IIRC, though, that isn't "give developer some more money and keep plugging". It was "take the game in its current state, hand it to another developer to get it into a releasable state, and ship it".

googles

Yeah. Basically, 3D Realms just kept kicking the can down the road. Gearbox took over, cleaned up what was there, and shipped it in half a year. It wasn't the perfect, ideal 3D FPS, but I suspect that cleaning up what was there and making what return was possible (and at least getting the people who had preordered the game many years back) was probably the right move. I don't think that 3D Realms was going to produce a huge success if they had another two years or something. It probably would have been a good idea to have wrapped up the project several years earlier than was the case.

https://en.wikipedia.org/wiki/Duke_Nukem_Forever

In 1996, 3D Realms released Duke Nukem 3D. Set apart from other first-person shooter games by its adult humor and interactive world, it received positive reviews and sold around 3.5 million copies.[8] 3D Realms co-founder George Broussard announced the sequel, Duke Nukem Forever, on April 27, 1997,[9] which he expected to be released by Christmas 1998. The game was widely anticipated.[8] Scott Miller, 3D Realms' co-founder, felt the Duke Nukem franchise would last for decades across many iterations, like James Bond or Mario.[8] Broussard and Miller funded Duke Nukem Forever using the profits from Duke Nukem 3D and other games. They gave the marketing and publishing rights to GT Interactive, taking only a $400,000 advance.[8] 3D Realms also began developing a 2D version of Duke Nukem Forever, which was canceled due to the rising popularity of 3D games.[10]

Rather than create a new game engine, 3D Realms began development using Id Software's Quake II engine.[8] They demonstrated the first Duke Nukem Forever trailer at the E3 convention in May 1998. Critics were impressed by its cinematic presentation and action scenes, with combat on a moving truck.[8] According to staff, Broussard became obsessed with incorporating new technology and features from competing games and could not bear for Duke Nukem Forever to be perceived as outdated.[8] Weeks after E3, he announced that 3D Realms had switched to Unreal Engine, a new engine with better rendering capabilities for large spaces, requiring a reboot of the project.[8] In 1999, they switched engines again, to a newer version of Unreal Engine.[8]

By 2000, Duke Nukem Forever was still far from complete. A developer who joined that year described it as a series of chaotic tech demos, and the staff felt that Broussard had no fixed idea of what the final game would be.[8] As the success of Duke Nukem 3D meant that 3D Realms did not require external funding, they lacked deadlines or financial pressure that could have driven the project. Broussard became defiant in response to questions from fans and journalists, saying it would be released "when it's done".[8] In December 2000, the rights to publish Duke Nukem Forever were purchased by Take-Two Interactive, which hoped to release it the following year.[11] By 2001, Duke Nukem Forever was being cited as a high-profile case of vaporware, and Wired gave it the "vaporware of the year" award.[12]

At E3 2001, 3D Realms released another trailer, the first public view of Duke Nukem Forever in three years. It received a positive response, and the team was elated, feeling that they were ahead of their competitors. However, Broussard still failed to present a vision for a final product. One employee felt that Miller and Broussard were developing "with a 1995 mentality", with a team much smaller than other major games of the time. By 2003, only 18 people were working on Duke Nukem Forever full time.[8] In a 2006 presentation, Broussard told a journalist the team had "fucked up" and had restarted development.[8] By August 2006, around half the team had left, frustrated by the lack of progress.[8]

According to Miller, the Canadian studio Digital Extremes was willing to take over the project in 2004, but the proposal was rejected by others at 3D Realms. Miller later described this as a "fatal suicide shot".[13] In 2007, 3D Realms hired Raphael van Lierop as the new creative director. He was impressed by the game and felt it could be finished within a year, but Broussard disagreed.[8] 3D Realms hired aggressively to expand the team to about 35 people. Brian Hook, the new creative lead, became the first employee to push back against Broussard.[8] In 2009, with 3D Realms having exhausted its capital, Miller and Broussard asked Take-Two for $6 million to finish the game.[8] After no agreement was reached, Broussard and Miller laid off the team and ceased development.[8] However, a small team of ex-employees, which would later become Triptych Games, continued developing the game from their homes.[14]

In September 2010, Gearbox Software announced that it had bought the Duke Nukem intellectual property from 3D Realms and would continue development of Duke Nukem Forever.[15] The Gearbox team included several members of the 3D Realms team, but not Broussard.[15] On May 24, 2011, Gearbox announced that Duke Nukem Forever had "gone gold" after 15 years.[16] It holds the Guinness world record for the longest development for a video game, at 14 years and 44 days,[17] though this period was exceeded in 2022 by Beyond Good and Evil 2.[18]

In 2022, Miller released a blog post on the Apogee website about 3D Realms' failure to complete Duke Nukem Forever. He attributed it to three major factors: understaffing, repeated engine changes and a lack of planning.[13] On Twitter, Broussard responded that Miller's claims were "nonsense", described him as manipulative and narcissistic, and accused him of blaming others. He blamed Miller for the loss of 3D Realms and the Duke Nukem intellectual property.[13]

I think that one key phrase there might be important: "As the success of Duke Nukem 3D meant that 3D Realms did not require external funding, they lacked deadlines or financial pressure that could have driven the project." Like, this is maybe a good example of where they really did need someone outside the project to say "I need you to get milestones and a schedule in shape", and where more money and time isn't the right answer. It's not that the project is on the cusp of amazing success and the people managing the project just mis-estimated the schedule by several months. It's that they just aren't anywhere near where they want to be and don't have a realistic roadmap for getting there.