In 1986, a twenty-something programmer in Garland, Texas released a text adventure called Beyond the Titanic through public-domain disk libraries and asked players to mail him whatever they thought it was worth. He made about ten thousand dollars, combined, off that game and a follow-up called Supernova — for two full years of work. Most people would have taken that as a sign to quit. Scott Miller took it as a data point, and the model he built in response to that failure became the business plan for an entire company, and eventually for large parts of the modern games industry.
The Numbers That Started It
Miller had been distributing games through BBS systems and shareware disk libraries since the mid-1980s under the name Apogee Software Productions. Beyond the Titanic and Supernova followed the “donationware” convention common at the time: try the whole game, pay if you liked it, on the honor system. The combined return — roughly $10,000 across both titles in a year — told him the honor system alone wasn’t enough leverage. People who already had the whole game in hand had little practical reason to pay for it afterward.
So Miller inverted the transaction. Instead of giving away the complete game and hoping for payment, he would give away a genuinely complete-feeling first act — enough content to be satisfying on its own — and hold the rest of the game back entirely until payment arrived. When a player finished that free episode, the game itself would display Miller’s home mailing address and ask them to send money for the remaining episodes. No honor system, no trust required: if you wanted episode two, you had no way to get it except to pay for it.
Kroz: Proof the Idea Worked
Kingdom of Kroz, released in 1987, was the first release built around this new structure, and it worked immediately. The ASCII-graphics dungeon crawler was primitive even by 1987 standards, but the economics were sound in a way donationware never could be, and Apogee — the company name Miller now used consistently — began generating enough revenue to fund its next releases without outside capital. Over the following years, this “Apogee model” let the company self-publish roughly its first twenty games with no retail distributor and no outside investment, at a time when getting a game onto store shelves normally required exactly that kind of backing.
The Letters
What Kroz’s success gave Miller wasn’t just money — it gave him a template he could hand to other people. Through the late 1980s, he began writing directly to developers whose shareware and public-domain releases had impressed him, pitching the Apogee model as a way to get paid for work they were currently giving away for nothing. It was, in effect, cold outreach as talent acquisition, years before “indie publishing deal” was a phrase anyone used. Several of the developers who said yes to those letters ended up defining Apogee’s catalog for the next half-decade: George Broussard, who released Pharaoh’s Tomb through his own Micro F/X label before folding his operation into Apogee outright; Todd Replogle, who’d go on to co-create the original Duke Nukem; and, most consequentially, a group of four programmers in Louisiana who called themselves Ideas from the Deep.
The Demo That Changed Everything
In 1990, Ideas from the Deep — John Carmack, John Romero, Tom Hall, and Adrian Carmack — were working for a disk-magazine publisher called Softdisk when Carmack figured out how to do smooth, console-quality side-scrolling on PC hardware, something the rest of the industry considered technically out of reach. The demo they built to prove it — a level-for-level copy of Super Mario Bros. 3 with the sprite swapped for Romero’s earlier character Dangerous Dave, sardonically titled “Dangerous Dave in Copyright Infringement” by Tom Hall — is covered in more detail in our piece on the original Dangerous Dave. The short version: the group tried to sell the underlying scrolling engine to Nintendo as an official PC port of Super Mario Bros. 3, Nintendo passed, and the team redirected the technology into an original character instead. Miller signed the resulting game before it was finished.
Commander Keen: Invasion of the Vorticons shipped in December 1990. It was a shareware sensation — the scrolling nobody else could do, wrapped around a genuinely charming original character — and it proved the Apogee model could carry a technically ambitious release, not just small platformers and puzzle games. Ideas from the Deep formally incorporated as id Software in early 1991, and the partnership between the two companies over the next three years produced Wolfenstein 3D in 1992, distributed through Apogee’s FormGen arm, and Doom in 1993, which id self-published but which validated the shareware model at a scale nobody — including Miller — had originally predicted. It’s worth pausing on how unlikely that outcome actually was: Miller was signing games from a group of programmers still technically employed by a disk-magazine publisher, betting his own distribution model on a demo built as a joke about copyright infringement. Within three years, that same relationship had produced two of the most commercially significant PC games of the decade.
Everyone Else Miller Signed
id Software is the partnership people remember, but Apogee’s catalog through the early 1990s was built on a much wider bet on independent talent. Broussard designed Pharaoh’s Tomb and Arctic Adventure before co-creating Duke Nukem with Replogle. Jim Norwood designed Bio Menace. Robert Allen built Dark Ages. Keith Schuler made Paganitzu and, later, Realms of Chaos. None of these were Apogee staff in the traditional sense at the start — they were outside developers Miller’s letters had convinced to try his distribution model instead of going it alone or chasing a traditional publisher. Because Apogee carried none of the overhead of physical retail — no manufacturing runs, no shelf space negotiations, no returns to absorb — it could afford to take chances on games a mainstream publisher of the era would have rejected outright, and the resulting catalog is unusually eclectic for a single company’s output: platformers, shooters, puzzle games, and educational titles from Karen Crowther’s Redwood Games, all funded by the same basic mechanic Miller had worked out from a $10,000 failure.
That eclecticism is easy to undersell in retrospect, because the games that made Apogee famous — Duke Nukem, Wolfenstein 3D, Doom — all skew toward action. But for every one of those, the catalog also carried something like Word Rescue or Math Rescue, aimed squarely at kids and classrooms, funded through the exact same shareware pipeline as a side-scrolling shooter. Miller wasn’t selecting for genre. He was selecting for developers who could finish a game and needed a distribution channel that didn’t require them to already have industry connections — which, in the late 1980s and early 1990s, described a lot of talented people with no other realistic way in.
From Apogee to 3D Realms
By the early 1990s, retail distribution and CD-ROM installs were becoming viable again in a way they hadn’t been when Apogee started, and the shareware-by-mail model that had built the company began facing real commercial pressure from publishers who could put a box on a Babbage’s shelf. In July 1994, Apogee created “3D Realms” as an internal label for its increasingly 3D-focused output. The transition wasn’t instant — Apogee kept publishing under its original name for two more years — but by late 1996, the company had dropped the Apogee trade name in favor of 3D Realms entirely, the same year Duke Nukem 3D shipped and became the character’s commercial and cultural peak. The Apogee name itself didn’t disappear for good, either: it has resurfaced since as an “Apogee Entertainment” label used for publishing and re-release efforts, even though 3D Realms remains the primary operating name for the company that traces directly back to Miller’s 1986 mail-order experiment.
What Actually Survived
The shareware-first, pay-for-the-rest structure that defined Apogee didn’t disappear so much as it dissolved into the rest of the industry. Playable demos, free-to-play games with paid content unlocks, freemium mobile app economics, even the entire idea of a public game “beta” that doubles as marketing — all of them are recognizable descendants of what Miller worked out from a text-adventure failure in 1986. The company itself no longer exists under the Apogee name in the way it did in 1990, but a genuinely large share of its back catalog — the platformers, the shooters, the ASCII dungeon crawler that started it all — has circled back to being free again, decades later, this time with nothing left to sell. Our verified list of what’s actually freeware today covers which of those titles you can legally play right now.