A love affair with TF2 mapping leads to shameless replica of Hammer for Unity 3D

[Fruity Snacks]

Well, the mesh manipulation system seems similar to the UDK I've worked with. Probably thats it?


EDIT: Management was a poor word choice, manipulation is better.

 

[(s2a) yahodahan]

Ah, well, that's a good thing- mesh manipulation is just awful in Hammer, haha

 

[grazr]

This looks far superior to hammer. That's a little sad We can only hope Hammer 2.0 will be as good as this

I think this goes without saying since Hammer is essentially 8 years old.

I'm hoping that although Hammer has had its minor upgrades with updates to the engine with games like L4D2, that it gets a severe overhaul for HL:ep3.

As, if and when it is ever released, of course.

 

[WastedMeerkat]

Oh my god, thank you! I just installed and started looking into Unity last week, and now this thread title catches my eye! I'm definitely going to be keeping tabs on this project.

 

[(s2a) yahodahan]

Is it ok to necro my own thread? It's been ... awhile XD I'm resurecting things, and was reminded of this post. So glad I still exists. Still miss the TF2 life.

So, the story has gone:

• ProBuilder did indeed get better
• Unity bought it, and hired me
• Unity laid me off, after 8yrs (not a bad run, really)

In case anyone is interested, myself and the co-founder are starting again, this time with a strong focus on the level design. Maybe see you over there (www.overdrivetoolset.com), hope everyone here is doing well! Messy times out there. Thanks for all the support as we started

 

[Sonoma]

Is it ok to necro my own thread? It's been ... awhile XD I'm resurecting things, and was reminded of this post. So glad I still exists. Still miss the TF2 life.

So, the story has gone:

• ProBuilder did indeed get better
• Unity bought it, and hired me
• Unity laid me off, after 8yrs (not a bad run, really)

In case anyone is interested, myself and the co-founder are starting again, this time with a strong focus on the level design. Maybe see you over there (www.overdrivetoolset.com), hope everyone here is doing well! Messy times out there. Thanks for all the support as we started

As someone working in Unreal, I understand the pain of not having good level design tools in modern game engines, BSP much better for block outs and y'know, level designin' ! Hammer++ has been a game changer in terms of getting some Quality-of-Life updates to the 20-year-old tools, even then it's crazy how so behind-the-times-yet-ahead-of-the-times Hammer is. It's just an extremely solid level design tool, something which Unreal lacks, what little Unreal has to offer in terms of in engine support for Level Design tools are not sufficient enough, I yearn for Hammer and other BSP tool's simplicity and efficiency. I also have a theory that the shift away from modern engines supporting Level Design tools in their engines has led to the decline of Level Design as a whole in the industry. Thankfully Source 2 should be the blueprint for how other modern game engines should support level design tools. For Unreal thankfully there is a community tool called Scythe which aims to bring Source 2's LD tools to UE

 

[(s2a) yahodahan]

Much agree. So many newer games (and creators?) don't have/know the tools, so they lean into procedural, or kit-bashing. The result is pretty mediocre, in my opinion. Good level design is an absolute art form, and requires specific tools (and artists).

 

[nickybakes]

I've used Probuilder for years, both in personal projects and professional ones at my old job. I always enjoyed how it felt similar to Hammer, yet I never knew it started out as a recreation of it! Thank you for your work on this, and I'm very sorry to hear about the layoff.

At my old job, I actually customized my editor layout and color settings to be close to Hammer, complete with a top-down 2D view

I really enjoyed using the vertex color options to make my blockouts really colorful haha

Your tool brought me much joy and productivity. I wish you and the co-founder the best of luck on your next endeavor!

 

[(s2a) yahodahan]

Hey @nickybakes ! Thanks so much. And wow, super cool work there! Vertex coloring is heavily underrated, such a great use. We'll definitely make it much easier to use and more prominent in the new toolset. Speaking of, any chance you'd like to preview some of that? I have a few helper tools ready to go, would love to get them tested in real games

 

[Tiftid]

As someone working in Unreal, I understand the pain of not having good level design tools in modern game engines, BSP much better for block outs and y'know, level designin' ! Hammer++ has been a game changer in terms of getting some Quality-of-Life updates to the 20-year-old tools, even then it's crazy how so behind-the-times-yet-ahead-of-the-times Hammer is. It's just an extremely solid level design tool, something which Unreal lacks, what little Unreal has to offer in terms of in engine support for Level Design tools are not sufficient enough, I yearn for Hammer and other BSP tool's simplicity and efficiency. I also have a theory that the shift away from modern engines supporting Level Design tools in their engines has led to the decline of Level Design as a whole in the industry. Thankfully Source 2 should be the blueprint for how other modern game engines should support level design tools. For Unreal thankfully there is a community tool called Scythe which aims to bring Source 2's LD tools to UE

The decision to leave BSP behind is also a heavy contributor to the rather poor performance these days which no one (including me) can seem to stop talking about.
Considering "raytracing" is the big topic of discussion among graphics programmers, I find it rather baffling that people have seemingly chosen to ignore that a BSP is literally the ideal ray-tracing acceleration structure.

Of course, it's not like there was no reason for this. After a certain point of cramming more detail into your BSP, the performance benefits of it start to become performance detriments due to the high iteration costs, and as we all know certain types of maps just aren't a good fit for BSP-based optimisation.
When the "solution" is to cram a bunch of extra-BSP detail into your map and use inferior optimisation techniques, you can start to understand that we're losing the plot, especially since these studio models are a MASSIVE reason for TF2's poor performance.

My theoretical ideal graphics engine would be a ray-tracer with a BSP that has the underlying geometry at a 1990s level of detail, but has the rays lie about what they hit by blending between the positions and normals at the edge between two BSP faces, and based on the heightmap of the texture they hit.
That way, you get next-generation, mathematically perfect visuals with what should be a hyper-fast renderer, kind of like procedural raymarching but far superior. It also takes minimal effort to build, which is an important part of why no one likes working with BSPs anymore - instead of just having a level designer who applies high-quality textures to your walls, you need a level designer and a team of 3D modellers.

This leads me on an unexpectedly lengthy tangent.
Everyone seems to want tools that lead to a higher-quality overall result, which is great. But NO ONE seems to be acknowledging that the way you get there is actually by taking the existing graphical quality, and reducing the cost and time it takes to produce it.
There's a thread quoting a 2005 talk by John Carmack on some graphics programming forum where he compares the amount of budget and time it takes to make a video game to a movie, and says it's only going to become more so in future - which turned out to be true.
This biases the industry towards the biggest, most well-established game development companies, who could easily fund a movie and in fact frequently do.
What were the most prosperous times in game development history, where all your favourite games came from? The pre-2000s, a.k.a. the era where games cost the least time and money to develop, and the 2000s, the age of Flash where a single person could make a web-game in a few months.

Mentioning John Carmack's name is also significant since he was the guy that decided to use BSP-based geometry in video games in the first place, and the result of that was that DOOM was a mind-blowingly amazing-looking game for 1993, because as it turns out, significant graphical advances are always directly preceded by landmark advances in graphical optimization.
The faster you can get existing graphics running, the more you can push them to their limit and make them look even prettier.

Unfortunately, I'm not NEARLY smart enough or rich enough to make all of these things that I want happen.

 

[nickybakes]

Hey @nickybakes ! Thanks so much. And wow, super cool work there! Vertex coloring is heavily underrated, such a great use. We'll definitely make it much easier to use and more prominent in the new toolset. Speaking of, any chance you'd like to preview some of that? I have a few helper tools ready to go, would love to get them tested in real games

Thanks, and that all sounds cool! I'd be down for doing some testing on my personal game projects! My Discord is just nickybakes if you want to reach me there.