Dig Dug and Mario Bros Arrive on Amiga in AI-Aided Ports

Dig Dug and Mario Bros Arrive on Amiga in AI-Aided Ports

Two early-80s arcade staples have just turned up on the Commodore Amiga, and both come from the same developer. TjLaZer has released conversions of Namco's Dig Dug from 1982 and Nintendo's Mario Bros from 1983. Both were built with help from modern AI tools.

That last detail is already causing an argument in the comments over at IndieRetroNews. Before we get to that, let's look at the games themselves.

Digging Through the Dirt

Dig Dug has been rewritten for the Amiga and runs on OCS/ECS or AGA machines with 512K of chip RAM. For anyone who didn't grow up with the hardware, OCS, ECS and AGA are the Amiga's successive chipset generations. OCS and ECS cover the classic A500-era machines, and AGA powers later models like the A1200. Chip RAM is the memory those custom chips can work with directly.

According to the developer, the port uses the original graphics, accurate round layouts and the familiar sound effects. Everything you'd expect from the arcade version seems to be accounted for:

  • Enemies that ghost through the dirt to chase you down
  • Fire-breathing Fygars
  • Falling rocks to crush anything foolish enough to stand underneath
  • A high-score table for the top five diggers

The clever bit is how it handles regional differences. PAL and NTSC are the old TV standards used in Europe and North America, and the game detects which one it's running on and adjusts itself. On PAL you get the score in a side panel. On NTSC the display scrolls instead. Either way, the game keeps the arcade's original 60Hz speed, which is a nice touch for anyone who remembers PAL conversions running noticeably slower back in the day.

It's also a good companion piece to the recent Mappy port for the C64, if you're in the mood for more early Namco.

Plumbers in the Sewers

TjLaZer's second release is a Mario Bros conversion for OCS Amigas. It recreates the sewer-clearing action with a 32-color low-resolution palette meant to capture the look of the arcade cabinet. There are smooth sprite animations, and the POW block still shakes the whole screen when you hit it.

The developer is aiming for 60 frames per second on both PAL and NTSC machines. If you want to take matters into your own hands, the F5 key lets you switch the display rate manually. Two players can also team up at the same time using two joysticks, which is how Mario Bros was always meant to be played. Or fought over, depending on your sibling.

The AI Elephant in the Room

Not everyone is cheering. One commenter on the original IndieRetroNews post greeted the releases with a single word: "Slopocalypse."

The site's own reply was more measured. It said these two ports aren't bad, and that it would save the "AI slop" label for conversions that look and run poorly. It also pointed out just how crowded things have become. By its rough estimate, something like 20 to 30 AI-made games showed up in the past two weeks, compared to one or two without AI. And that's for the Amiga alone.

Other readers pushed back on the criticism. One argued that complaining about AI-assisted ports has become fashionable, and said that's unfair when these run well on a base spec A500. Another welcomed the releases, noting there used to be almost nothing in this space, and expects the quality to improve over time.

Our Take

For players, the practical upshot is simple. Two well-known arcade games now exist in Amiga form, with PAL and NTSC handled thoughtfully and features like two-player co-op intact. If the performance holds up as described, that's a solid result regardless of how the code was produced.

The bigger story is volume. This fits a pattern we've been seeing for weeks, from the AI-assisted Cho Ren Sha 68k port to a steady stream of arcade conversions. Traditionally handcrafted projects like Acidbottle's Centipede for the Amiga now share the spotlight with much faster AI-aided releases. If IndieRetroNews's estimate is even roughly right, curation is going to matter more than ever.

It is worth watching whether the community settles on clearer quality standards, and whether coverage starts to separate polished AI-assisted ports from rushed ones. For now, the best advice is to judge each release on how it actually plays.