The linked paper is waaaay more interesting and detailed. The motivating context of early video games as digital artefacts is much clearer. In Section 3 they actually explain the code. There's a...
The linked paper is waaaay more interesting and detailed. The motivating context of early video games as digital artefacts is much clearer. In Section 3 they actually explain the code. There's a much weaker conclusion than the article implies:
Our conclusion is that the table values were manually chosen, or manually tuned, by the maze algorithm designer.
Great. So much for "a mystery bit of code they couldn't explain", whose logic "has been lost forever". Might as well give up analyzing Ulysses because of a missing page in the manuscript. 🙄
They also discuss the PRNG (pseudo-random number generator), which contains a small bug ― this bug is present in a few other contemporary games, which suggests that they're related. Who was the original author of the PRNG? How did the code get into these games? These are interesting archaeological questions.
Yes, a bit. I was really interested in the maze generation story. It was a bit like reading a crime drama only for the last paragraph to be “All that said, Detective Bartlebe retired from the...
Yes, a bit. I was really interested in the maze generation story. It was a bit like reading a crime drama only for the last paragraph to be “All that said, Detective Bartlebe retired from the force having never found the killer. The End.”
It's funny because I've often referred to doing software archeology, but on a slightly more recent time scale. I'm currently working on a 20-year old codebase. It's been updated continuously...
It's funny because I've often referred to doing software archeology, but on a slightly more recent time scale. I'm currently working on a 20-year old codebase. It's been updated continuously during that time, but its history is preserved in source control. I have gone back through the commit comments and older revisions to see how parts of it were written previously and to understand why some constructs exist. (I've also shown that some existing constructs are no longer necessary because the things they supported no longer exist!) It's really interesting to do! I don't usually have much time to do it, but when I do, it's quite satisfying.
The linked paper is waaaay more interesting and detailed. The motivating context of early video games as digital artefacts is much clearer. In Section 3 they actually explain the code. There's a much weaker conclusion than the article implies:
Great. So much for "a mystery bit of code they couldn't explain", whose logic "has been lost forever". Might as well give up analyzing Ulysses because of a missing page in the manuscript. 🙄
They also discuss the PRNG (pseudo-random number generator), which contains a small bug ― this bug is present in a few other contemporary games, which suggests that they're related. Who was the original author of the PRNG? How did the code get into these games? These are interesting archaeological questions.
Did anyone else find the article really disappointing?
Yes, a bit. I was really interested in the maze generation story. It was a bit like reading a crime drama only for the last paragraph to be “All that said, Detective Bartlebe retired from the force having never found the killer. The End.”
They could have at least shown us this magical table and prove that it's incapable of producing impassable mazes.
Edit: Here is the table (page 12)
Or, like, a screenshot of the game?
That would be nice.
It's funny because I've often referred to doing software archeology, but on a slightly more recent time scale. I'm currently working on a 20-year old codebase. It's been updated continuously during that time, but its history is preserved in source control. I have gone back through the commit comments and older revisions to see how parts of it were written previously and to understand why some constructs exist. (I've also shown that some existing constructs are no longer necessary because the things they supported no longer exist!) It's really interesting to do! I don't usually have much time to do it, but when I do, it's quite satisfying.