How accurate is GZDoom to vanilla doom?

Discuss anything ZDoom-related that doesn't fall into one of the other categories.
Posts: 67
Joined: Fri Jan 14, 2022 8:31 pm

How accurate is GZDoom to vanilla doom?

Post by Casualfan »

I've heard a lot of people avoid playing vanilla doom on GZDoom because it has some inaccuracies to the original. Personally I don't care if it isn't 100% the 1993 experience (I use mouse look like a heathen 😁), but im curious: From purely a gameplay perspective, how accurate or inaccurate is GZDoom from the original. Like, if I wasn't intimately familiar with all the different ports, would I notice any of the differences.

Just want to be clear, I'm not here saying "GZDoom sux!" or "GZDoom is the shiz!" I'm just curious. Personally GZDoom is my go-to, and I'm just interested in what other ports have to offer :)
User avatar
User Accounts Assistant
Posts: 6002
Joined: Thu Feb 26, 2004 3:02 pm
Preferred Pronouns: He/Him

Re: How accurate is GZDoom to vanilla doom?

Post by Caligari87 »

Most of the changes are too numerous to list. Physics, rendering, input, RNG... almost all of it has changed in some way. Some might be noticeable, but lots of it isn't readily apparent.

Taking RNG for an example. Doom and GZDoom are both deterministic, meaning that given the same "seed" (starting conditions) and actions, a game of Doom will play out exactly identical every time, because the exact same random numbers are generated. This is how demos work. But since the RNG engine in GZDoom has changed, it outputs different random numbers than the RNG in classic Doom. Superficially this won't be immediately apparent because the main source of "randomness" between games is the player's actions (since players aren't frame-perfect machines). Monsters will appear to move in similar patterns in GZDoom, but some players might notice the SSG feels very very slightly more powerful, since due to problems in Doom's original RNG the maximum damage per pellet is never rolled.

Most non-visual changes are like that. Very subtle, hard to notice unless you know what you're looking for or make a direct 1-1 comparison, but might contribute to very slight almost imperceptible differences in game-feel.

Posts: 17754
Joined: Fri Jul 06, 2007 3:22 pm

Re: How accurate is GZDoom to vanilla doom?

Post by Gez »

The differences are mostly visible on maps specially contrived to make them blatant. For example, there was a map that featured a long lava corridor. In vanilla Doom there's a glitch in the movement code where sometimes you can get a big boost to your movement speed if you are running along blocking things or a wall. (Those are called "thingrunning" and "wallrunning" in Doom jargon.) So in this map, the trick is to exploit wallrunning to go supersonic and cross the corridor before you melt. In GZDoom, the glitches in the movement code were fixed (not specifically to remove wallrunning, but to solve other problems), but as a consequence the wallrunning and thingrunning exploit were removed and you cannot cross that corridor without cheating.
User avatar
Graf Zahl
Lead GZDoom+Raze Developer
Lead GZDoom+Raze Developer
Posts: 48598
Joined: Sat Jul 19, 2003 10:19 am
Location: Germany

Re: How accurate is GZDoom to vanilla doom?

Post by Graf Zahl »

There's also a common misconception of thinking that ports that are demo compatible are also vanilla-accurate for actual playback. This really isn't the case. Most of those demo compatible ports have multiple code paths where the vanilla-faithful one is only activated when explicitly requested or for vanilla-compatible demo recording.

As an example, Eternity Engine uses a lot of ZDoom code and its own additions for playing advanced maps because the original code wouldn't be capable of handling portals or other 3D constructs the engine features. Boom also changed the RNG and only uses the vanilla one when playing back demos.

Return to “General”