a basket

 

After playing a round with "B," I found an almost surefire way to beat the computer in pro mode. Play as Bird, and always take your shots from the three-point line. When you're on offense, they'll usually go right over Erving's head and sink, as long as you move directly down and not at an angle. On defense, go for the rebound when you can, but don't fatigue yourself with too much exertion in pursuit of the ball. The AI as Erving only ever goes for two-point shots, so it's not that difficult to gain a big lead over him through reliable 3-pointers, even if you don't block his own offense most of the time, and even if you miss on occasion. On my first try, I beat the AI on pro mode, scoring over 100 points and flipping the counter.

Stupid game didn't even acknowledge my victory.

GAB rating: Average. I'm not really the target audience for this game. I recognize the strides made toward realism compared to contemporaries, but neither "B" nor I had that much fun playing it.

 

Data Driven Gamer will return in August to wrap up 1983 with Wizardry III!

Games 278-279: Floppy Frenzy & Digger

Windmill Software is an oddity; a software studio that made arcade-style games exclusively for the expensive, business-oriented IBM PC, and did it during the first few years of the product's existence, back when it wasn't especially popular as a home computer or very well suited to these sorts of games.

Windmill's first game, made in 1982, was Video Trek 88. A variant of that ancient BASIC game Star Trek, I'm not very interested in playing it, but its color text display does illustrate one advantage that CGA had over its 8-bit contemporaries; text is very crisp, effectively rendering at a 640 pixel wide resolution, and supports 16 colors or monochrome.

Screenshot by Mobygames

Most of their subsequent games, however, utilized that notoriously ugly 4-color 320x200 mode, though they certainly did their best to make the most of it.

 

Game 278: Floppy Frenzy

 

Windmill's second game, released in 1982, on first glance looks like a Pac-Man knock-off. You guide a floppy disk around a maze, avoiding magnets and dust which, as every computer user in 1982 knew, was deadly to floppies. By laying down traps, you can weaken them and then chase down and kill them. Fail, and you get a cute Game Over animation.

 

On closer inspection, though, this game is a dead ringer for something far more obscure than Pac-Man - Heiankyo Alien! You go through a maze, lay down stationary traps, and have to just hope that your enemies wander into them. Apart from having to chase them down once they fall in, it's nearly identical. Did Windmill somehow get to play this Japan-exclusive arcade game, which was obscure even then? Did they arrive at virtually the same gameplay concept and execution by accident? I like to think that they might have played Space Panic and copied the formula but without the ladders and platforms, coming full circle back to a game that plays just like Heiankyo Alien.

And just like Space Panic and Heiankyo Alien, the enemies move with such randomness and indifference to you and your traps that the game becomes chaotic and defies strategy. The F1 key, used to set traps, is unreliable too - the game always acknowledges the attempt with a beep, but may still decline to leave a trap at all, simply because your floppy wasn't perfectly aligned with the invisible pixel position needed to set one down.

GAB rating: Bad. Even on the early IBM, Windmill could do a lot better than this. And they did.

Game 279: Digger

Digger, the best known game from Windmill, is a shameless clone of Mr. Do! You run around a dirt maze, digging tunnels that you and the monsters can pass through, and win by collecting all the stuff - diamonds rather than cherries - or by killing all of the monsters.

There are some differences, mostly simplifications:

• You have a simple forward-firing bullet instead of a SuperBall.
• The trick of pushing a moneybag just over the edge and waiting doesn't work like it does with Mr. Do's apples. They'll just fall.
• There is no equivalent to Alpha Monsters, and no way to spell EXTRA.
• Instead of junk food there are cherries, which also appear when the final monster spawns, but collecting simply makes you invulnerable for a few seconds. However, any monsters killed during this time will respawn, and possibly overwhelm you, so you've got to decide whether to spend this time pursuing monsters (and points) or safely collecting diamonds to end the level.
• Moneybags break open when they fall and the contents can be collected for 500 points which is significant, but it's risky to backtrack to get it, and they can be taken by the monsters as well.

 

The somewhat simplified nature of Digger makes it feel even more difficult than Mr. Do!, where you had more options, both for dealing with the monsters and for gaining more lives, and the craziness sometimes worked to your advantage. In Digger, when you lose control of the playfield action, you die. Allowing the monsters to go berserk and start digging on their own will be your doom.

 

Windmill Software makes the best of the period PC's hardware, with surprisingly okay 4-color CGA graphics, cute little animated sprites, and a funky beeper rendition of Popcorn, though I would have liked to see 16-color composite support. As with most early PC games, logic is tied to the CPU clock, and default DOSBox settings run everything much too fast. I found 350 cycles was enough for smooth gameplay, but the Funeral March that plays when you die gets cut off slightly too soon, and 250 cycles allows you to hear the whole thing. DOSBox's timing emulation isn't 100% precise, so maybe my inability to find a perfect setting isn't necessarily the game's fault.

GAB rating: Good. Sure, it's not exactly original, nor does the system it's made for the one the best suited for it, but Digger is one of the best arcade-style computer games I've played yet.

Game 277: Chuckie Egg & early A&F Atom games

 

Atomulator's instructions said to do this.

I think I have a new least favorite machine to emulate. The Acorn Atom, a predecessor to the BBC Micro series, presents an environment where disk and tape images are hard to find, provenance doesn't exist, and guides on how to actually use the computer once you're emulating it are sketchy, inaccurate, or just plain wrong. Even without these issues, the Atom seems like a real dog of a computer, with only 2KB of RAM, a weird form of BASIC, a terse and confusing command line interface, and an unresponsive keyboard that turned LOAD "MINEFIELD" into LOD"MINFL" when I typed it too fast for the polling interval to keep up.

A&F Software, British bedroom game design studio named after its teenage founders Nigel Alderton and Mike Fitzgerald, started out with a focus on this arcane analytical engine before making Chuckie Egg, our next whale.

The Chuckie Egg Professional Resource Kit catalogs an impressively complete-looking list of A&F's games, featuring many titles that aren't even listed on Mobygames, though it curiously lacks Minefield for the Atom, even though Mobygames list this as their first published title.

I've selected Minefield and Cylon Attack, both for the Acorn Atom, as ancestors to Chuckie Egg. Figuring out how to emulate these games was a tedious set of trials and errors, and I'll spare you the details on the many paths I took that wound up going nowhere, but the emulator that wound up doing the job was Atomulator, which in version 1.30 anachronistically emulates a MicroSD card reader, of all things. Loose program files dropped into the "mmc" folder can be loaded into the machine's memory and ran.

Ultimately, though, neither proved substantial enough to be worth diving into. Both are pretty poor games, even considering their humble ambitions.

The goal of Minefield is to clear a field of mines with your tank by shooting at the mines before they detonate and steering clear of the explosions. You probably can't shoot them all, but as long as you don't get caught in the resulting chain reaction, you'll be fine. It does raise the question of why you even need to be there. The tank drives forward on its own and may be rotated with the A and Z keys. Ammo and fuel are limited but can be replenished on the field.

The chain reactions of exploding mines are an interesting mechanic, but this game is overall primitive and controls awkwardly.

Cylon Attack fares better, but still plays like a poor man's version of Exidy's Star Fire, or any number of better clones that followed. I'm actually moving my ship around in the above shot, but since this only moves the enemy ships' relative positions and not the star field or my crosshair, it doesn't seem like it, and it's unclear what determines when you get hit.

Anyway, that was all just build-up to what A&F are clearly best known for.

Game 277: Chuckie Egg

 

Released in 1983 simultaneously on the ZX Spectrum, BBC Micro, and Dragon computers, the original design seems to be the Spectrum, programmed by Alderton while Fitzgerald put together a Micro conversion.

One neat feature here is the ability to redefine keys. I don't think I've seen this in a computer game yet, but it's nice to not be forced to use the Spectrum's weird QAOP tradition. A feature I couldn't care less about is having up to four players, who can only alternate. The cassette case also claims compatibility with a "Fuller Speech Box" peripheral, but I couldn't figure out how to enable this.

It's another single-screen side-view platformer, with ladder mechanics reminiscent of BurgerTime or Space Panic. You're just trying to collect all twelve eggs while avoiding the geese, who are assholes and will kill you, just like in real life.

For once, the first level is pretty easy, in spite of awkward-feeling jumping mechanics that I've come to expect of these Spectrum platformers. The geese move slowly, and aren't all that interested in you. They just roam around the level, eating bird seed if they happen to come across it and being a mild nuisance. They'll get in your way, but when you can't find a way to get around them, you can usually wait for them to go somewhere else. There's no fall damage, and although the ladders are maddeningly precise about where you need to be standing before you can climb them, the game recognizes diagonal input and using this makes grabbing them so much easier.

Level 3 is where things start to get irritating.

Moving platforms combined with clumsy platforming mechanics and awkward jumping controls that make it difficult just to do a standing forward hop make for a lot of frustrating deaths. Every single failure here felt like a consequence of losing a fight with the controls.

The subsequent levels, unfortunately, don't get any more interesting. You're just looking at the same platforms, elevators, ladders, and geese, and about the most clever thing I saw were occasional eggs suspended in places where it was impossible to reach them and live, forcing you to collect them last or sacrifice a life.

Level 8 had no elevators, and consequently gave me less trouble than any level before it with them. The main challenge was having enough patience to wait for those stupid geese to get out of my way. The second hardest part was timing and aiming standing jumps correctly to get the eggs suspended in air.

Unlike every stage from 3 to 7, I beat this one the first time I reached it, though not on a single life. After that, it loops back to the first level, only now instead of dealing with idiot geese, a giant duck chases you around the level.

While this does admittedly make things more interesting, my run came to an end once I hit level 3 again, where waiting for the elevator isn't an option, and rushing onto it is suicidal.

I stopped playing here. The game will repeat forever, but subsequent loops bring the geese back, then add additional geese, and then make them move faster.

GAB rating: Below Average. There's just not much to see here, and like Manic Miner, that other ZX Spectrum platformer of 1983, the challenge comes mainly from its unwieldy controls, though Chuckie Egg is far more forgiving.

Game 276: Alley Cat

I've been looking forward to playing Alley Cat on its original platform. Like most of its players, I'm sure, I first played its PC CGA port, which had been programmed as a self-booting disk by the game's original creator Bill Williams, and got distributed widely as a pirated DOS conversion.

To be honest this is one of the better looking 4-color CGA games.

Nowadays, I have a better appreciation for how much more well-suited Atari and Commodore computers were for this sort of game, and with cycle-accurate emulators with modern features and conveniences, accessibility isn't a problem. At the time, though, Alley Cat was just a fun little DOS game that I could play for a good hour when I had one to kill, and it happened to work perfectly in Windows XP with no effort.

I spent a few days playing Alley Cat via Altirra. My best attempt cleared three rounds, making it to the fourth and most difficult setting but could not clear it.

 

The premise is that you, Freddy the alley cat, roam the back alley of the sleazy Catalina Condominiums, hoping to court the local white-coated queen Felicia. It's a dangerous place where vicious dogs patrol the ground level, and should you ascend to the upper levels by hopping from the trash cans to the fence and then grabbing the clothes lines, the tenants will, displeased by all the racket you're making, toss things at you through the windows. You can catch mice up here for extra points, and if you can slip through a window, you'll enter a new area and get a chance to cause some real trouble. Do well enough and Felicia might give you a shot, but first you'll have to get past her six brothers.

Alley Cat on the Atari feels a lot more difficult than I remember it. Freddy moves more fluidly, as you might expect, but his standing hops don't have a whole lot of horizontal travel, and you've got to run a fair distance before picking up enough speed for a leap to count as a running jump.

No place is safe for you either. Stay on the street too long, and a dog rushes onscreen and kills you. Stay on top of a trash can too long and a feral cat knocks you onto the street, where a dog rushes onscreen and kills you. Up on the clotheslines, the garbage thrown at you through the windows is nearly impossible to avoid if it comes out close enough, and will also kill you. And it's easy to fall from your perch thanks to a mistimed jump, or to get knocked down by the mice skittering across the clotheslines, but at least this isn't fatal.

If you can sneak into an open window, one of five minigames plays out.

The Aquarium Room is dead simple. Just get to the fishbowl, done simply by jumping from the floor to the table it rests on. A broom here, found in all of the rooms, will chase you around, shooing you out the window, and if you take too long the dog will enter, but reaching the fishbowl should be easy. Once you do, the real minigame starts.

Wait, is this fishbowl a TARDIS?

You've got to get all of the fish, don't drown, and don't touch the electric eels. This is trickier than the PC version, as here your swimming controls are heavy and imprecise. Freddy bobs slightly upwards, and fights inertia has he changes directions or stops moving. It's best to go for the easy fish first, and then wait in a safe spot while the trickier ones hopefully move away from the eels. You don't have a time limit; you just need to surface for air when Freddy starts to turn purple.

Succeed in a minigame and you get a time-based bonus, and the next window you enter will be Felicia's lair.

You've got to reach the top platform where Felicia lives, avoiding her brothers which is no simple task as they track your horizontal movements and do it fast. Contact with one knocks you down a platform, and probably right into contact with the one below. The whole time, cupids are firing arrows at you, which turn the heart platforms into broken hearts and then ricochet off the bottom and revert the broken hearts back, which very quickly screws up the platform configuration of the room. You do get a box, which can bribe one of the cats to leave you alone for awhile, but on the harder difficulties this might not be enough unless you get very lucky with the platform layout.

Thankfully, you don't lose a life if you fail, though you will be dumped back into the alley and have to complete another minigame before Felicia gives you another try. When you make it in again, you'll get an extra box to play with, up to a maximum of three. Succeed and you go back to the alley, and the difficulty level increases.

The fourth and hardest difficulty is where my game ends every time. Just getting onto the fence is nearly impossible; only one trash can is high enough to reach it, and the only way onto it is to take a running jump from the right edge of the screen. You're screwed if you're to the left of it; running all the way to the right and running back takes too long to avoid triggering the dog. I've made it into the windows here before, but have yet to finish a minigame.

In another minigame, you sneak around a room full of sleeping dogs, trying to drink their milk without waking them up. It's probably the easiest minigame, requiring more patience than skill. Back off when you hear them growl.

The broom makes things annoying, but you can distract it by leaving pawprints on the floor, which works in every other minigame as well. A carton of milk periodically refills the bowls, but once a bowl is empty, it's gone. It's best, therefore, to prioritize the bowls closest to the dogs first, allowing the farther ones to get refilled, as you can easily go for those later.

The necromancer's spiders return! The manual even refers to them as the pets of an eccentrick "Nick Cromancer." Climb the bookshelf and knock his prized flowers off the top to finish this room. Mastering the jumping trajectories and understanding where Freddy will and won't grasp the shelves is key to reaching the flowers without touching the spiders.

 

In my favorite minigame, four mice crawl around an enormous and unsanitary hunk of Swiss cheese. Catch them all to win. You can also crawl through the holes to reach tricky areas or even intercept the mice if you know where they'll take you, but the broom will once again chase you, and leaving pawprints to distract it is risky as this attracts the dog.

In my least favorite minigame, you have to knock a birdcage off a table while avoiding the broom. You've got to nudge it several times, and every time you do, you get pushed back onto the floor, where you've got to quickly jump back onto a chair and make your way back to the table while avoiding the broom. And each time there's a chance that the dog will enter.

If you're lucky, the broom might knock you right onto the table, where you can just push the cage off of it, but you'll still have to catch the bird afterward. And I just love accidentally jumping out the window while frantically jumping around the room in its pursuit.

 

GAB rating: Good

Alley Cat is just so charming, original, tightly designed, and fun that I can't not give it a harpoon. It's not an epic adventure that will take days to finish; it's a time waster that you'll play for a few minutes, but they'll always be enjoyable minutes, which is what every arcade-style game strives for, and few achieve. Williams always had a knack for creative ideas, even if they didn't always translate into a playable game, and for the first time, there's enough polish and refinement that they work.

We're done with Synapse Software, which continued to have a prolific, Atari-focused output through 1984, when they were bought out by Broderbund. Williams continued to have a career as a developer up through 1993. His next several games targeted the Amiga, and most were independent works save for Cinemaware's Sinbad and the Throne of the Falcon, but Alley Cat remains his most most enduring classic.

Games 274-275: Salmon Run & Necromancer

One year before writing Alley Cat, its designer Bill Williams had written two other Atari computer games. This post is about my experience playing them.

Game 274: Salmon Run

Unlike their incredibly unsuccessful attempt at creating a walled garden of internally developed and uncredited Atari 2600 cartridges, the Atari Program Exchange encouraged programmers, whether amateur or professional, to write computer software. Select titles would be listed in a quarterly catalog distributed to all registered customers, with credit, royalties, and occasionally, cash prizes given to their authors.

Williams' first game, Salmon Run, was distributed this way. Save for the oddly realistic-sounding water effects, it feels like something that might have existed on the Atari 2600.

 

As Sam the Salmon, you've returned to the rivers of your youth to embark on that one last adventure that millions of your ancestors took before; to swim upstream, leaping over waterfalls, avoiding opportunistic bears, seagulls and boys with fishing nets, and, em, find true love.

Oh, Bill.

Unlike in real life, phenoptosis doesn't set in, but instead you gain an extra life and get to have another run, only it will be more difficult, not only because the river will have additional waterfalls and hazards, but Sam isn't getting any younger. The longer you go on a single life, the weaker Sam gets - jumps become shorter, currents affect you more, and you swim slower. It's an interesting take on the limited lives mechanic that I've never really seen done before.

Unfortunately, I can't seem to beat a second loop at all, because Salmon Run is really hard. The random hazards just seem to sometimes appear in configurations that are completely impossible to avoid.

Like many of Atari's games, Salmon Run has multiple game modes, 8 to be exact, which represent 8 combinations of three binary settings:

• General waterfall concentration (Normal/High)
  
• Subsequent rounds strengthen the current (Off/On)
  
• Subsequent rounds add more waterfalls (2/4)
  

 

Mode 8 has all of these enabled, and it was here that I achieved my highest score.

 

GAB rating: Bad. I feel bad about condemning this earnest, well-meaning, and novel effort, and almost think I must be missing something, but Salmon Run just feels frustrating and annoying to play. It handles weirdly, collision detection seems glitchy, and I just can't see a viable strategy for success when you're at the mercy of randomly spawning predators in places where you can't avoid them.

Game 275: Necromancer

Williams' second game was his first with Synapse Software, and is a bit more substantial than Salmon Run. And a lot more weird.

The first and most striking thing about Necromancer is its atmosphere. This game is dark, grim, and oppressive! Cliche nowadays, but there's a style and sincerity that I just don't get from modern corporate-created grimdark settings. Williams' low resolution artwork feels more surreal rather than abstract, partly due to skillful use of palette cycling tricks, but also, I think, because Atari games aren't usually this black. Let's also not fail to mention the title screen theme, a simple but haunting melody, made at a time when most computer games didn't have any music at all.

You play a druid, and must defeat a necromancer and his army of spiders with your magic powers and army of trees, over the course of three stages, where your performance in one affects conditions of the next.

The first stage is about raising your trees while orcs threaten to cut them down.

Your joystick controls a wisp, and it's difficult to move it around precisely. Pressing the button plants a seed, which will grow into a tree, sometimes very quickly, sometimes agonizingly slowly, but it's vulnerable until fully grown. The wisp can chase away the orcs or destroy them, at the cost of some strength, and can retrieve more seeds from the occasional dryad. Spiders will poison your trees, causing them to start screaming, and will die soon unless you heal them with your wisp. As you plant more trees, the "level" increases, and the tempo of the action with it, up to a maximum level of 5. Once your energy runs out, which is reduced by killing orcs or getting hit by a spider, you gain 200 strength for each level attained, and take your trees with you to the next level.

I found a pretty good strategy for ensuring that you reach level 5.

Plant your first four trees in the corners, one at a time while the level is still low, and watch them like a hawk to make sure each one of them grows. Then go nuts on the top and bottom rows - the corner trees will protect them. You've still got to watch for spiders, and they can infect multiple trees in one pass, but curing your sick trees is a lot easier when they're all standing in a row on an edge of the screen.

Stage 2 is the catacombs, and here, things fall apart. The goal is to crush as many sacs as possible with your trees, which is done with a truly bizarre control scheme and needlessly convoluted rules.

To command a tree, first move the wisp to your tree, and then your joystick moves the tree, but the control scheme is an uncomfortable scheme where you tap a direction to start moving, and tap up or down to stop moving (unless you're by a ladder), like keyboard controls in an Apple II game - totally unnecessary as these are joystick controls. The druid may be moved by holding the joystick button, and his direction is changed by tapping a direction while still holding it, which is also pretty confusing. Question mark icons may be collected to extend ladders, but you must be careful to avoid the grasping hands - if they get your druid it will cost 300 strength, and you just can't afford that.

To crush a spider sac, move a tree above it and press the button to plant it. Eventually roots will take hold, the ground will crumble, and the tree will fall. Then you may burn the tree with your wisp, reclaiming it.

There's a lot of luck involved here, because the sacs will hatch eventually, and in a random order, with random timing. A sac will start glowing when it's about to hatch, but sometimes it's too far away to do anything about it. The trees move very slowly, and the grasping hands slow you down even more. The amount of time it takes to hatch is random, the amount of time it takes a tree to take root is random, and sometimes a spider just hatches before you can do anything, even if you acted immediately.

That's not the end of the world on level 1, where everything moves slowly and you can kill the spider with your wisp, gaining 100 strength as a reward, but around level 3 they start moving so ridiculously fast and deftly avoid your wisp that they'll wreck havoc. If you planted a tree above, it will certainly destroy the tree. If you wrote that sac off as a lost cause and tried planting trees far away, and they didn't take root fast enough, the spider will likely destroy them. It might even get into your tree reserve and destroy everything. And ultimately, unless you get extremely lucky, it will touch you, costing you 100 strength and ensuring you'll see it again in the next stage.

What was I even supposed to do?

Even after learning the controls and mechanics, I found this stage random, chaotic, and unfair, and frequently did things I hadn't intended thanks to the weird controls. Often I'd plant a tree, only for this to cause the druid to walk right into a hand, as the button for planting trees and moving is the same.

Full disclosure - I had to horrifically abuse save states to pass this stage with a minimum of eggs left uncrushed, partly due to the weird controls, but mostly due to the extreme amount of luck manipulation involved. And even then it wasn't perfect by any means.

If you just want to pass the stage, you can do this fairly easily by ignoring the eggs and just rushing to the bottom, being extremely careful to avoid the hands. Better yet, on levels 1 & 2, wait for the eggs to hatch, and kill the spiders with your wisp until your strength maxes out at 2000. This, I can do without save states, but you'll be screwed on the next stage. Or at least that's what the manual says.

Descend past level 5, and you arrive at the graveyard, and showdown with the necromancer himself.

 

Topple the headstones, avoid the spiders, and shoot the necromancer for bonus strength. The wisp is incredibly hard to control here! Sometimes it kills spiders, sometimes it doesn't, but it hardly matters as the necromancer will just regenerate any spiders that you do kill. The controls still use the backwards scheme of controlling the wisp when the button is released and moving the druid by holding it.

The manual says that the more sacs you leave uncrushed (or hatched) in the previous level, the more spiders you have to deal with here. In truth, it seemed like even when I manipulated luck to get most of them, his army was still overwhelming and unmanageable.

There are five levels, and it is critical that you "farm" the necromancer for strength during levels 1&2 where it's still possible to do this faster than the spiders can drain your strength. Do not finish these levels with less than the maximum of 2000 or you'll be very, very dead during the next three. Even if you do, you're probably still going to die.

Once again, I had to use save states to ensure not too much life was lost on each subsequent level. You can't control where the spiders spawn, they move faster than you and unpredictably, it hardly seems worth the trouble to try to kill him, and the necromancer himself moves too quickly to farm for more life. You can easily lose 500 strength in a fraction of a second. It all feels very luck-based, and almost hilariously skewed against your favor. And that's if you can even decipher all the chaos going on the screen.

Beat level 5 and the necromancer asplodes and fills the screen with rainbows.

GAB rating: Bad. Once again, I feel awful about condemning this. It's an original, authorial title, with strong art direction and so many novel ideas. We've seen minigame collections before, where each stage is a different arcade game knockoff, but not only does every Necromancer stage seem like something new, but there's actual gameplay cohesion, and not just a bunch of unrelated minigames. You plant trees in stage 1 so that you can squash spider nests in level 2, so that you can deal with fewer of them in level 3.

But obtuse controls, weird mechanics, and grossly unfair gameplay ruin the experience. I had no fun playing Necromancer during my attempts to do it fairly, and no fun savescumming during stages 2 & 3 every few seconds to ensure the game would be beatable.

Thankfully, I already know the next game is good. Really good.