Half-Life 3 Leak/Speculation By Tyler McVicker

Thats completely unrealistic - Epic has 100+ devs working on just the engine.
Even companies like CD Project Red who have a custom engine are switching to UE5 now because there is such a huge cost in terms of resources and manpower needed to handle your own tech.
To be honest, it's not even about the manpower; a lot of these companies have more than enough bodies to throw at the problem.

The issue is that it can't really be done in parallel with game development anymore. It's like trying to lay the tracks in front of the train while it's moving. You just can't make a game until the engine is feature complete. And having to put your whole project on ice for years while you develop an engine is not really a viable option either.

This is what messed up CDPR with Cyberpunk and also what detailed every Valve game from the time they moved off of Source until Alyx came out. Including the earlier attempts at HL3. The reason Alyx managed to succeed was because Source 2 was finally in a more finished state.

It's still totally possible to make your own engine but it involves planning ahead, working on it for the NEXT game.
 
Last edited:
To be honest, it's not even about the manpower; a lot of these companies have more than enough bodies to throw at the problem.
Well by manpower I meant people who have the necessary skills to work on a specific part of the engine.
Its not like you can just throw random programmers at a problem, you need people with experience in a specific subset.
Someone experienced in animation programming might be completely useless for the audio part etc.
Those experts usually already work for the big players - Valve has no such people, as they didnt need them.
The issue is that it can't really be done in parallel with game development anymore. It's like trying to lay the tracks in front of the train while it's moving. You just can't make a game until the engine is feature complete. And having to put your whole project on ice for years while you develop an engine is not really a viable option either.
Yep, that is correct. It is still feaseable for smaller projects, but for AAA it has become way too risky.
This is what messed up CDPR with Cyberpunk and also what detailed every Valve game from the time they moved off of Source until Alyx came out. Including the earlier attempts at HL3. The reason Alyx managed to succeed was because Source 2 was finally in a more finished state.
It's still totally possible to make your own engine but it involves planning ahead, working on it for the NEXT game.
I think the only reason Valve made Alyx a VR game is because that way they could get away with using their heavily outdated Source 2 tech.
For VR you dont need AAA visuals or levelstreaming, something Source 2 cant do.
But I dont think Valve has any interest in competing anyway, they will stick with this tech till the end.
Hence why they did Deadlock - no new tech needed.
 
Last edited:
As long as were dreaming.
During the Nvidia presentation in January, Gabe will walk out on stage and talk about a new partnership working with Nvidia to get full driver support for Steam OS.
And show a trailer for HL3 running on Steam OS and a 5090 with all the bells and whistles, exclusive to Steam.
I mean... this is all within the realm of possible innit?
A HL3 exclusive on Steam? Shocking! Using 50xx to show it off for PC users? Sign me up for a 5080! Proper Nvidia driver support for Linux? The masses are clamouring for it! Will Nvidia show stuff in January at CES? Absolutely! (Jensen's keynote speech is Jan 6th). Gaben walking out on stage? It happened before.

This is not some harebrained crazy talk, it's prophecy!

124861.webp
 
Half Life Alyx is fantastic. But what's not discussed so much......is HL1 and HL2 are equally amazing in VR with the free VR mod on Steam.

Point being...if HL3 is a real thing make it a hybrid title.

VR and Half Life are just too good together.

I don't disagree.

But it's not me you need to convince. It's Valve.
 
I agree you're not going to get a huge team working on single player campaign. But HL: Alyx was made by 80 people and Alan Wake 2 was made by 130 people. So it's still possible to produce a high quality single player experience with a smaller team, even if it's not full AAA. I think the ridiculous expectations around Half-Life 3 are a distraction from the fact that we still have reason to get excited about a new single player game made by the Half-Life: Alyx team.

If you're satisfied with a non-AAA HL3 that is much smaller in scope than HL1&2; being made by a smaller team, then I agree your expectations are much realistic than those expecting something of the same scope of the original games, but updated with the latest rendering tech.
 
Good lord, some of you lot are so sceptical.

You'll all be eating your words and have egg on your face when this game launches, just three weeks after Bloodborne PS5.

No need to eat your words when you're face is covered in delicious egg, just use that (hopefully very long) tongue.
 
I get it...I really do. I mean you see somebody on the internet saying they have inside information about Half-life 3, you get excited, next thing you know you're posting on GAF. It's alright, I don't blame you. Next time though, just remember to THINK.

T - Throw your keyboard out the window
H - Hold your breath
I - Incapacitate yourself (by holding your breath)
N - Nap
K - Know after waking up to not post stupid shit you can't corroborate.
 
Last edited:
 
Hard to believe it's going to happen, but then again several years ago I wouldn't expect a big project like HL Alyx to be released.

"Procedural generation alongside handcrafted experiences" still sounds awful though. For me stuff like this is always an instant turn-off, you can never beat handcrafted stuff, even if it means there's less content in the game.
 
Last edited:
Well by manpower I meant people who have the necessary skills to work on a specific part of the engine.
Its not like you can just throw random programmers at a problem, you need people with experience in a specific subset.
Someone experienced in animation programming might be completely useless for the audio part etc.
Those experts usually already work for the big players - Valve has no such people, as they didnt need them.

Yep, that is correct. It is still feaseable for smaller projects, but for AAA it has become way too risky.

I think the only reason Valve made Alyx a VR game is because that way they could get away with using their heavily outdated Source 2 tech.
For VR you dont need AAA visuals or levelstreaming, something Source 2 cant do.
But I dont think Valve has any interest in competing anyway, they will stick with this tech till the end.
Hence why they did Deadlock - no new tech needed.
Did we play the same game? HL:A looks absolutely stunning even 4 years later.
 
Did we play the same game? HL:A looks absolutely stunning even 4 years later.
Good looks has nothing to do with how well an engine is suited for modern development needs.
Quake RTX also looks amazing, but no one in their right mind would use that engine for a big game nowadays!
 
Last edited:
Honestly this is the most realistic smoke we've had on Half Life 3 ever. I'm internalizing it and beginning to make plans for launch day.
 


  • Valve is working on Deadlock, with updates including changes to ragdoll physics, localization, and shaders.
  • Valve partnered with a tech startup specializing in realistic natural environments, possibly for HL3.
  • An original Half-Life 2 developer rejoined Valve to work on an unannounced project, possibly HL3, after participating in the documentary.
  • Valve accidentally leaked internal SteamVR builds, revealing future ARM compatibility.
  • SteamVR updates hint at broader ARM ecosystem integration.
Deckard VR Headset

  • Next-generation all-in-one VR headset.
  • ARM-based CPU, similar to the Quest series.
  • Inside-out tracking, supporting both standalone and tethered gaming.
  • Functions as a "Steam Deck VR," playing Steam Library games on a virtual screen.
  • Potential 3D depth feature for flat games, similar to Nvidia's GeForce 3D.
Roy Controllers

  • Evolved design from knuckles to Quest-like controllers.
  • Different control schemes, including D-pad and face buttons.
  • Potential finger tracking via onboard cameras.
Steam Controller 2

  • Resembles a screenless Steam Deck with VR tracking.
  • HLX might use assets from Arty, which could be revived as a tech demo for Deckard.
  • Leaked SteamVR builds revealed ARM compatibility and confirmed hybrid gaming for Deckard.
Speculation from Tyler

  • HLX (possibly Half-Life 3 or a related project) may incorporate assets from Arty, which was ported to Source 2.
  • Arty could be revived as a tech demo for the Deckard.
 
I do hope Valve just shadowdrop the announcement and don't do a fucking ARG or some other prolonged period of teasing.
 
That would actually be kinda neat if the third game makes it way back to Black Mesa... or the ruins of Black Mesa.... Is Opposing Forces even cannon anymore? I don't think it is...

Pretty sure Black Mesa getting nuked at the end of Opposing Force is canon.
 
Pretty sure Black Mesa getting nuked at the end of Opposing Force is canon.

That part might actually be cannon, but I swear that Valve said in the past that Opposing Forces is not cannon overall, because of the Race X stuff that doesn't really fit in with the HL2 narrative.
 
I think the Half-Life style of single player FPS is basically dead at this point, replaced with open world games and military shooters. So if Half-Life 3 came out it wouldn't have any direct competitors.
Then, everyone will copy it.

I just wish more people used source engine
 
every game in the franchise has perfect VR mods now... so wait maybe 6 months after release and we'll have it.
It's more about the message this sends than the ability to play it in VR that I'm most concerned about: We need more big names doing VR (even if it's only an optional mode) to increase adoption.
 
Last edited:
I just wish more people used source engine
Why would any dev ever want to do that?
- It is heavily outdated tech lacking a shitton of features other modern engines have
- The editing tools/scripting is way worse than those of competitors
- Upgading Source to incorporate modern features is akin to writing a new engine
- Valve doesnt licence or support the engine and there is no direct console support

Outside of modding or people making retro stuff it makes zero sense to use Source1/2 for a project.
 
Why would any dev ever want to do that?
- It is heavily outdated tech lacking a shitton of features other modern engines have
- The editing tools/scripting is way worse than those of competitors
- Upgading Source to incorporate modern features is akin to writing a new engine
- Valve doesnt licence or support the engine and there is no direct console support

Outside of modding or people making retro stuff it makes zero sense to use Source1/2 for a project.
I will say that Source 2 has impressive scaling across devices... HL:Alyx can run on surprisingly low-end hardware for VR, and very seamlessly upgrades to looking gorgeous on higher GPUs. I rarely see a game that looks that amazing at the high end while still running smooth on GPUs from nearly a decade back.
 
I will say that Source 2 has impressive scaling across devices... HL:Alyx can run on surprisingly low-end hardware for VR, and very seamlessly upgrades to looking gorgeous on higher GPUs. I rarely see a game that looks that amazing at the high end while still running smooth on GPUs from nearly a decade back.
Of course it scales, its old tech. Upgrading materials etc can make it look very shiny.
But that doesnt change the fact that its completely useless for modern development.
 
I will say that Source 2 has impressive scaling across devices... HL:Alyx can run on surprisingly low-end hardware for VR, and very seamlessly upgrades to looking gorgeous on higher GPUs. I rarely see a game that looks that amazing at the high end while still running smooth on GPUs from nearly a decade back.
I fully agree. Played it on my 4080 super + PSVR2

Stunning
 
But that doesnt change the fact that its completely useless for modern development.
I mean... I don't really have a strong opinion about engines to be clear, but: a game like Alyx is still easily one of the top looking & running games on the entire market today 5 years later, so (putting aside all the licensing barriers etc that you mentioned, of course) if that doesn't count as "modern development" then the term has little meaning to me.
 
I mean... I don't really have a strong opinion about engines to be clear, but: a game like Alyx is still easily one of the top looking & running games on the entire market today 5 years later
I played through Alyx and it still looks good thanks to its art direction, but all the old tech is clearly visible in it and is nowhere near any top looking game today.
Source Engine is based on the Quake engine - so the best comparsion nowadays is idtech - and if you compare Indiana Jones with HLA, then it looks like more than a generation behind.
Good art direction can hide tech deficites up to a certain point.

, so (putting aside all the licensing barriers etc that you mentioned, of course) if that doesn't count as "modern development" then the term has little meaning to me.
Well if you arent a developer then of course this has no meaning for you.
But I am a developer for over 22 years and I worked with various engines.
I just tried to make you understand why no one is using it!
 
If this is true that HL3 game will launch this year, this could be the biggest shadow drop in history of gaming ever.
 
Then, everyone will copy it.

I just wish more people used source engine
I don't think there were many games that tried to do what Half-Life 2 was doing, despite its success. Even Half-Life 1 didn't really have any "clones". There was Red Faction, System Shock 2, Deus Ex, but they were all pretty different.
 
Knowing Valve, they'll make it SteamOS exclusive and launch the PC distro like a week prior to get people ready, then Linux gaming will be standard in like 8 years or so
 
Top Bottom