Half-Life 3 is in the Final Phase of Development – Rumour

I can't imagine it will have any ground making tech like 1 or 2 and will be a lame corridor shooter.
Funny thing is, if Half Life 2 never came out and dropped tomorrow with modern poly counts and texture res, people would be drooling over the "now THIS is NEXT GEN" physics and reflections.
 
Last edited:
Honestly, I just don't care anymore.

It's been 20 years since HL2. I'm over it.

If it had been 5-10 years afer the release of HL2 then I'd be all over HL3, but I just don't care about Half-Life anymore.
 
If it comes, i wonder what kind of visuals it'll have. Source 2 visuals are pretty outdated compared to what's available elsewhere, but if they take the ultra realistic destruction route with some amazing weather effects that affect gameplay, everyone will forget it not looking as good as other titles.
 
I think if does come out it will disappoint people and not because the game is gonna be bad or anything, mostly because after these long years people will have very unrealistic expectations for it which will never meet.
 
Last edited:
Reality check:
- Source 2 is an extremly outdated engine that is 1-2 decades behind the competiton. Even HL Alyx still didnt have level streaming, like wtf! UE3 & others had that in 2004!
- In order to make a HL3 that meets the crazy high expectations of modern players, they would either have to use a modern engine, or make a new one.
Using a competitor´s engine is unlikely and making a new one on the same level as Anvil, UE5, Frostbite etc, requires a very large amount of engine coders.
On top of that, you need another large team to develop a Triple A title with the new engine.
- Valve does not currently have the staff numbers or expertise to pull off anything like that, and we havent seen them going on some sudden hiring spree of experienced devs either.
- Half Life is an IP only old people like us care about. Releasing some AAA HL3 to a very limited audience without having remade HL1 and HL2 first will lead to disaster.

Conclusion:
If we ever see a HL3 - which I doubt - it will most likely be done in an upgraded Source 2 variant aka Alyx and not try to compete with anything that is on the market.
How is source 2 1-2 decades behind the competition? Wasn't the first game released on it half life Alyx? And that looks great even now
 
Half-Life Alyx was absolutely terrific, so we know Valve can still walk the walk, but at this stage I'll believe it when I see it, frankly.
 
the question is what the standout design element will be.
they can't just make another Half Life 2, and I bet they also wouldn't want to.

so they probably have some innovative concept attached to it... but what could it be?
maybe the first game with a truly intelligent AI based on a machine learning algorithm? maybe some crazy new physics based mechanics like controlling gravity... as an evolution of the gravity gun. or maybe they'll just make it a Portal crossover
 
Reality check:
- Source 2 is an extremly outdated engine that is 1-2 decades behind the competiton. Even HL Alyx still didnt have level streaming, like wtf! UE3 & others had that in 2004!
- In order to make a HL3 that meets the crazy high expectations of modern players, they would either have to use a modern engine, or make a new one.
Using a competitor´s engine is unlikely and making a new one on the same level as Anvil, UE5, Frostbite etc, requires a very large amount of engine coders.
On top of that, you need another large team to develop a Triple A title with the new engine.
- Valve does not currently have the staff numbers or expertise to pull off anything like that, and we havent seen them going on some sudden hiring spree of experienced devs either.
- Half Life is an IP only old people like us care about. Releasing some AAA HL3 to a very limited audience without having remade HL1 and HL2 first will lead to disaster.

Conclusion:
If we ever see a HL3 - which I doubt - it will most likely be done in an upgraded Source 2 variant aka Alyx and not try to compete with anything that is on the market.
I think Valve should be capable of implementing a level streaming system, given that they have some of the best talent in the industry. And I don't think we're in a world where games need to look better than HL:A say, to do well. Let's not forget that Episode 2 and Portal 2 were good looking but hardly cutting edge when they released.
 
Last edited:
Reality check:
- Source 2 is an extremly outdated engine that is 1-2 decades behind the competiton. Even HL Alyx still didnt have level streaming, like wtf! UE3 & others had that in 2004!
- In order to make a HL3 that meets the crazy high expectations of modern players, they would either have to use a modern engine, or make a new one.
Using a competitor´s engine is unlikely and making a new one on the same level as Anvil, UE5, Frostbite etc, requires a very large amount of engine coders.
On top of that, you need another large team to develop a Triple A title with the new engine.
- Valve does not currently have the staff numbers or expertise to pull off anything like that, and we havent seen them going on some sudden hiring spree of experienced devs either.
- Half Life is an IP only old people like us care about. Releasing some AAA HL3 to a very limited audience without having remade HL1 and HL2 first will lead to disaster.

Conclusion:
If we ever see a HL3 - which I doubt - it will most likely be done in an upgraded Source 2 variant aka Alyx and not try to compete with anything that is on the market.

Yeah on a tech level hl 3 is gonna be aa+ level at the very best.

Valve don't have the people to do aaa releases anymore.
 
my man tyler mcvicker
 
Gaben's Yacht already let us down
Love seeing this ad in the body of and between every fucking post

9g7QL6Se6mungvSf.jpg
I keep getting that annoying gif too. Strange thing is that normally I could click a small x ir 3 dots and say I'm not interested in this ad but I can't with this one. Must be a different ad provider or something.
 
Last edited:
Reality check:
- Source 2 is an extremly outdated engine that is 1-2 decades behind the competiton. Even HL Alyx still didnt have level streaming, like wtf! UE3 & others had that in 2004!
- In order to make a HL3 that meets the crazy high expectations of modern players, they would either have to use a modern engine, or make a new one.
Using a competitor´s engine is unlikely and making a new one on the same level as Anvil, UE5, Frostbite etc, requires a very large amount of engine coders.
On top of that, you need another large team to develop a Triple A title with the new engine.
- Valve does not currently have the staff numbers or expertise to pull off anything like that, and we havent seen them going on some sudden hiring spree of experienced devs either.
- Half Life is an IP only old people like us care about. Releasing some AAA HL3 to a very limited audience without having remade HL1 and HL2 first will lead to disaster.

Conclusion:
If we ever see a HL3 - which I doubt - it will most likely be done in an upgraded Source 2 variant aka Alyx and not try to compete with anything that is on the market.
You people don't seem to realize that not everything that good for dev will leading to a better graphics, it's all about PERCEPTION, if the game looks good to the eyes of the average consumer it doesn't matter if the engine has "level streaming" or some other feature that's deemed essential to today mainstream

And alyx looks amazing
 
Last edited:
That's the essential issue with this industry:

"This engine is bad because she doesn't have the X feature that will make my life better therefore she'll not be able to produce high quality graphics"

After they launch a extremely good looking game

"Well the game looks good but they don't have the X feature..."


NOBODY CARE, the people who pay the devs wage, the consumer, only care about how the game looks not if they use X or Y feature or techniques
 
As someone who's been gaming since the NES I do wonder how much people even care at this point.

Obviously it would be big news when announced and a lot of people would be very excited - including myself. But if you were around 10 or 15 years ago I feel like the announcement would have had a significantly bigger impact than as opposed to now.

I think this is partly because the people that grew up with half life, many of them just don't game anymore.

If you put this question into Chan GPT and ask it for statistics it gives you some pretty interesting information. I just did a relatively simple query on the impact of a Half-Life 3 announcement in 2025 versus 2015. It concluded that the impact would be big but nowhere near as big as 10 years ago.

I think this speaks volumes:

Generational Shift: Many men in the HL2 demographic are now 40–55. Many are still gaming, but fewer do so daily or follow gaming news religiously.
 
Half Life 3 could be one line of code away from being finished. Doesn't mean Valve is happy with it and wont scrap it and start over again. I have a feeling this is something that happens a lot at Valve. Take something an inch to the finish line and do not cross it, simply because they can.
 
How is source 2 1-2 decades behind the competition? Wasn't the first game released on it half life Alyx? And that looks great even now
Because engine technology moved on? There have been big technological shifts how engines work and get made in the last decade.
Source 2 is an upgraded Source engine which itself stems from GoldSrc which is based on the Quake 1 engine.
The first Source 2 game was DOTA2. Alyx looks amazing despite the engine it is running on thanks to strong art direction.

I think Valve should be capable of implementing a level streaming system, given that they have some of the best talent in the industry. And I don't think we're in a world where games need to look better than HL:A say, to do well. Let's not forget that Episode 2 and Portal 2 were good looking but hardly cutting edge when they released.
And yet HLA didnt have it, so what did the "best talent in the industry" do in those 17 years between HL2 and HLA? Other companies wrote 2-3 completely new engines from scratch in that time!
The reality is most of that talent left a long time ago when they realized Valve doesnt really do games anymore.

You people don't seem to realize that not everything that good for dev will leading to a better graphics, it's all about PERCEPTION, if the game looks good to the eyes of the average consumer it doesn't matter if the engine has "level streaming" or some other feature that's deemed essential to today mainstream. And alyx looks amazing
How well games look that are done with a specific engine doesnt matter anymore to devs nowadays.
What matters to them is how good the tools are that allow you to make the game as fast and efficent as possible.
Most of the dev teams outside of Valve themselves that used Source1/2 in the past have switched to Unity, Unreal or something else for exactly that reason.
Sure, you can cut the grass in your garden with scissors, but why do that when there is mowing robots that do it 100 times faster.
 
And yet HLA didnt have it, so what did the "best talent in the industry" do in those 17 years between HL2 and HLA? Other companies wrote 2-3 completely new engines from scratch in that time!
The reality is most of that talent left a long time ago when they realized Valve doesnt really do games anymore.
They were stuck in development hell on continually cancelled projects. Valve's issue is not with the individual employees but with a flat management structure and an endless source of money (Steam) that means they never have to actually ship anything. But we're talking about the hypothetical scenario where HL:X does get shipped, because like HL:A, the project managed to build up enough momentum internally to avoid being cancelled. (Valve actually re-wrote HL:A's story from scratch like a year before release).

Looking at the experienced engineers who still seem to be working at Valve, I count Yahn Bernier, John McDonald, Jay Stelly, Dave Kircher and Brian Jacobson, who are 20 year+ veterans – plus a bunch of more generalist programmers who joined in the late 2000s. The way Valve works is each game will have a specific set of technical challenges that they will be focusing on, and they will update the engine to meet those challenges. The focus of HL:A was on confined spaces, due to how VR limits traversal, so the lack of level streaming would not have been a major limiting factor. If HL:X is a lot more open, then we would expect level streaming to be more of a focus. Being an AA sized developer means taking on a limited set of challenges at a time. It doesn't mean that no technical challenges can be accomplished!

Just look at Remedy: they limit the scope of their games, but what they are able to accomplish is still highly impressive in its own right.
 
HL:A, the project managed to build up enough momentum internally to avoid being cancelled

They bought the firewatch company basically canceled the project they were working on and those guys helped finishing hla.

People forgot in the valey of gods.
 
How well games look that are done with a specific engine doesnt matter anymore to devs nowadays.
What matters to them is how good the tools are that allow you to make the game as fast and efficent as possible.
Most of the dev teams outside of Valve themselves that used Source1/2 in the past have switched to Unity, Unreal or something else for exactly that reason.
Sure, you can cut the grass in your garden with scissors, but why do that when there is mowing robots that do it 100 times faster.
devs are irrelevant to this equation, what matters is what those who pay the devs think AKA the consumer, the players

when you said that the source engine was not capable of meet the crazy high expectations of today players that just not true at all, the alyx proves this
 
They bought the firewatch company basically canceled the project they were working on and those guys helped finishing hla.

People forgot in the valey of gods.
I mean I doubt Valve put a gun to their head and made them stop working on it. Likely no one joined their team or they were attracted by HL:A. It looks like Jake Rodkin and Sean Vanaman are still at Valve and Chris Remo moved to London. (Though his LinkedIn bio says he is still working at Valve, so maybe he is a consultant/contractor).
 
Last edited:
They were stuck in development hell on continually cancelled projects. Valve's issue is not with the individual employees but with a flat management structure and an endless source of money (Steam) that means they never have to actually ship anything. But we're talking about the hypothetical scenario where HL:X does get shipped, because like HL:A, the project managed to build up enough momentum internally to avoid being cancelled. (Valve actually re-wrote HL:A's story from scratch like a year before release).
The flat management structure is exactly the reason why those projects ended up in development hell, and as you say correctly, limitless money does not help either to create any pressure to deliver anything.

Looking at the experienced engineers who still seem to be working at Valve, I count Yahn Bernier, John McDonald, Jay Stelly, Dave Kircher and Brian Jacobson, who are 20 year+ veterans – plus a bunch of more generalist programmers who joined in the late 2000s. The way Valve works is each game will have a specific set of technical challenges that they will be focusing on, and they will update the engine to meet those challenges. The focus of HL:A was on confined spaces, due to how VR limits traversal, so the lack of level streaming would not have been a major limiting factor. If HL:X is a lot more open, then we would expect level streaming to be more of a focus. Being an AA sized developer means taking on a limited set of challenges at a time. It doesn't mean that no technical challenges can be accomplished!
Sure there is still some around, but look at the complexity of any modern engine to realize that 5-10 people are nowhere near enough.
See CD Project Red who had a huge dev team just for the engine, but now switched to UE becasue they couldnt maintain it anymore.
If you think Valve would magically do better with less people, think again!

Just look at Remedy: they limit the scope of their games, but what they are able to accomplish is still highly impressive in its own right.
Remedy´s Northlight engine was written from scratch with no old baggage like Source.
 
Reality check:
- Source 2 is an extremly outdated engine that is 1-2 decades behind the competiton. Even HL Alyx still didnt have level streaming, like wtf! UE3 & others had that in 2004!
- In order to make a HL3 that meets the crazy high expectations of modern players, they would either have to use a modern engine, or make a new one.
Using a competitor´s engine is unlikely and making a new one on the same level as Anvil, UE5, Frostbite etc, requires a very large amount of engine coders.
On top of that, you need another large team to develop a Triple A title with the new engine.
- Valve does not currently have the staff numbers or expertise to pull off anything like that, and we havent seen them going on some sudden hiring spree of experienced devs either.
- Half Life is an IP only old people like us care about. Releasing some AAA HL3 to a very limited audience without having remade HL1 and HL2 first will lead to disaster.

Conclusion:
If we ever see a HL3 - which I doubt - it will most likely be done in an upgraded Source 2 variant aka Alyx and not try to compete with anything that is on the market.
Source 2 is not "decades behind competition". Yes, it does not have all the features that Unreal 5 has (which btw, quite a few of them are still in experimental, even years after they've been pushed to Unreal 5...). But Source 2 is not intended to directly compete with Unreal, it's something for Valve to use on Valve stuff, and they don't need Unreal 5's feature set for what they want to do. And if they do need Source 2 to do new shit, they'll implement it (as can be seen by the commit diffs). And in some ways, Source 2 is ahead of the competition. For example, Source 2's implementation of VR and its feature set for VR is more advanced than what Unreal 5 offers. So is its physics system. I highly doubt that Valve's devs "don't have the expertise". They have shown that they know what they're doing (Half-Life Alyx, CS2). It's just that they are very selective with what they spend their time on. Epic aims for Unreal 5 to be used for a variety of genres and platforms, and they need to stay relevant on the market.
 
Last edited:
devs are irrelevant to this equation, what matters is what those who pay the devs think AKA the consumer, the players
Wut? The consumer doesnt pay the devs, the publisher does.
The publisher tells the devs what to do, not you!

when you said that the source engine was not capable of meet the crazy high expectations of today players that just not true at all, the alyx proves this
Alyx proves absolutely nothing - it was a great small scope project for a super niche audience on a platform that had basically no competition.
 
Source 2 is not "decades behind competition". Yes, it does not have all the features that Unreal 5 has (which btw, quite a few of them are still in experimental, even years after they've been pushed to Unreal 5...).
Not just UE5, any modern engine like Anvil, Frostbite, CryEngine. And yes Source2 is way behind the competition in pretty much every single aspect.
But you would know that if you were a developer and actually worked with such engines!

But Source 2 is not intended to directly compete with Unreal, it's something for Valve to use on Valve stuff, and they don't need Unreal 5's feature set for what they want to do. And if they do need Source 2 to do new shit, they'll implement it (as can be seen by the commit diffs).
Yeah it doesnt compete with those other engines becasue it cant. They would have to start a new engine from scratch as Source is based on "old" engine principles that can only be pushed so far.

And in some ways, Source 2 is ahead of the competition. For example, Source 2's implementation of VR and its feature set for VR is more advanced than what Unreal 5 offers. So is its physics system. I highly doubt that Valve's devs "don't have the expertise".
They were ahead 5 years ago, but time doesnt stand still, UE5 VR stuff is now superior as unlike Valve, Epic, Ubi, EA & Co continue to develop their engines.
As for the physics, thats completely laughable - especially after what we just saw what Frostbite can do in Battlefield 6.

They have shown that they know what they're doing (Half-Life Alyx, CS2). It's just that they are very selective with what they spend their time on. Epic aims for Unreal 5 to be used for a variety of genres and platforms, and they need to stay relevant on the market.
They have shown that their engine is good for AA type games with limited scope and thats about it.
Nothing wrong with that, but thinking Source2 is some technical wonder is only something people think that never worked with engines professionally.
 
Last edited:
Sure there is still some around, but look at the complexity of any modern engine to realize that 5-10 people are nowhere near enough.
See CD Project Red who had a huge dev team just for the engine, but now switched to UE becasue they couldnt maintain it anymore.
If you think Valve would magically do better with less people, think again!
I count 38 software developers listed on Valve's employee page. In the past Valve said that 20 - 30 people were working on the Source engine at a given time. While according to the below interview, the Northlight team has over 40 programmers.


So I think it's reasonable to estimate that the Source 2 team could be half of Northlight's. On the other hand all Valve need to do is update Source 2 to support the kind of gameplay they require in HL:X, while maintaining the same visual bar. It's what they did with Episode 2 and Portal 2, neither of which were "cutting edge" on release, but still delivered a compelling experience.

So no, I don't expect Valve to "beat" CDPR. I expect them to deliver compelling gameplay at a reasonable visual fidelity.
 
Not just UE5, any modern engine like Anvil, Frostbite, CryEngine. And yes Source2 is way behind the competition in pretty much every single aspect.
But you would know that if you were a developer and actually worked with such engines!
I actually did, which is why I think that "decades behind the competition" is exaggerated.

Yeah it doesnt compete with those other engines becasue it cant. They would have to start a new engine from scratch as Source is based on "old" engine principles that can only be pushed so far.
And it doesn't compete with other engines because that's not Source 2's goal. It needs to do the shit Valve needs it do to. It's not Unreal 5, which is a thing Epic aims to sell to developers.

They were ahead 5 years ago, but time doesnt stand still, UE5 VR stuff is now superior as unlike Valve, Epic, Ubi, EA & Co continue to develop their engines.
As for the physics, thats completely laughable - especially after what we just saw what Frostbite can do in Battlefield 6.
No, Unreal 5's VR shit is still pretty much a pain to work with. I know this because I have the displeasure to work with it. And in terms of physics, yes, Frostbite in the iteration for Battlefield 6 seems to have quite a few tricks upon its sleeve, which is what should be expected, though. It's been specifically iterated upon for Battlefield 6. We haven't seen anything of Half-Life 3. So, yes. Frostbite is "better" in that regard, because it's a new iteration for a new game, with stuff added to it for said new game. Everything else would be weird, given how much time DICE had.

They have shown that their engine is good for AA type games with limited scope and thats about it.
Nothing wrong with that, but thinking Source2 is some technical wonder is only something people think that never worked with engines professionally
I have never said that Source 2 is a "technical wonder", I have said that Source 2 is competent in what Valve needs it to do. And I still think that. I do admit that I have only worked with Source 1, not Source 2, for obvious reasons. But Source 1 was very much the same way back then. It was good for what it was intended to do.
 
I count 38 software developers listed on Valve's employee page. In the past Valve said that 20 - 30 people were working on the Source engine at a given time. While according to the below interview, the Northlight team has over 40 programmers.
Northlight is being worked on constantly, former Valve devs have said that basically no one really works or updates Source2 unless there is a need to for a project.
That´s also why the promised Source2 SDK hasnt been released for years.

So I think it's reasonable to estimate that the Source 2 team could be half of Northlight's. On the other hand all Valve need to do is update Source 2 to support the kind of gameplay they require in HL:X, while maintaining the same visual bar. It's what they did with Episode 2 and Portal 2, neither of which were "cutting edge" on release, but still delivered a compelling experience.
So no, I don't expect Valve to "beat" CDPR. I expect them to deliver compelling gameplay at a reasonable visual fidelity.
Yeah good art direction can even make old engines look amazing.
 
I actually did, which is why I think that "decades behind the competition" is exaggerated.
I worked with around 10 different engines professionally (not all 3D tho) - and yes fine - decades is an exaggeration.
But 10-15 years is probably realistic. Heck Source2 still uses BSP and non metric units!

And it doesn't compete with other engines because that's not Source 2's goal. It needs to do the shit Valve needs it do to. It's not Unreal 5, which is a thing Epic aims to sell to developers.
So you actually agree with me!

No, Unreal 5's VR shit is still pretty much a pain to work with. I know this because I have the displeasure to work with it. And in terms of physics, yes, Frostbite in the iteration for Battlefield 6 seems to have quite a few tricks upon its sleeve, which is what should be expected, though. It's been specifically iterated upon for Battlefield 6. We haven't seen anything of Half-Life 3. So, yes. Frostbite is "better" in that regard, because it's a new iteration for a new game, with stuff added to it for said new game. Everything else would be weird, given how much time DICE had.
Oh hey a fellow UE5 VR dev, but I cant say I have a reason to complain!
Well given the amount of time and money Valve has, its weird they havent delivered anything better.
Even if HL3 cant compete with other modern games, im sure old ppl like me will be happy to finally see the story end ;)

I have never said that Source 2 is a "technical wonder", I have said that Source 2 is competent in what Valve needs it to do. And I still think that. I do admit that I have only worked with Source 1, not Source 2, for obvious reasons. But Source 1 was very much the same way back then. It was good for what it was intended to do.
So you agree with me here too. I never said that Source was a bad engine, just that it is hella outdated for modern development.
And yes same here, I only tinkered with Source1 abit back then, but then UT99 released with UnrealED and I never looked back ;)
 
Northlight is being worked on constantly, former Valve devs have said that basically no one really works or updates Source2 unless there is a need to for a project.
That´s also why the promised Source2 SDK hasnt been released for years.
Valve said in 2020 that they didn't have plans to release a Source 2 SDK, because it would be a ton of work. It seems that S&box will somewhat serve this purpose.

But this is my point: Valve are not competing as an engine developer. They're not dedicating employees to build a stable platform for a variety of internal or external projects. Instead, they're going to take the tech they have from HL:A and modify it to suit the gameplay requirements of HL:X. All the technology they are developing needs to do is support HL:X. It doesn't need to work as a general purpose platform. If from the end user's perspective, the game looks and runs decently and offers a compelling experience, it doesn't matter how much of the underlying technology is smoke and mirrors.
 
Last edited:
Top Bottom