3 ms·
The Zork series (particular Zork 1) are very fond childhood memories for me. I still pull it out every year or two, and try to "speedrun" it, by completing it i
by nineteen999 7y ago
The Zork series (particular Zork 1) are very fond childhood memories for me. I still pull it out every year or two, and try to "speedrun" it, by completing it in the least possible moves. There are some great methods for pulling this off, but you are sometimes at the mercy of the RNG with respect to the roaming thief and the troll fight. Save scumming doesn't seem to work in most versions I've tried since the RNG state appears to be stored in the save game file.
When I started messing with UE4 about 4 years ago, the very first project I wanted to try was recreating Zork I, as I pictured it in my mind.
I was still learning the ropes in terms of creating good assets for UE4 (and still am to a lesser degree), and formulating my own art style, but there are some examples (Living Room and Entrance to Hades) as well as a sample treasure (Platinum Bar) here:
https://imgur.com/a/7XnzvY0 https://imgur.com/a/7XnzvY0
Other than that, I had a large part of the overground/underground map done in greybox form, and were starting to add interactive components like the rug, trapdoor, and lamp.
From time to time, I feel like going back and starting over, it's still something I would like to do one day. The conceptual problems I faced were:
1) whether to limit the game to "on rails" style game play when moving between locations, or to allow the player to freely roam in normal first-person fashion; and
2) the original game relies so heavily on the players imagination to flesh out the environment on their mind. If you take that away from them by providing a visual interpretation, would the game be as fun or interesting?
I never answered those questions, so I gave up on it and moved onto another equally as ridiculous concept (and then another, and another...)