https://www.youtube.com/watch?v=GGTOYDIwCFo

A lot of people mention that the stock pipes are "less consistent" because they tumble through the air. This video and its description have a very detailed explanation of why this is not the case, the Iron bomber and stock pipes have the exact same projectile arc. The only difference is the fuse time and rollers.

Basically there are 2 things that make pipe launchers have "inconsistent" arcs.

1. Pipes are given a random amount of velocity when shot (within a range)
2. Pipes are given a random amount of spin when shot.

These are the same on all pipe launchers (except for the loch n load which doesn't spin) and if all of them had the same stats, they would all be exactly the same if it wasn't for VPhysics models.
The VPhysics model is what people talk about when they say the Grenade Launcher is "more inconsistent" with respect to its air arc. The source engine simulates drag, and so projectiles with a larger surface area to their model will be slowed down more in the air and travel slower. This happens with loose cannon projectiles which get slowed down more because they're big, so they go slower than they're supposed to.

In theory, the Iron Bomber's arc should be different to the Grenade Launcher's arc even when the random spin and velocity are disabled, because the projectiles are different sizes and shapes. However, this video shows that the Iron Bomber and Grenade Launcher actually have projectiles that fall in the exact same spot every time when these things are disabled. This is because the physics model for the iron bomber is actually a cylinder the same as the stock pipes. So iron bomber pills tumble in the air and experience the same orientation dependent drag that pipes do.

The sigsegv video that a lot of people reference when talking about this actually shows stock pipes and iron bomber landing at very similar positions.

Edit: It seems that the Grenade Launcher and Iron Bomber behave differently with respect to how their models interact with their collision hulls, effectively giving the Iron bomber a bigger "hitbox" see: https://www.teamfortress.tv/post/1008432/iron-bomber-regular-pipes-have-the-same-physics

Does the seed always treat the second bullet fired so that it goes at the same angle or is it based on something else?

If you want to hit an "average amount of spread" you can probably calculate what the average concentration would be with a random seed and choose one which is close to that average for each weapon.

Also, just out of interest, is spread just calculated using an azimuth and angle away from crosshair? If so is the angle away from the crosshair drawn from a uniform distribution or do they implement it some other way?

In the second example the red guy has the disadvantage because rockets come out of the right hand side of the rocket launcher.

https://i.imgur.com/etvHLMB.jpg

This my tf2 team,. just me n the boyz.. chillin.

https://www.youtube.com/watch?v=4dcQO6Zb8Eg

gun

https://i.imgur.com/CA3zsGv.jpg

One for all you boomers out there

It's the visual, rather than the networking effect that people are using for airpogo. It's not that the stickies are actually spawning and detonating earlier, it's that you're able to see them doing so which means it's easier to time them, even with the tradeoff of potentially worse networking. I can see how this would be annoying if you're used to it, but tbh I didn't even know it was a thing, I can't always see the projectiles I'm firing at walking speed lmao.

aye she's turned the weans against us

https://pbs.twimg.com/media/EWWmQqPU8AAlOls?format=jpg&name=120x120

https://i.imgur.com/1VrIsPM.jpg

I can't believe Viewmodels are Fucking Dead.

There are multiple small qol things that might be good to change e.g. melee interacting with teammate hitboxes, but it's not clear whether many these things are bugs or intended features especially considering how long so many of them have been in the game. Am I right in assuming you want this to only be things that are certainly bug fixes/performance improvements and no changes to the default game?

In any case, a few things I can think of which I think are definitely bugs and haven't been changed for years as far as I can tell are:

When skipping ticks in demos, snd_soundmixer sometimes gets set to "FreezeCam_Only" instead of "Default_Mix" leading to a bug where the game appears to be muted.

The thing where spies have certain cosmetic body groups (most notably pyro heads and demoman feet) just disappear when disguised (this might have already been fixed now, not sure, but probably not in the version of the source code you're using).

Bug where certain projectiles collide with func_lod bounding boxes for objects marked as solid as detailed in this video.

It might be interesting to see if you can fix physics issues associated with running servers at higher tick rates (pipes bouncing oddly, sentry gun turn rate) but that might not be a priority as its a non standard configuration.

https://i.imgur.com/aFn3qBP.jpg

https://pbs.twimg.com/media/Ee_XwEDXoBQVxJl?format=jpg&name=small