Hacker News new | past | comments | ask | show | jobs | submit login
John Carmack’s presentation at Facebook Connect [video] (youtube.com)
295 points by tosh on Sept 18, 2020 | hide | past | favorite | 166 comments

This talk was such a treat. After all the carefuly crafted hype from Zuck and the gang, Carmack threw huge bucket of cold water on everything. I think it's just wonderful that he's allowed to speak so openly, about the things he's proud of, the things he thinks are failures, internal politics, the possibilities for improvement, the real hard physical limits VR is running into, and his growing pessimism for the future of VR. I could really hear it in his talk why he moved on to spend more time on his next passion.

Since it sounded to me like the low hanging mobile VR fruit has been picked, I went ahead and pre-ordered my very first HMD afterwards. Time to see for myself if this is the computing future or an evolutionary dead end.

I watched the whole thing and I don't understand how you arrived at him having a growing pessimism for the future of VR?

I had the opposite impression. He did seem to care about VR as a general purpose system (alternative to traditional monitors), even if it's not quite there yet, both due to hardware and software limitations. He also seemed proud that they were actually shipping stuff.

Same here. As an engineer, who is naturally very blunt about work, I just heard another engineer speaking and being honest about all the pros and cons. My overall impression is he is very optimistic about things, and proud of what they've accomplished, and there is a lot they can do to make it even better. So overall positive, and I'm glad Carmack is working on the Oculus and I'm happy to buy one.

Yes, I do think he was very proud of the shipping product and saw that there were many more achievable gains.

My impression of pessimism came from his skepticism that Moore's law would continue long enough to get them to where they really wanted to be on mobile VR, his struggles with latency probably making cloud rendering a poor experience for most, recognition that they were already at the ceiling of their heat and power budget, his belief that foveated rendering was not going to save the day, his recognition that the lenses were now a limiting factor vs the screen for image quality and that other lens designs would be fragile and suboptimal in a shipping product.

I didn't mean to say that I thought he was pessimistic about VR overall, just that he didn't seem to think there were any large gains in hardware quality coming soon and it was down to the software to make full use of what was there.

I'd recommend anyone interested in the future of VR displays to take the time to watch Doug Lanman's (FRL) recent presentation (IS&T Electronic Imaging (EI) Symposium: Quality Screen Time: Leveraging Computational Displays for Spatial Computing) to get a taste of what's coming: https://www.youtube.com/watch?v=LQwMAl9bGNY - it's great, and one of the more optimistic technology talks I've seen in a while.

As far as the limits of mobile computation goes, I think the thing to consider is that Qualcomm is already providing Wi-Fi 6E chips that have VR-class latency (<3ms) - this is bog standard, but of there are of course companies like Amimon that have been offering proprietary ultra-low latency (<1ms) solutions for a while now. We'll see if the PS5/PSVR2 offers a wireless solution, but I don't think it's too much to ask for those looking for cutting edge experience to use untethered, local compute - why do you need to strap an RTX 3080 to your face? There's already more than enough power on mobile to provide the basic I/O for presence and engagement in VR.

That is an amazing talk. Thanks for sharing. Lanman also seems to indicate that eyetracking is really hard and is still unsolved to the level it would need to be to drive the halfdome 3 prototype. And yes, that bouncing iris totally did give me the willies. It seems that there may eventually be a varifocal HMD with eyetracked fovated rendering, but I'm doubtful it's coming in Oculus Quest 3 or even 4.

Personally, I'd wouldn't count out electronic varifocal coming in Quest 3 (2022?) as it looks like the Half Dome 3 prototype was already TRL6 w/ the electronic varifocal, and Facebook has been making additional moves like acquiring Lemnis earlier this year, which already has a similar-approach full-stack commercial varifocal system. Varifocal would be a huge win for comfort/long-term usage (eg, if FB continues to push on the remote work/productivity/collaboration use cases).

While I'm less sure on foveated rendering (although I'm really impressed by DeepFovea https://research.fb.com/publications/deepfovea-neural-recons... I wonder what kind of rendering pipeline will be required), I think basic eye tracking will be coming w/ Quest 3 as well - it's not only needed for DeepFocus, but also their Codec Avatars.

The stuff that FB has been showing off the past few years is insanely impressive - I think their mixed reality approach beats the pants out of what MS has been doing. I'm really curious to see what Apple has been whittling away at in secret - I'm doubtful if it can compete, but who knows? Google seems a bit out to lunch on all this.

I mean, I think the last thing he said regarding Moore’s Law was the comment about what that AMD exec said, which was basically that they still have tons of room for future improvements.

Specifically, that was Jim Keller who gave the "we've got this" response to that. I've been grumbling about moore's law for some time, but if Keller says that it's possible, I'll believe him on this.

He said he was skeptical of that exec's claim, just not in a super call them out way. I did not take that section to be particularly optimistic. We are probably near the end of die shrinks anyway, and with them, the thermal advantage they bring.

The limit is quantum dots, and that doesn't afford very many more doublings.

> Time to see for myself if this is the computing future or an evolutionary dead end.

There's a third possibility - VR is finally here and consumer market is what it is. It's not going to take over the world but it isn't going away either.

I'd say that is the same as a dead end. Right now this is a very expensive, multi-billion dollar toy with toy demonstrator apps and proof of concept games. It needs to be a game changer and the future of something or the research funding and developer ecosystem are going to dry up. And given the amount that Facebook must be losing on the Quest 2 and the all-in nature of this launch, if this doesn't take off now, the runway for Oculus at FB may be over.

>Right now this is a very expensive, multi-billion dollar toy with toy demonstrator apps and proof of concept games.

We are far past tech demos and concept games.

But it's clearly not a mass market thing either— mass market is mobile games raking in millions, Switches selling out during quarantine, the upcoming PS5 launch, and Epic having splashy platform wars with Apple and Valve. Those are billion dollar gaming events which all have nothing to do with VR.

There's no non-nerd out there who thinks "maybe I'll try VR", and goes into a store to buy... whatever it is, goes home to have a great time with it, and immediately tells their friends. No, it's Linus Tech Tips upgrading his water-cooled living room PC with a special USB3 card just to get enough bandwidth to all the peripherals: https://www.youtube.com/watch?v=HFuLvGf0g0c

Others have said this, but I'll reiterate: your description of mass market is precisely what's happening with Quest. Your view of VR seems rooted in pre-Quest (i.e., exclusively PC-tethered) VR and not the stand-alone world.

In fact, so many people literally went to Best Buy or Walmart to purchase a Quest that it has been sold out since the beginning of the year, online and off, with eBay pricing at 50%+ premiums.

You literally open the box, run an app on your phone (for configuration), put the HMD on your face and you're in VR. People not only tell their friends, but buy multiple versions to give to friends and family.

With Quest 2 at a $299 price point, the non-nerds saying "maybe I'll try VR" will happen even more often. The OculusQuest subreddit has a constant stream of questions about this very behavior. It is currently the ~2,000th most popular subreddit, with an impressive growth curve[1]. Some are literally buying the Quest 2 instead of the PS5 (though clearly a small number).

So far, 35 games have made over $1m on Quest alone, with one clearing that figure in 4 days and another in 8 days. Top games have made over $10m, not even including revenue from Steam, PSVR, or the Oculus PC VR platform. These aren't Epic / Apple level mega-wins, but many mobile developers would be happy with $5-10m revenue figures.

Upcoming VR games are coming from top tier publishers like Respawn (Medal of Honor), Ubisoft (Assassin's Creed, Splinter Cell), and Rockstar (unnamed AAA open world game)[2], Crytek (The Climb 2), and EA (Star Wars: Squadrons).

Additional games include Sniper Elite, Warhammer 40,000, Star Wars: Tales from the Galaxy's Edge, Jurassic World Aftermath, and Myst. And this is just the tip of the iceberg from brands/publishers you may know.

Point being, stand-alone VR has really addressed many of your (and the market's) concerns.

And, to be fair, in your linked video Linus was trying to use optional, expensive peripherals for full-body tracking while streaming - not exactly a mass-market use case.

[1] https://subredditstats.com/r/oculusquest

[2] https://uploadvr.com/rockstars-new-vr-game-guesses/

Great perspective! I think things are early stage, but that the next couple of years are going to be the most exciting period for VR with possibly several of the upcoming games you mention hitting very large sales numbers. There's no guarantee, but just like the Wii was the must-have console of its generation, the Quest 2 has the potential to be a runaway success with sales numbers simaler to the consoles. As technology watchers, sometimes we watch these incremental improvements over many years and its hard to recognize the moment something is about to pass into the mainstream.

The problem is, those numbers are nowhere near close to the price of making triple-A games. I heard Assassins Creed Origins cost 80 mio and Odyssey was 100 mio. With that in mind, "Top games have made over $10m" sounds like we won't be seeing high quality VR games anytime soon.

A few points to consider:

1. Facebook funds many of these games; think of it like a first-party game made to drive sales of a new console. The developer gets paid up front, Facebook recoups the revenue via incremental sales of HMDs and software on their store.

2. Note that many games are VR offshoots of existing IPs. This means the opportunity for asset re-use and consequently lower costs.

3. "High quality" and "big budget" are not equivalent. There is a huge spectrum between a tech demo and a flagship title with 400-500 people working on it (as is the case of AC). AAA-caliber, lengthy games can be made by teams of 50-75.

4. Again, those revenue numbers are on Quest only. PSVR is an even larger market, and PC VR (Vive, Index, WMR, Rift) is probably about the same size.

The most important point though is that if your entry point requires Cyberpunk 2077, Red Dead Redemption 2, or The Last of Us, then it is indeed too early.

However, the two hottest non-VR games right now are Fall Guys and Among Us - great, high quality games come in all sizes.

So now it's "AAA or it's a flop"?

It does feel like the goalposts are being placed wherever VR isn't.

> "..."maybe I'll try VR", and goes into a store to buy... whatever it is, goes home to have a great time with it, and immediately tells their friends."

This is definitely a thing that happens with the Oculus Quest already. The hardware is easy to use and the software library is pretty much in place now. A major marketing campaign could make it really take off.

I don't have any insight into FB VR's decision to require the login on Oculus devices now, but my guess is that they consider the device to be ready for a mass market, and believe that the gains from the social graph will be worth the short-term pain of leaving early-adopter money on the table from here on.

Yes, really agree with this perspective. Last year Carmack's talk was equally interesting and he stated that they had all the usage data on several different visions of what VR could be (Rift, GearVR, Go, Quest) and that the data showed the Quest had the most engaged users and the product finally showed signs of being sticky. At the time it convinced me not to buy because their v2 was clearly going to get a big investment now that they had plausible product market fit.

I think the Quest showed FB that they had a potential winner on their hands and the Quest 2 is moving out of MVP phase to mass market phase. They are gambling that this is going to take off in a big way and they pushed hard to get an improved version out there and ramp up their manufacturing faster than most thought they would. They clearly want to be first to the "Oasis" with Horizon if that's ever going to be a thing and the FB login is a piece of that. If this is a flop, it will be a huge disappointment to the execs. If it's a win then FB may have been first to plant their flag in our VR future, for better or for worse.

I keep seeing the word "Oasis" being mentioned on the recent Oculus threads. What does it refer to? Some company? Some meme I'm missing?

It's a reference to a VR environment in the sci-fi novel "Ready Player One" by Ernest Cline.

Thanks! Haven't read the book yet, and forgot the name from the movie.

Agree. Compare the popularity of Beat Saber to what Wii Sports was and you have a measure for how far out of the mainstream VR still is.

The PSVR has sold 5 million units or so to date, and there's going to be a second generation. There are a lot of games which support it due to this install base. That's pretty mass market.

(It's popular enough that Sony actually has made a second iteration of the headset, but it's not an upgrade.)

It's not as accessible as the Oculus Quest, but the PS4 has sold roughly 106 million units to date, and gamers are probably the biggest market at present for VR devices.

I completely disagree. Quest and a game like Beat Saber are about as broad as you can get. Right out of the box you're swinging your arms around to popular EDM music. The price point is low enough to make it a solid holiday buy for a good amount of people as well. VR is also one of those things that you can show anyone and really wow them. It's not like handing a controller to a non-gamer and having them discouraged because they don't know how to use 2 joysticks.

Many games like Beat Saber, in which you move with the beat, have been popular fads, but home hardware for those games (like Dance Dance Revolution dance pads or Guitar Hero guitar controllers) has never been more than a niche product.

VR needs more than that.

Quest also sold out during quarantine.

There was a time when everyone in America knew who Mario was.

Is there anything in VR apps/games that’s even passingly familiar to 1% of average people?

Personally, I know Myst has been ported to VR and that there’s some way to watch NBA games in VR but that’s about it.

For example?

Half-life alyx is wonderful. I cannot praise it enough.

I played too much half-life in my college days and think of it as fun campy sci-fi. Stuff like press e to move this dead body is suddenly this viscerally horrific experience.

You picked up a box of matches accidentally? Read the label. Found a pen draw something.

Echo Arena, onward, pavlov, beat saber, arizona sunshine, stormland, the climb, journey of the gods to name a few.

oh and myst is coming, MYST.

Beat Saber is the obvious answer with the most general appeal.

Right. Beat Saber works because the virtual world does not rotate relative to the real world. That rotation is what induces nausea. The player is confined to a small box in the virtual world; they never move very far, and the virtual world is locked to the real world.[1]

Games with that constraint are limited. You could do ping-pong. Shooting galleries.

[1] https://www.youtube.com/watch?v=gV1sw4lfwFw

Ok but it's not a tech demo or some proof of concept. It's a fully fleshed out game with high profile DLC and a strong mod community. Nausea has nothing to do with this.

I'm not sure your understanding of VR motion sickness is quite right. Most people are completely fine with teleport locomotion which is sufficient for a huge class ofs games.

Also cockpits seem to reduce nausea thus opening up another broad category.

It's mostly smooth first person motion with strafing and smooth turning that is problematic.

> It needs to be a game changer and the future of something or the research funding and developer ecosystem are going to dry up.

I hate that we live in a world so dominated by network effects, economies of scale, corporate consolidation, and shareholder value that there are fewer and fewer realms where "small success" still exists.

Me too, but then given the complexity of this tech, I think it's much closer by order of magnitude to a moon landing worth of engineering than Pat's Sustainable Bamboo Bicycles. Big projects, big scale, big money.

Palmer lucky literally did built the first Occulus in his garage. I don't think it's fair to compare this technology to the moon landing.

You're right that some sort of ready player one metaverse thing is probably on that level, but creating genuinely interesting, fun VR experiences that may in many ways be more interesting and unique does not to me seem beyond the scope of small scale business.

In a sense VR is one of the purest versions of that sort of cyberspace idealism, it's an escape hatch away from curated, corporate products to whatever people want to make. So I'm not pessimistic that creative people find things to built.

Me, too. I think economic pressures like you described generally harm quality of life. We end up with less "real wealth" --all the work-products that would otherwise make net QoL better, for the sake of monetary wealth concentration. As usual this comes down to the incentives and disincentives we've baked into our system not aligning with the best overall outcome. They align with the perceived needs of the biggest fish only.

Being able to buy a VR headset for 400$ is "real wealth".

I have to disagree. This argument comes up in every discussion about wealth inequality. 'The non-wealthy can't complain, because everything is so cheap now.'

Real wealth is, you know... actual wealth. It's being able to start a high 6-figure to 8-figure business (or even smaller if you are satisfied) and not being laughed at as a failure. It's being able to buy rather than rent everything you want to buy. It's having productive assets.

> It's being able to start a high 6-figure to 8-figure business (or even smaller if you are satisfied) and not being laughed at as a failure.

I suspect this is a bit of a SV bubble viewpoint. I don't think that in the general population, if you tell people you are pulling 8-figures in your business, you would be considered a failure.

Yep, sounds like someone who spends way too much time on forums. Do they even have a 6-figure business that's literally getting laughed at? And was that "laughter" more than maybe an internet comment, at most? Starts to sound more like superstition.

>> It's being able to buy rather than rent everything you want to buy. It's having productive assets.

If by "productive assets" you mean income/revenue producing, then I'm not sure we can all have that. Many assets like that require some else to pay rent.

People can now buy a haedset for 400$, rather than 100000$. That creates wealth.

That you personally care more about starting a business than about using a VR headset is not really relevant.

It creates purchasing power only in VR, but that is not the same thing as wealth. I think it's important to keep that distinction. Wealth is more directly connected to financial stability than the price of specific technology products.

I don't think you have a proper concept of "wealth".

I appreciate that. Economy of scale makes it possible for really amazing products, concentrating the efforts of many, to be spread far and wide. That is a great thing, but there are very serious negatives that come along with that incentive to scale when it's not tempered. Hegemonic industries disproportionately empower owners over workers and limit opportunity, competition, and diversity to the point where the "best" products can shove anti-consumer features down people's throats. You get unsteerable inertia.

It's not cut-and-dry bad the way we do things, but there is a lack of balance. There's a lot wrong that could be better if it wasn't someone's perverse bread and butter. Extreme wealth and power inequality, whether for people or companies, don't make for a just, healthy society.

I appreciate your concerns, but I think they are mostly propaganda. Without the industries, there would be no jobs and no workers that could be deprived of their power.

If you want to strengthen workers, create more demand for jobs, so that workers have options. By for example creating a successful VR industry, you create demand for more jobs that haven't previously existed, thereby empowering workers. Somebody who has previously been exploited to manufacture fridges can now give their boss the finger and go manufacture chips for VR headsets instead. One easy way to create more jobs is to hire a cleaning person, by the way. Everybody who is concerned about "living wages" and things like that should start by hiring cleaners and possibly gardeners for their homes.

I personally think the whole concept of "workers" is bunk. Everybody is a free agent entering contracts with other market participants. That's it. "Worker" is a social construct, actually, a socialist construct. More so, being a "worker" does not sound like a very desirable state of affairs, so I think society should work towards getting rid of workers.

Wait a minute, small successes still exist. There are orders of magnitudes more objects available to purchase than two decades ago.

The success of mega projects (iphone, etc.) creates a massive supply chain (mostly in China) of people putting things together in different ways to make random stuff.

E.g.: http://www.arcadeguns.com/index.php?main_page=product_info&p... I was just looking at this to turn our entire wall into a shooting arcade with Aimtrak lightguns to help kill time during the winter pandemic.

This was not possible ten years ago.

There are plenty of places where "small success" still exists, usually in the form of tools that solve a problem for a specific trade/industry/etc. Though you won't see these small successes on HN.

As a society we need a way to determine where we invest resources. Some of those metrics you mentioned, though not perfect, are proxies for evaluating whether or not we should be investing more in certain tech. Otherwise, how would we determine where we should spend our limited resources?

Small successes haven't gone anywhere. However a system that needs billions of dollars in corporate funding worth of research to come into existence cannot be one by definition.

What makes you think VR headsets take billions of dollars to launch? The first Oculus was a Kickstarter project.

I think what Facebook and its vendors are building takes huge investment and lots of the best engineers you can hire. It's not just an lcd panel with lenses glued on, it's a room scale computer vision system, a VR optimised cpu/gpu with a suite of sensors, a VR software platform, controllers packed with sensors, computer vision tracking of controllers and hands, developer tools, a consumer app store, first party apps, a social VR metaverse, a telemetry system, a manufacturing and distribution supply chain, and a marketing push to get it into millions of hands (and it's also highly likely each headset is sold at a loss). I don't know how much money it takes to do all that, but I know you can't raise it on kickstarter.

Some of that might be difficult, but none of it is impossible and all of it is getting easier and easier. Some parts don't need to be invented or kickstarted because it exists already and is ready to go.

It also used to cause motion sickness, had terrible resolution, lots of lag and no games/software that supported it. AR/VR as a whole has come a LONG way since then to make it an actual consumer product.

Um, it is simply very expensive to create a good VR experience. It's not expensive BECAUSE of network effects and economies of scale and corporations. It is only POSSIBLE and AFFORDABLE because of those things.

For another example, to make a modern movie can cost hundreds of millions of dollars. It is only possible to make them because there are many people who will possibly watch it and pay for it. Nobody would be able to film the Lord of The Rings trilogy in their own basement just for their own entertainment. Not yet, anyway - tools of the future may make it possible ("GPT-300, make me a good movie based on the LOTR books").

In my opinion it’s just like every other games console: an expensive toy that no one needs and lots of people buy anyway because it’s decent fun.

Meh, the resolution is super close to where it needs to be to read comfortably in the headset and do real work (and with wireless headsets with inside-out tracking, the annoying setup overhead is all but eliminated). That can make it a valuable peripheral, like a nice monitor without all the space requirements.

It won't likely ever be as big of a platform as smartphones are, though. So I overall agree with you.

If the market is about 10 - 20 million users, that's plenty to support developers and future development. The investors who made big bets probably aren't going to see much of a return on that money, but that's okay. VC money is probably bad for the sector in the long run anyway.

Don't forget that the current generation of VR hardware has roots back to a Kickstarter project. Billions of dollars are investment aren't needed.

> proof of concept games

Half Life: Alyx is a fantastic AAA game that in no way feels like a proof of concept. Strongly recommend playing it. (Beat Saber too)

Have you played Onward on the Quest?

>I think it’s just so wonderful he’s to speak so openly...

John is known for being completely candid in his talks. It comes with the territory. It’s like, you get carmack’s mind, but you also get his mouth... I for one, would accept that. I listened to his quakecon talks about the physics of light many many times. The guy is fascinating how his mind works. But more importantly, he speaks his mind because he has the chops to back it up and ACTUALLY FIX the things he complains about.

I don't understand how you equated Carmack speaking about implementation challenges with being "pessimistic about VR". He still seems very gung ho about VR

EDIT: if he's moving towards other projects, actions speak louder than words. I didn't get that from his talk, or maybe I just missed it.

Yeah, that dude definitely didn’t comprehend Carmack’s talk very well. Carmack wasn’t in any way pessimistic about VR, he’s just disappointed with the pace that things are moving in. As he said, he always wants things to move faster.

But if anything, Carmack expressed overall optimism. He said that we pretty much have everything we need to implement his ideal vision of VR. It’s just a matter of putting all the pieces together.

Farther up I detailed the specific things I thought Carmack was most pessimistic on. I clarified that I don't think he's down on VR, just on big hardware gains on mobile in the near future. I took it to mean "Hardware's done. Have at it, devs." And I think that's the reason he's moving on to the next project, since this tech is now in the polishing phase and out of the pioneer phase. I think it's a good thing and a sign of a maturing platform, just like how slowing advances in smartphone hardware marked the begining of their global sales s-curve (LTE retina iphone 4 in 2010).

>I could really hear it in his talk why he moved on to spend more time on his next passion

what is he moving onto?

AGI, artificial general intelligence as an independent researcher.

He was complaining a couple days ago in twitter that he couldn't buy a ton of expensive gpu machines on GCP as an independent researcher, but being John Carmack someone from google stepped up and fixed it for him.

Was that where he was talking about how Google wanted him to have a website to prove his validity (before people stepped in of course). Reading that thread was bizarre...

haha, yes that was it

I wonder why he didn’t just move over to Azure, they don’t make a distinction between a corporate person and an individual, as long as you can pay your bill. (I’ve never hit the oh no limit, anyway. He’s probably using a ton of power though).

Link below. Anyway why do they have such a low quota? To stop people from using it for Crypto? But surely they could have make multiple account if it needs to be.

So why put a limit?


Actually I just realized, after thinking about the forward looking AR research talk by the VP or whoever, that AGI is actually the long-term research goal for Facebook's AR. They want a personal assistant that understands the world well enough to insert content that you need.

So John is actually still working on the next phase of the Facebook AR project in my opinion.

Nope. He's collaborating with his son on it, and he's being extremely careful to keep the work highly partitioned for IP reasons. He doesn't want a repeat of the ZenniMax lawsuit. I'd read it as his sunset project for his career, where he gets to take a longshot at something really ambitious, while enjoying hacking with his son.

I didn't mean he was actually working for Facebook. I meant it's an extension of the project there. Even though it seems unrelated.

I sure hope you want for a pc vr headset and not the oculus quest. HTC Vive Pro and Valve Index are simply stunning and can see them become the future of gaming.

As an owner of both, it's hard to see them as the future because they are not accessible or democratized unlike the Quest product line.

Let's start with cost. Both headsets alone run around $1000 - $1200. If you build your VR ready PC, it's about $1500. If you pay Dell or HP, then it costs about $2000. We're already at $2500 minimum and we haven't even added the costs for Vive's wireless option which runs at an extra $350, so with wireless the cost is $2850 minimum; for the non-technical it could skyrocket to $3250 vs the Quest's $299.

Now let's talk about the "initial hill to climb" that most techies forget about. If you must deal with wires, then you have to do ceiling based wire control, where you mangle your room with ceiling hooks. For base stations, unless you have the perfect room with perfect furniture and you don't mind drilling into walls, it's a pain to set it up just right. It took me a lot of trial and error before I got it working flawlessly all the time. Let's also not forget that you need to install a wireless card for Vive's wireless accessory. Not everyone has both the skill and patience to deal with these time consuming annoying things.

Sure, there's no option for full body tracking, visual fidelity isn't as good, and nothing beats external tracking performance; but with Quest, it's $299. You put it on your head and you can play. I can buy a Quest for my mom.

Don't get me wrong, I really love both the Index and Vive Pro, but I am not forgetting about the work and costs associated with either platform

EDIT: I forgot the ritual I have to go through to get Vive and Index to work

1. Turn on computer

2. Put wires on ceiling

3. Turn on each base station

4. Turn on VR HMD

5. Turn on controllers.


1. Turn on Oculus

2. Turn on controllers

It's bad enough where my SO "doesn't want to hassle" me just to play VR.


Except for ceiling cabling, which i personally dont need as i usually play seated games, i agree with everything else. What i meant tho (sorry for my phrasing) is that their _specs_ are the future. Less cables is great and indeed oculus is pretty straightforward. So i suppose once they cut the number of cables (or make htc’s wifi addon cheaper), make trackers remotely manageable and drop the price, i would see increased adoption.

Yes, as the other poster said, it is the best of both worlds device, especially since the USB 3 link now has several times the bandwidth of the Quest 2. I think PCVR is interesting in a supercar kind of way. It's going to provide an incredible experience at a price that will limit it to a small niche of the overall gaming industry. I purposely bought a mobile HMD because I think that will be the only way VR hits the mainstream, gaming or otherwise, so I wanted to understand what mobile can offer on its smaller performance budget.

Yeah I dont think vr’s future is mobile - cant be. There is only this much performance you can squeeze out of a mobile platform (even carmack himself seems to suggest that). Desktop gaming pcs are becoming cheaper by the day and gpu and cpu power are going up. My previous “gaming” pc cost less than 600 gbp and it handled oculus quite well. Now i reckon i own a quite powerful pc, cheaper than a macbook pro, and its flying through games. And most gamers are willing to spend money on such pcs. Anyway up to you, but i would recommend trying pc based vr as i have the feeling you will get hooked!

Thanks, I hope to soon! I'm personally in your boat and would happily tether to a space heater if it got me some extra polygons. I think the data shows there are far more gamers on console, mobile, and laptops than on rigs. And that's part of Google, MS, Nvidia, and Sony's big bets on cloud gaming, that plenty more people would game on their potato laptops if they could. I think the PS VR is an interesting model as well since I can see general consumers going for a wireless headset powered by the big box under their TV. If the Quest 2 doesn't meet FB's expectations then Sony's model may win.

With the Quest being able to be used as both a PC VR and and standalone, why not Quest? I'm saying this as someone who owns 4 different VR headsets, and I find myself using my Quest 1 the most.

Interesting - i steered away of oculus generally speaking as it was by far the buggiest headset I owned. From what gather tho it can only play low spec games decently. For instance, how does it handle Alyx?

I think it can only play Alyx through the link on PC, unless Valve decides to port it at some point in the future, which seems unlikely since they have their own investments in VR as a platform.

Valve is not able to port it - and likely never will be - simply because as a low end headset, the quest cant handle it. From what i read, oculus link leads to latency and compression quality losses so it cant really he played via cable either.

There are two ways to play pc vr games on quest. First is via wifi, with apps like Virtual Desktop, this gives you very good quality, but depending on your setup can range between perfectly acceptable latency, or terrible jittering and lag.

Oculus link gives you very good latency, but with the default settings the quality leaves a lot to be desired. You can tweak the render settings though, which fixes the issue, and it ends up looking about as good as Virtual Desktop. Only real unsolvable problem is the cable, but it's not a huge issue for me.

Alyx is quite playable through both of these setups (assuming good latency for a virtual desktop setup).

Take a look at 29:40 in Carmack's video. The Quest 1 link was running at just 150mbps due to encoder limitations and it apparently ran on a USB 2 cable and you didn't even need their fancy long USB 3 cable. Sounds like there's a lot more available encoder bandwidth on the Quest 2, so I'm hoping link will end up almost as good as traditionally wired VR.

The visual quality of the Index is hard to compare to the Quest.

Valve Index is great, but tbh, I've kind of abandoned mine for the RiftS in the front room. Losing the sensors and extra wires made the wife happy and the difference between the two isn't that great for me. As soon as Facebook lock you into FB accounts though, it'll be tossed to one side. I deleted my FB account 8 or 9 years ago and there is no way I'm going back over a VR headset. Thankfully i've purchased very little through the Oculus store.

Hopefully you went with a real VR headset instead of the Quest. I guarantee you will be disappointed with the Quest.

I own several headsets including a Quest and I'd be happy with any of them. I guarantee you're mistaken.

He is an incredible public speaker.

The first time I heard him, almost 10 years ago, speaking to a journalist at E3, I was shocked. So much detail, but so well communicated. I can happily listen to him talk on almost any subject.

Oh, he's improved over the years. You should watch his very first presentations, during the Quake era. I think he's taken some speaking lessons to diminish his "uhms". But yeah, he's fearless of the public, which removes his mental blocks. I used to stutter in front of the public, but the one thing that solved it for me was a course, where we had to improvise in front of a camera. At the end, the coach replayed every video back to us. Every participant felt so much cringe that it ironically solved the issue.

I think for a lot of people, their fear of public speaking comes from not having gone through that first time you sound legitimately dumb. For example, saying something obviously flat-out wrong and tripping over your words doing so.

Up until that point, you are terrified of it happening. Once it does and you get past it, everything else seems much easier.

Watching recordings of yourself sounds like a mild form of exposure therapy.

A good approach I think is to put together screen cast demos for work - like with obs studio. After you do that a few times pretty much all your fear of public speaking will tend to go away. It's so easy to do this.

This is a good point - I worked as a sports ref in HS, and let me tell you, you'll be made aware when you screw something up even the smallest bit.

But you know what? You learn to (1) tune it out quick, (2) maintain your confidence, and (3) move on - a skill that's been invaluable for me going forward.

His Joe Rogan podcast is pretty incredible, if you haven’t seen it yet.

I love how he goes everywhere from making cars go faster with compressed air rockets, to getting inspiration from the ductwork in the ceiling. Such a great talk.

Thank you for this comment, didn’t know he had been on Joe Rogan

Rogan was a massive Quake player back in the day

If you want even more: after this talk John took questions in Horizon


It's interesting the way that talk demonstrates the limits of the software he was talking about.

In a big picture way, it's also interesting to hear a prominent Facebook employee being frank about the limitations of the product. Perhaps they're just not allowed to do that when it comes to privacy issues.

I think his point about latency is a big one.

I suspect a lot of the problems with videochat (and VR) has to do with this.

Bad latency is frustrating and wears you down, at some point it's not worth attempting to speak up because it's tiring to deal with.

In real life communication transitions between people are fast and you can overlap a bit.

VR could solve the issue of what people are looking at least which I think is the other issue, though I'm not sure how well avatars will fit in for that.

Spatial audio is also a huge thing for conferences. If you want a good virtual conference experience, you need a hallway track - for the meaningful conversations which happen outside of sessions. And there are some noble attempts but it's really hard to replicate that in a system where everyone is speaking through the same pipe.

So, add a 3d space with spatial audio, and it's natural to form groups. Add really good spatial audio, and the system can handle the different unconscious movements we do to pick up snippets of activity around us and figure out which conversations we want to be closer to. (You still can't replicate the free food, though, so it's probably doomed anyway, but it's worth exploring).

His bit about how you can't really control volume in that kind of experience was really interesting to me.

Chatvr seems to have plenty of “players” - in the millions actually. Premature optimisation is not ideal, and a delay in the tens of ms is not unheard of in regular chat apps either. I mean how many live on a 10 Mbps home connection with high latency yet they make perfectly fine calls.

The full quote of that premature optimization trope is this:

"We should forget about small efficiencies, say about 97% of the time: premature optimization is the root of all evil. Yet we should not pass up our opportunities in that critical 3%"

People often ignore that second half.

> "I mean how many live on a 10 Mbps home connection with high latency yet they make perfectly fine calls."

I'm not sure I agree with 'perfectly fine', people just suffer through things being terrible because they don't know they could be better. Latency on a 1:1 call is bad, but tolerable - things get a lot worse as you move towards group interaction (dealing with group video chats often because of covid it's clear how much worse this is than in a real life room).

It's not premature optimization to get the latency down when that's probably the most critical feature tied to usable performance and better experience, arguably everything else should be lower priority.

We may have experienced different kinds of latency but in my experience i havent noticed much of an issue when using chat vr. I may be biased as i am keen on vr going mainstream and looking back at the times of dialup i cant help but wonder what would happened if we waited for fiber optic before using the internet, because it was laggy. So it may suck for a while but its just fine - no worse than a laggy skype call 5 years ago.

It occurs to me that there are a bunch of fairly simple things that could be changed to make wireless headsets more practical for actual work.

My smartphone hasn't been working, so I've been using the headset a lot more as a secondary computing device. The keyboard is not nearly as practical as it could be task switching requires quitting out of each application. The pass-through environment is only available in the base applications, and it's difficult to use when you have multiple guardians and are switching rooms. Also, the hand-tracking interface is really bad. If there were some way to just grab controls directly in 3D space instead of using the pinch-as-a-mouse mechanism, that'd be way nicer. it borrows the poor windowing support of mobile that was driven by screen real estate constraint, even tho there is less constraint on real estate even than desktop.

If Oculus engineers have any notions of making the Quest 2 (or later headsets) into a general use platform, they should try doing all their work in one. Get rid of their laptop and smartphone and do everything in the headset. I'm sure we'd soon see better text rendering, better support for keyboard and mouse, better switching between applications and guardians, better copy/paste behavior between applications, ability to take pictures with the tracking cameras, more control over how the headset goes to sleep or content pauses (or doesn't) when donning and doffing the headset, etc.

They are working on it, it's just not there yet. Quest 1 has a bit too low resolution for readable text. The hand tracking is fairly recent and not fully integrated. They are inventing pinching keyboards[1]. It's all moving in the direction of AR glasses to fully replace your phone. I think Apple is moving the other way, they will probably drive AR through iPhone.

The camera access is a huge privacy issue. I would also love to get access to cameras. It's in OpenXR standard, so we can hope they will implement it at some point.

That said, Quest is already useful for productivity. There's a modern browser, keyboard support (usb or bluetooth), Termux app that gives you capable posix shell, also apps for streaming your PC monitors.

I've built an in-VR live coding environment that I use to prototype and debug Lua 3D [2].There is also another environment by MrDoob [3]. I'm actually surprised how much Quest is open platform and usable as standalone headset. Very unhappy about FB account though.

[1] https://www.youtube.com/watch?v=2BV1lYp04vk [2] https://twitter.com/j_miskov/status/1275135068282007555 [3] https://twitter.com/mrdoob/status/1304465880462643200

Yeah, Quest 2 has twice the usable subpixels (50% greater pixels plus a full RGB stripe, better than the pentile of Quest 1), so I suspect text rendering is pretty good.

Yeah, Carmack mentioned some official tracked physical keyboard being announced soon, which is definitely the way to go for useful work. Good tracked keyboard and mouse support plus some decent windowing will make it dramatically better than a cellphone and perhaps even a typical tablet as far as productivity whereas by default (i.e. just controllers or default hand tracking) the Quest 1 is slower for precise input than a smartphone.

I actually don't think it'll replace phones. More likely to replace large laptop displays for on-the-go productivity as resolution approaches retinal quality in the next few years.

> If there were some way to just grab controls directly in 3D space instead of using the pinch-as-a-mouse mechanism,

You mean virtual controllers?

think of the scrubber bar on youtube videos. why cant i literlly grab and move it instead of the weird pinch mouse control?

EDIT: Sorry, I was typing the above in VR using my controllers. (Honestly!)

I thought he locked himself up in a cabin in the woods and tried to get into AI lately.

As I understand it he now works part-time at Oculus, dedicating the rest of his time to AI, he mentions it in passing in the Horizon session [0] by but he tries to keep it really separate because he doesn't want to be the subject of another billion dollar lawsuit [1]

[0]: https://www.youtube.com/watch?v=P7mFEGah27c

[1]: https://en.wikipedia.org/wiki/ZeniMax_v._Oculus

Still, too bad IMO - I personally enjoyed him talking about games and CG much more than about VR (which I don't care too much about) even though there is, of course, an overlap. I wonder whether I will enjoy him talking about ML even less since I hear that all the time already in my day job.

You should really try out a good VR headset. I think it will change your mind.

As much as 3D TV did?

Sorry old timer, wasn't around for the 3D TV phase. Sorry your tech back then sucked so hard though!

A bit like in Ex Machina? It would be cool to have a GPT-3 cloud powered clippy like guide in the Quest2, not only responding verbally but non-verbally as well.

Not only that, but he's said his primary interest is in embodied AI.

He remains in a “consulting CTO” position at Oculus.

If you're in an isolated cabin in the woods are you still allowed to use cloud servers?

If you're in an isolated cabin in the woods I suppose you can do whatever you want.

Has Occulus ever tried electromagnetic motion sensors. Sixense probably did that once, thought they never went into production but they said it was super precise and has very low latency.

Occulus is tracking motion with camera which is way more CPU intensive than calculating position based on electromagnetic signals.

One trouble they had though that it wouldnt work well with high energy electric items nearby.

So dooes anyone know if Occulus or HTC vive or other well in market consumer VR device ever tried that or has plans to do so? is this even possible?

EDIT: They also made controllers for Razer. https://support.razer.com/console/razer-hydra/

The problem with magnetic tracking is distance. Unless the motion you're tracking is in a relatively small space, you need a terribly powerful magnet to overcome background EMI, and that introduces problems of its own.

Adding compute power to compensate is a much easier problem to solve.

I think that the goal for Oculus is to do everything using headset-mounted cameras, without the need for a base station, so it makes sense for them to focus on image processing instead of specialized sensors. Cost also matters, especially for a $300 standalone headset, and extra sensors aren't free.

One thing to realize is also that cameras (or lighthouses in the case of the Vive) are not the primary sensors. What really do the job are accelerometers and gyroscopes. Cameras only provide reference points to correct for drift.

Cameras are also nice for enabling pass-through views. Works much better than you might think on the Quest. My phone hasn't been working the last few days, so I've been watching videos (etc) in the Oculus browser with the passthrough environment, and I can walk around a bit and do stuff (brush teeth, etc) while watching a video on a large screen, and it's relatively practical (hand tracking is still annoying and largely 2D so it's still better to use a controller), but this could be improved with software updates and the improved Quest 2. The view is stereoscopic and so eye-hand coordination works basically flawlessly (much better than I thought it would).

It'd be sweet if someone eventually made a headset with pass-through view with high resolution and ability to choose wavelength (UV to IR, maybe even a phased audio array to visualize sound sources for the deaf).

If you could make the tracking cameras work in the Sun and figured out a way to harden the display (perhaps using a protective electronic shutter system) to sunlight, it'd be closer to a general platform.

Magic Leap's controller uses this technology, and as far as I am aware most people said it has inferior tracking to Oculus/HTC/Valve's approaches.


Someone said you can’t have a PC in the same room at all while magnetometer based tracking is operating, so...

Wouldn't be a problem if homebrew gamer PCs were FCC compliant and didn't have gaping RF holes in them.

Yes they did, and it didn't work well enough.

Would anyone be so kind as to enumerate some of the things he talks about?

It's a wide ranging and frank talk on the many technical and management challenges that come with creating a cutting-edge product. His bits about latency and compromise should be required reading in all Compsci curriculums.


I always love John Carmarck's talks and try to not miss it when I come across a new one. He genuinely makes me feel smarter. The worst thing about him leaving id Software was the end of his QuakeCon talks !

Consumers should stand their ground and not buy Oculus headsets until Facebook login is removed.

What is next? Facebook login required to use a pointing device, keyboard, printer? Before you know it you will be completely oppressed.

Recent related discussion (Oculus Quest 2 announcement): https://news.ycombinator.com/item?id=24495138

I really love John's presentations. His candidness and passion make his talks not only compelling, but also entertaining.

question for VR people: is the tech at a place yet where a headset could be a monitor replacement for productivity work? i mean my laptop screen is 13" -- could i instead be coding on a virtual 100" screen, still using my regular OS, keyboard and trackpad?

Carmack talks about this a bit, the Quest 1 had enough pixel density that a big screen in VR (such as the Netflix app) could give you about 720 of useful resolution. With the new screens, it's more like 1080p. But if you're doing multiple smaller screens rather than a big movie theater, they'll be less than that.

They're launching an official project for this, called Infinite Office. You see the world in black and white through the tracking cameras, but you have colorful VR screens around you.

thanks. the infinite office demos are pretty funny. all the users doing "office productivity work" are just browsing facebook

Welcome to Facebook's vision of the future of work!

I would say no. The resolution is not high enough for typical monitor sized text at typical monitor distances. If you wanted a virtual 100" screen but with only as much on it as you put on a 24/27" screen today, maybe, but I'm not sure what the benefit to that is.

Nearly. Some people do, but I get too much eye strain. Having said that, Oculus2 might be enough of a step up.

See this pic for an example:


FOV is way too narrow for that.

Anyone know if this is likely to change? I was playing super hot VR last night, and thought to myself, “this game wouldn’t be so hard if I didn’t feel like I was looking through a toilet paper roll”.

Pimax 5K/8K? Those have 200 degrees FOV.

Could Oculus 2 serve as an ok display for coding?

I use it with a Quest, and the ImmersedVR application.

I love it for learning, or working out a coding challenge.

The benefit to "code inside a void" is immense for people who really struggle with ADHD. I've never felt the ability to zone in like I can in VR.

But I have to blow the text size up quite large. And I don't do it for more than an hour, because the resolution isn't quite there for an all day reading experience.

I have a lot of fun playing my 2D games in VR this way as well. Starcraft is crazy fun on a screen that feels like its the size of a movie theater.

Oculus 1 is nearly there, so I have a feeling Oculus 2 could be great. Having said that, there are people who use Oculus 1 all day. There is an app called Immersed that let's you create giant virtual VR monitors and put them wherever you want, which is pretty wild - see:


It's easy to imagine how useful this could be eg: when working with complex web apps where having 4 screens (IDE, terminal, browser, browser console) actually has great utility.

It's not really comfortable enough for very long sessions. You really need something as lightweight as a pair of glasses for productivity work, in my opinion. Maybe we'll get there eventually, depending on how mobile SOCs improve.

From some of the reviews I’ve seen it seems like it’s either there or really close. I think Tested, in their review said it was possible to have a usable virtual monitor with 1080 resolution, but the virtual monitor had to be large.

I see many developers betting on VR. I could see it happening everywhere but at home.

At best VR somehow is a perfect argument on why people should play escape games or paintball/airsoft. It sound like a second version of the google glass.

You just need to use some common logic: a video game works on a screen. It's too difficult to add new sensors. You cannot bridge reality and virtuality. You either have a video game or a normal game/toy/puzzle, you cannot have both.

Maybe the military will come up with new sensors, but until then, I would not bet on VR.

I tend to disagree. I won an Occulus Go at a meetup and almost gave it away- I didn't realize what I had won and just assumed it was some kind of upgraded google cardboard.

I was really surprised and delighted at how cool it was. It felt raw and fresh- almost like when I first got on the internet, there was just a whole new world to explore, and different types of content and games- some were cool, some were not, but were at least interesting in their mechanics. Especially during covid, I found it super neat that I could watch a basketball game in a "section" of a stadium and chat with those around me, and similarly watch a band perform.

The biggest drawback is that the form factor isn't all that comfortable, and battery life is an issue, but I think there is a tremendous amount of potential in terms of experiences.

You should try a 6Degree of freedom headset like the Quest or a PCVR headset too. It really does click into a sense of reality.

VR games are great at home already. I'm not sure I understand your common logic about video games working on a screen.

Is playing real paintball more immersive and real than playing VR paintball in Rec Room? Of course it is. It's also an expensive hobby with a bunch of equipment that requires you to set aside hours of time and drive to a paintball arena.

You can't just jump from "This isn't a 100% faithful reproduction of real life activities" to "This is pointless and will never catch on."

Pong is clearly inspired by ping pong, but you wouldn't have played it in the 70's and said "Pong is not the same as ping pong, video games aren't going anywhere."

Funnily enough, ping pong is one of the more convincing VR experiences because you only need to move in a small area, and the small haptic feedback from the controller is close enough to a ball hitting the paddle. Just don't try to lean on the table, it's not there.

Thanks for the input

I just have more faith in augmented reality. Less latency problems, more possibilities of games like pokemon go.

I'm curious why they didn't try a version with transparent screens.

It doesn’t sound like you have tried a Quest.

I’m not sure I understand your argument at all - what is the logic you are following?

I don’t think OP’s tried VR either. Any of my friends who tried VR once was immediately blown away by it. It’s such a paradigm shift. You want a play paintball in space, in a medieval setting? It happens immediately. There is nothing in real life that makes up for it.

Watching someone try a good VR headset for the first time is so fun. It takes mere seconds for it to “click”, without fail.

I don’t get your comment fully, but I’m an VR advocate, have Google Glass, 1st gen LG G Watch, Recon Jet, 3rd gen Apple Watch with FeliCa payment, and also an airsoft guy, and one of my friend was talking about cool escape room ideas combined with VR pre-COVID.

uh oh.

Applications are open for YC Summer 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact