How do you want hammer integrated with scenes?

[DrakeFruit]

So far the options I've got are:

1. Leave hammer alone and just place prefabs in hammer instead of entities
2. Integrate hammer in-scene, sort of like probuilder in Unity
3. move the hierarchy, inspector, and scene view into hammer and leave the editor for the game view and project management

[SamFadi-dev]

2

[matekdev]

Option #2, have an in-scene replacement for Hammer. Everything would be done in C# as well which would allow for faster iteration in the future.

[handsomematt]

The plan is already 1 first and then 2.

[DrakeFruit]

meaning 1 as an intermediary to 2 or both at the same time?

[handsomematt]

intermediary, it's better to make hammer work with scene first before we try to match it's extensive feature set

that's not to say there's an immediate plan to replace hammer, we'll just add and do features that make sense as we explore the scene system

[anthonysharpy]

Integrating Hammer with the scene editor... yes that could be good. But IMO the idea of getting rid of Hammer is dumb:

• If all the screens/buttons/controls that are in Hammer were ported to the scene editor, there wouldn't be much space left on the screen to actually do anything. Sure, you can strip out features and/or hide stuff away, but the result is inevitably a worse piece of software.
• There is no real benefit to doing it. Maybe there's a benefit to adding some of Hammer's features (mesh editing, lightmapping etc) into the scene editor, but that's not the same thing as getting rid of Hammer. In fact, why do we need these things in the scene editor at all? When I double-click a mesh, why can't that just open it up in Hammer for editing, for example?
• Hammer isn't perfect, but it's a complicated piece of software with a lot of features. I don't see Facepunch really having the resources to create a polished replacement in the short or medium term. And even if they did, what's the compelling reason to be reinventing the wheel when Hammer already does a semi-decent job?
• If it makes sense to merge Hammer into the scene editor, why wouldn't it make sense to merge the particle editor into the editor, or the material editor into the editor? The reason is because they're different pieces of software that specialise in different things, just like Hammer.
• A lot of people like Hammer, and it has a long history. Removing it would be unpopular with a lot of people, and I would venture a guess that it's not something that anyone who actually uses the software really wants to see.
• If we are going to do everything "the way that Unity does it", then what really is the appeal of s&box? Why wouldn't someone just use Unity? Sure, s&box can and does do some things better. But at the end of the day Unity is a much bigger company, so it doesn't make sense to compete over the same (or similar) feature set. Also, what was the point of basing s&box on top of the Source 2 engine if everything is just going to be stripped-out anyway?
• If Hammer goes then compatibility with Source maps goes out the window with it. Maps made for games like Counter Strike could no longer be ported to s&box, and vice-versa, which will shrink and shut s&box off from the rest of the Source community.
• There are literally a hundred more productive things to be focusing on like networking, matchmaking, physics, fixing bugs, optimisation, creating assets for people to use, QOL features, improving asset.party etc etc.

Just my opinion as always 👍

[DrakeFruit]

1. No features should be stripped out, the UI just needs a bit of retouching (do we really need all the buttons in the top bar?)
2. Geometry is something that needs to be edited often and compared to other parts of the scene. we're not reinventing the wheel here, just attaching it to a new vehicle.
3. That's subjective and not really relevant.
4. You edit one particle / material / etc. at a time, so putting it in a separate window you can have on your other monitor makes sense. Map geometry is much bigger than that and doesn't make sense to have in a separate window when you need the reference of the rest of the scene.
5. Merging hammer into the scene view wouldn't take anything away from it's functionality or design. Done right it should be as noticeable of a change as the new entity menu / asset browser. And s&box isn't meant for people with experience in the source engine, it's made for game developers and content creators regardless of background.
6. It's not the goal to do everything the way Unity does it. The goal is to take the best parts of every engine as inspiration. Probuilder has a terrible design and really isn't fun to use, but hammer definitely is and I think Facepunch will do a good job at merging the strengths of the two.
7. Hammer isn't going anywhere? it will still be hammer, just integrated into the UI.
8. Hammer integration isn't that high of a priority. Short / medium term is just to get prefabs in the current hammer.

[CarsonKompon]

As someone who's actually been making games with s&box (whether i use hammer to model or for maps or both) I love hammer, but I'm very eager to see it leave.

Hammer is a great toolset, but there's no reason why Facepunch can't make a similarly feature rich toolset that works best for the s&box workflow. Hammer was made for FPS games, so it caters towards that. We don't need a lot of its now ancient features such as entities, baked lighting, or anything else that isn't required for s&box.

I'm also sure it's much easier for them (and favourable) to create new tools rather than work with the spaghetti code that makes up Hammer. The end result is going to be an improved modelling/mapping tool that has near feature-parity with Hammer and is built in whatever way Facepunch finds works best for the new workflow (could take some iteration like other systems in the past, but that's all part of the process).

That's also not to mention that once Facepunch builds their own toolset, they can make sure users are able to extend it in whatever way they want. Meaning you'll have more control as a developer than you ever had over Hammer (custom tools, panels, views, ect). This is a future I'm looking forward to that Hammer locks me away from as it stands.

[MrBrax]

baked lighting is crucial for performance, having hundreds of dynamic lights on the map looks awful and performs awful

[handsomematt]

wrong

[MrBrax]

i get like 20 fps on my map with all them lights going on so i disagree, but on the other hand i don't wanna spend 700€ on a new mid tier gpu to upgrade from my 1070

[CarsonKompon]

The performance gains from scene system are already so large that once vis is working again I'm pretty sure that will be a massive lag reduction for all maps. Baked lighting doesn't necessarily mean more frames since most maps use both baked and dynamic lighting together

[MrBrax]

what about light bouncing and global illumination? it's all the craze today with ray tracing but not everyone has that