Hacker News Reader

Comment by FartyMcFarter | Hacker News Reader

I love Carmack's general message here, but this part is not something a lot of developers have the luxury to do:

> "Most of game development is a panicky rush to make things stop falling apart long enough to ship, so it can be hard to dedicate time to fundamental software engineering, but there is a satisfaction to it, and it can pay off with less problematic late stage development."

When you're at id Software and don't have to answer to anyone with regards to deadlines or funding, it's fine to think like this. Unfortunately a lot of developers have a lot more constraints than Carmack did when he made his best games.

View on HN

I took it as an appeal to management to consider the trade offs they're forcing.

That might be the case, but in games development you often have a company (studio) answering to another (publisher), and it's the publisher's management that decides the deadlines and funding. The people managing the engineers would have no control over those.

Carmack's company didn't have this problem when they made the games they eventually open-sourced. When they did get bought by a publisher, they no longer open sourced their games as far as I know, so who knows what the engineering practices were like at that point.

Games are this interesting medium of art where they're really not easily replicated, but very easy duplicated. There's practically no equivalent to a cover for a song that exists for a game. You can distribute a ROM, maybe swap some assets, but the code only exists in the binary it was compiled to. It was only up until recently (with decompilations) that a game was effectively set in stone unless it was open-sourced.

I don't know. I think it's silly for decades-old-games to run into licensing issues when trying to go open source. The tech is effectively obsolete except in the instance of that specific game.

Look up “video game demake”.

I don’t really understand what distinction you’re making between “replicating” and “duplicating”

Duplicating = copy/paste digital binaries, replicating = reimplementing the codebase.

There have been engine reimplementations without decompilation or source releases - OpenMW being perhaps the most complex example.

I'm also not sure why you think remakes are that different from cover bands - neither are a 100% identical copy.

And let’s not pretend that Doom and Quake didn’t have several point releases after the fact. And that the team at id didn’t pull all-nighters to ship the games.

Those games are far simpler (by at least an order of magnitude, maybe two or three) than most of the modern AAA titles.

What was the last AAA game Carmack shipped? Rage? Hardly a smashing success, even if it was technically impressive.

I have a huge amount of respect for Carmack’s skills and work ethic and track record, but I wonder if he really has a good handle on the sheer amount of work that goes into a modern game. With often dozens of programmers, and hundreds or thousands of contributors, the coordination overhead is just massive. I think that’s why overall quality control is harder to maintain.

> When you're at id Software and don't have to answer to anyone with regards to deadlines or funding

actually Id Software people worked in much harder conditions than game devs of today.

https://preview.redd.it/iehd38ucuuz71.jpg?width=450&format=p...

Didn't id software lose a lot of relevancy in the early 2000s? They stopped producing new, interesting IP, and their engines lost popularity.

I think he programmed Wonfenstein 3D alone in about 4 months, it's not like he took his time

Reply on news.ycombinator.com