Valve drops VR support for macOS
techcrunch.com
techcrunch.com
1. No cross platform builds allowed. We can compile to Windows, Linux, Android, Switch, PS4, Xbox One all from one windows build machine.
2. Shifting sand of non-backwards compatibility. Notice how of the above list everyone except Linux and Android have solid back compat stories. Video games are in development for 2-3 years, and stay on the market for 5+ years with no planned recompiles. So a 7 year period of support after inital build machine setup is expected, anything less is going to get your platform side lined.
3. Small market. Mac users make up a share of sales similar to Linux. "What are you talking about! Linux is 1% while Mac is a whole 4%!", nope! Those numbers are the same insignificance per a video game business plan. Platforms with such low sales, like Stadia, might be support provided the platform owner pays porting costs.
4. Culture. Steve Jobs disliked games, it shows in Apple's support.
5. Horrible hardware. Supposing you've decided to ignore all the downsides, and the small upside still appeals to you. Now you have to tune your game to run on hardware which would put cheap Wal-Mart machines to the test. Apple's install base is almost all integrated older intel GPUs. If you do not support Metal, then you are forced to support these old intel integrated GPUs with bonus hacked up OpenGL drivers.
All points are ones I am speaking from experience on. We shipped TINY METAL on MacOS, the lack of cross build delayed the release. The horrible sales numbers meant carrying forward support was a net drag on the game. All further games I work on do not get MacOS ports for these reasons.
Do video game makers plan for updates? Am I getting this wrong, or are patches and such deployed without a game recompile?
And yes, many expansions and additional content are usually data-driven, so it might not require a recompile.
Every game is different, but for every Subnautica with 5+ years of active development post initial sale you'll get a bunch of successful games and only marginal updates for 2 months.
For my next game, Railgrade, I am planning a couple years of updates, but even that will be dwarfed by the long tail years of continued sales after updates are done. I expect Railgrade to be on the market for 5+ years. For consoles it should be purchase-able for at least a decade until Nintendo shuts down the eShop. Then it should continue being playable for decades more.
At some point emulators will take over and my effort can keep bringing players joy for eons more. I am not interested in Apple's "expiring" model of art.
In a way, this is the whole point of the emulation community. Proprietary software was developed for really old systems and they don't work on modern computers. The old systems had to be emulated as virtual machines on top of which the old software could run.
Hardware isn't horrible from Apple. It's just more expensive for macOS and the complete build quality that has been slipping over the years. All Apple users know we're paying more for the hardware than machines preinstalled with windows.
Did Steve Jobs really dislike games? I assume he might have viewed them naively as a waste of time. Most people consider that assumption true and unless you consider them some type of therapy to unwind from the world but even then it can be argued there are healthier options for the majority of people that play games.
edit: ah yes, I get downvoted for writing an honest comment because that's HN crowd.
Re Steve Jobs on games, you can believe in your own skepticism, or you can hear it directly from John Carmack, the creator of DOOM, who tried to work with Steve Jobs to get a proper foundation for gaming on the Mac. He did not succeed.
https://www.digitaltrends.com/gaming/john-carmack-talks-stev...
As far as I know it is against the terms of service to run Mac OS on non-Apple hardware, so you can't just run a macOS VM.
> Hardware isn't horrible from Apple. It's just more expensive for macOS.
Is that not the same thing, effectively? For a given buying power, a player that chooses mac will have objectively poorer hardware.
Yes, there are a few capable rigs for Apple systems, but they are ludicrously expensive and thus provide a minuscule install base for your game, so it's basically not worth the effort for a triple-A studio. Less demanding games might be more viable, but it's still a lot of costs in porting, build infrastructure and publishing for very little return.
If you start with the premise that you should just barely achieve 60fps with all effects enabled on this year’s top-of-the-line PC video card you’re artificially constraining your market. Make your game fun and make it run on as broad a set of users’ systems as possible and you have much better chances of doing well in any market.
Targeting macOS is only feasible for simple/casual games and some eSports.
Basically every triple-A studio banks on having The Latest Technology for their games, so that is a whole nonviable segment.
Sole developers or small indies often can't bother keeping up with the product lineup (XCode often requires the latest macOS version and that means buying a new machine way more often than for other platforms) just to be able to keep building their game. Remember, these are possibly multi-thousand dollar purchases, not to count the actual porting and testing costs. So, another segment made largely nonviable, at least for initial/planned releases.
This leaves us with what's informally called the "double-A" part of the industry: Studios large enough to have a bunch of cash to spare and one or two people who can work on sussing out the multitude of platform requirements and compatibility/performance issues that arise from the Mac platform.
Not only that, but if you are a "double-A" studio that works with a in-house engine, the costs might be increased still due to the required proprietary Metal graphics API, that must be integrated or at least integrating something like MoltenVK (assuming the engine is new enough to have a Vulkan based renderer already).
The only remaining viable projects are on an exception basis, and that leaves very few candidates for porting.
The mac market has always been small, and the supply of games even smaller. So that supply has sold pretty well and for example Aspyr and Bungie made pretty OK money off of that.
If the mac supply of games was more competitive I'd wonder if those economics work anymore.
For the same reason as macOS games must be build on Apple machines, iPhone and iPad applications must also be build on Apple hardware. There is an entire market of companies that rent out "server farms" of MacMini's in datacenters for this exact reason. It massively raises the cost and complexity of a modern development/integration system to have "special snowflakes".
As it comes to hardware, the problem is you can't get something "reasonable". You can build a $500 Windows compatible PC that's not half bad. You can stick a low in Nvidia RTX card in that (for a few hundred more) and get some really reasonable gaming performance out of it. However to get an RTX level gaming performance out of a mac... you're out however much it cost to buy the new MacPro... so let's call that $10k. Most normal people are not going to pay that.
Renting out a mac mini isn't raising the cost or complexity to something unreasonable either. It's like people are forcing their ideology upon a company that decided whats best for itself.
The hardware comment your wrote in regard to mine is cherry picking out the part where apple users know they're overpaying for the hardware they get.
As a company, adding an additional different testing infrastructure (rent a macmini farm) with the additional pipeline costs (setup, build pipeline etc) as well as the additional opex, again, for a small potential market, is not economical.
Firstly, the hardware cost. It's expensive and slow, but for a business it's a justifiable expense. The real cost is in the ongoing hosting and maintenance. Keeping Mac minis or whatever in a datacentre is a costly waste of space. There's zero remote management, and any failure means a trip to visit the system along with a monitor to hook it up to. Then there's the ongoing system administration cost, keeping it updated, keeping the development tools and build dependencies current. This is an ongoing expense.
Typically, if you need to keep clean environments you need to virtualise, and this means using VMware Fusion or similar. It makes a slow Mac mini system even slower. But if you need to support multiple MacOS versions, or multiple environments, what other choice do you have? It's not like there's a containerisation system for MacOS.
Every other platform can be run virtualised on big metal, like VMware and OpenStack clusters. Linux, BSD, Windows, everything we need to care about. MacOS is a special snowflake exception. This nonconformity adds costs.
As for renting. It's bloody expensive for very little in return. That entire market segment exists solely due to how terrible the hardware and software licensing options are. And the hardware and software procurement and management pain is reflected in the uncompetitive pricing. If Apple offered a developers-only generic VM licence for MacOS this market would vanish overnight.
Dislike may be a bit strong, but John Carmack described him as indifferent:
> Over the years I've been through a number of initiatives where Apple wants to get serious about games, and we've done things with them. The idea way back with Quake 3 on there, that was my deal with Steve Jobs: if Apple adopts OpenGL rather than going off and doing QuickTime3D or something else of their own which was going to be a bad idea, then I'll personally port the Quake 3 stuff rather than working with a partner company on that. And we went through all that. All of our Apple ports have been successful - they've all made money - but it's marginal money, and we have worked with Aspyr usually on all the other ones after that, but I do think it kind of comes from the top.
> The truth is Steve Jobs doesn't care about games. This is going to be one of those things that I say something in an interview and it gets fed back to him and I'm on his shithead list for a while on that, until he needs me to do something else there. But I think that that's my general opinion. He's not a gamer. It's difficult to ask somebody to get behind something they don't really believe in. I mean obviously he believes in the music and the iTunes and that whole side of things, and the media side of things, and he gets it and he pushes it and they do wonderful things with that, but he's not a gamer. That's just the bottom line about it.
> There are people at Apple who want to support all this - and there's no roadblocks for us right now, we're going to support the Mac on Rage, we hope to get a version of Quake Live going up on the Mac there - but it's just that's not what the Mac platform's about, and I don't really expect that to change because it's a tough equation now that you've got everybody dual-booting their Macs and everything: why would you want to go to the extra trouble of [developing games for Mac]?
https://www.eurogamer.net/articles/id-softwares-john-carmack...
Apple later lucked into a gaming market with iPhone, but they're still awkward at it (just see the games they choose to showcase onstage) and there is very little cross-pollination that considers the Mac as a gaming platform.
[0]e.g. see the concluding paragraph on https://www.folklore.org/StoryView.py?project=Macintosh&stor...
Apple II, but Macintosh? I'm not sure I agree with that.
I think a lot of Jobs' opinions about personal computers gelled in the 80's. Back then, with the various Sinclair machines, the Ataris, the C64, and the Amiga, there were a lot of people who said that games actually harmed those platforms. In North America at least, having a large library of games made businesses see those platforms as toys and game consoles. Why would you buy an office full of game consoles? You're a serious business man wearing a serious business suit, running a serious business, and you need computers from a serious business computer company that can run serious business software! You need an IBM!
To be fair, that mindset predates the use of personal computers to do serious graphics business. The Mac is largely responsible for making that mindset obsolete with things like desktop publishing and Photoshop. The same attributes that made the Mac a good platform for those use cases also made it much more suited to gaming than DOS-era PCs.
Is that really true? I was alive during DOS-era gaming and still remember kids from Mac families standing out on the sidewalk in the rain crying like Oliver Twist weenies because they couldn’t get their game on. Meanwhile PC - read _intel_ - games we’re writing their own low-level memory managers to eek everything out of the platform possible. I was inside playing Ultima VII or playing Falcon 3.0 with a remote friend _over modem_ drinking hot chocolate. I think any review of the games catalogs would show that the market clearly judged the Mac as less well-suited for gaming.
'[Steve] claimed to have never read a comic book in his life (“I hate them more than I hate video games,” he told me)...' — https://www.vanityfair.com/news/2019/09/bob-iger-remembers-s...
Thus the average matters.
Only top-end MBPs are half decent for gaming anyway.
I was happy with this setup until Catalina killed 32 bit support. Most of the games I played on the MBP was 32 bit.
Killed Mac Gaming for me.
I’m a Mac user and develop on the Mac so you might think I’d bend over backwards to make an excuse, but...not possible in this area. Those gpus do render beautiful pictures and can do some cool gpgpu crunching, but aren’t appropriate for gaming.
What does this mean?
Now that I’ve had more time since I read the post, I think they may have been referring to adaptive refresh rate.
I think you have to do this if you want to be able to switch between integrated and external. Which we didn't care about but perhaps most people do?
Most people need to be able to power down the discrete GPU when they're done gaming in order to have reasonable battery life, and the most user-friendly way to make that possible is for every application but the video game to use the integrated GPU.
Another factor in play might be that, if rumors are to be believed, future Macbooks will be shipping with ARM chips, with yet-to-be-determined CPU and Graphics performance. That might be a bridge too far for crossplatform projects who are already frustrated about having to maintain a Metal backend just for Mac.
Took about 2 days to take a win project to linux, that same win project took about 3 weeks to port to mac.
That was the first time I’d done it of course (to either mac or linux), but I just found the mac documentation to be severely lacking.
Also all macs are now stuck on OpenGL 4.1 forever, it may not matter much now, but it will show in future.
And yes I know Vulkan is a thing but I think that’ll take a while to fully be adopted by the games community.
Not forever: OpenGL is deprecated on macOS and will be removed in the future. I will be very surprised if OpenGL survives the ARM transition.
As a Mac user, I buy other hardware to play games on! Probably the only real Apple game platform is iOS. And even then I’m a bit more careful of what games I buy.
The only Mac gaming I do is with emulators like DOS Box. The market must be really tiny.
Actually, I’d say the ‘platform’ is Apple Arcade. I imagine that’ll grow more.
You can’t swap sound outputs while an application is running. Fullscreen Direct3D games can freeze in a way that obscures all other windows, including the Task Manager, preventing you from ever killing them. USB device power management is abysmal, and devices often just won’t wake up after sleep. The OS itself is a complete clusterfuck of half-new, half-old, half-ancient things like the Control Panel. Everything prompts constantly—usually with modal windows—for absolutely pointless decisions (and no I’m not talking about UAC).
This is before I even get to what awful citizens of the platform their developers are. Everyone wants to hijack the right-click menu. For all the complaints about the Mac App Store, the Windows Store is—as best I can tell—bordering 100% crapware with no way to find anything of reasonable quality. Everything’s UI seems to have been built with entirely custom components designed by a color-blind child who’s seen too many sci-fi movies.
To top it all off, I am continually just floored whenever I read Raymond Chen’s blog. He often has posts explaining why $thing works some unexpected way or has some sharp edge, and the answer is always because thirty years ago a team at Microsoft built something insane and convoluted (admittedly through the lens of someone watching today), it’s now baked in at the lowest levels of Windows, and it can’t be touched, changed, or removed because of their commitment to backward compatibility.
I struggle to find any redeeming qualities of Windows as a user.
I get that the Mac is a poor platform for gaming development. I really do. But goddamn would I kill for that to change, and for me to be able to run games on my Mac some day.
And yes, I understand that there are people who are forced to use Windows professionally or for gaming (like myself), there are people who use Windows because they don’t like (or want to pay for) Mac hardware, and I understand that there are even people who aren’t ignorant about other operating systems who prefer Windows. On the whole though I conclude that the overwhelming majority of users who could use a Mac and don’t simply do so out of a combination of unawareness and/or inertia.
To your second point, literally none of my complaints are something I could have any possible level of control over, no matter my level of personal investment. Windows’ nested layers of awfulness driven by slavish adherence to backward compatibility are quite literally something I could never visibly impact even if I dedicated my life to doing so. The fact that Mac apps, by and large, try to be “good citizens” of the platform while Windows is a free-for-all of developers doing anything and everything however they see fit is in a similar boat.
Then the application in question wasn't properly written and assumes that the output sound device never changes. About the only application running on my Windows gaming desktop that doesn't follow the sound adapter I select (headphones, integrated, USB DAC, etc) is the Java version of Minecraft.
I use MacOS, Linux and Windows and every single one of them has their own different pile of warts to deal with.
This is SO annoying. I can cross compile to Windows from GNU/Linux quite easily. In fact with Wine I can even get MSVC running. But Apple, even though they employ many of the Clang/LLVM developers, believes that you must use their OS to target any Apple product.
Your only reprieve outside of buying their extremely expensive hardware is using CI services like Travis or Github actions.
Of course, if you want to test your software on Macs, you'll have to buy their hardware sooner or later, but even then you have to set up the build on the system instead of keeping it as pure test system.
If you do it for your day job, it's probably justified, but if you are just a hobbyist, then even hundreds of dollars are expensive. Especially if you live in a poorer area or are still a student.
Looking at System76's equivalent mini, it's $555 for the low end, which has the same RAM and a faster, newer processor. Going to 32 GB takes it all the way to $700, or half what Apple's charging. So yes, "extremely expensive" is what it seems like to me.
Especially given that it should be entirely unnecessary. Just let me do a cross-platform build!
It is a bit of a pain and took me about 45 minutes, but I was going slow.
This has been an issue for decades. Twenty years ago I was doing grid/HPC computing at Sun and there were a lot of bioinformatics firms that used macs and wanted compute clusters of them. It was a huge problem even then.
[0] https://twitter.com/brianstucki/status/800778773998145536
Do you want to be the genius that saved a few dollars on hardware with a Hackintosh only to get the company sued?
It also takes a lot of time to setup so hackintoshing will never gain mainstream numbers.
(I’m not saying that I disagree with some of the other issues)
I really like a lot of the ideas that Apple implements, but the upgrade treadmill is notorious and not one of them.
Another major issue is broken OpenGL: bad drivers, very old version and nowadays deprecated.
In 2010 macOS gaming was in a good spot. Nowadays, Apple is making sure you don't use Macs for games.
32bit was dropped in Catalina, the prior release a year earlier started making it clear it was going away.
Neither intel nor amd sell 32bit only desktop processors, and possibly not even laptops.
Game developers complain about performance all the time, but given simply recompiling you 64bit gives (IIRC) a 15+% perf win it continues to amaze me that devs insist on producing 32 bit only software.
It is software support that was dropped in Catalina.
"Simply recompiling" for games is impossible for technical, legal, financial and other reasons. Some companies do not even exist.
> 4. Culture. Steve Jobs disliked games, it shows in Apple's support.
IMO their lack of support has less to do with hate and more that they just didn't find value in that demographic. No one is making money off gaming hardware. On the other hand Apple has put in a LOT of effort into iOS gaming. Steve Jobs himself had multiple games showcased in his keynotes.
Halo was originally intended as a Mac release, and Steve was livid[1] when Microsoft bought Bungie and made it an Xbox exclusive. I think his apathy toward the games business probably grew from experiences like that, if not that one specifically.
1. https://www.mcvuk.com/business-news/publishing/steve-jobs-ra...
Huh? I mean, surely nobody is making Apple-money from gaming hardware. But plenty of companies of various sizes are decently profitable and making money in this space.
It has a controller that looks like a console gamepad and is one of the target platform of Apple Arcade.
There's the potential for the Apple TV to become a serious gaming platform in the future, but it would require a major shift in product strategy from Apple. That shift would have to include putting much more expensive hardware inside the Apple TV, increasing the BOM by more than the current retail price.
I agree on the need of a shift in product development, but the hardware isn’t what is lacking here.
The high end has quite generous margins and the volumes are absolutely insane, which is why we're now doing HPC and ML on what is basically gaming hardware. Nvidia have a broad spectrum of markets, from the high margin and low volume Quadro to the low margin and high volume GTX - in a business that's dominated by NRE costs, that's a good place to be.
Apple providing support for gaming wouldn't necessarily be profitable in itself, but it'd potentially provide the volumes to improve their offerings for professional rendering applications. Apple refuse to do business with Nvidia and they've been hamstrung by AMD's weak offerings, but a gaming-focused Apple could acquire some IP and tape out their own GPUs.
Some gaming keyboards full of fancy LEDs and software and stuff and military-grade whatever retail for the price of a small car.
Where does this nonesense come from? There's a whole market segment just catering to PC non-console games with companies manufacturing RGB lighted RAM, specialized motherboards and plenty of gaming periphelias. Even gaming laptop market is booming.
And that's even ignoring "tiny" companies like nVidia.
Their graphics performance on iPhones has been top of the line for many years now, they always boost about the graphics performance of their devices they have game demos and game introductions incorporated into their product announcements and finally they introduced game controllers support in the latest OS.
I don't know what else they need to do to target gamers. RGB LEDs maybe?
Mobile gaming is gaming now, if you count where gamers spend their money. Apple sell the easiest, most powerful mobile platforms for developers, and theres are loads of iOS exclusives as a result.
Big budget, cutting edge graphics, fiddly controls, single player big screen adventures... Those games are a shrinking niche for developers. And in that niche, MacOS can only be a 3rd place behind consoles and PCs.
Top Grossing seems to me to be about 50/50 casual / core games.
The medium is more approachable, therefore attracts more casual gamers and many of these casual games have stunning graphics, physics and gameplay.
There are also games that require more serious skills, games like Pubg Mobile and to be honest, I like the mobile version more than the desktop version.
Out of curiosity, what exactly were your struggles with Linux/Android backwards compatibility? Considering how many apps I run still on my phone still use Gingerbread-era UIs, and how Linux executables from as far back as the mid-90's are supposed to be able to run on even the absolute latest kernel (Linus Torvalds having publicly chewed people out on the LKML for breaking said backwards-compatibility), it's surprising that these would somehow be classified as not having solid backwards compatibility.
The main thing that comes to mind with non-Android Linux is library versions, since those do occasionally break backwards compatibility, but this shouldn't matter too much if you're compiling static executables, or using some isolated runtime like the one Steam (IIRC) provides.
As a point of comparison here, World of Goo's Linux version came out in 2009, and it still works on my system (Slackware64-current) which - while not at the absolute most bleeding edge - is still using library and driver and kernel versions substantially more recent than most "LTS" distro releases.
I build for OSX from Linux using clang all the time. You just need to grab the sdk from xcode and get your toolchain all setup.
You value controlling as much of your computing environment as possible? Want source to all the software you use?
Do we really need to explain the merits of libre software still in 2020?
Mozilla cross compiles Mac (and soon Windows, I believe) Firefox builds on Linux build machines.
We do have Flatpak[0] which could be used for indy, DRM-free games.
Valve is also working on a Steam Linux Runtime[1] which uses Linux namespaces to make older titles work on newer distributions.
[0] https://flatpak.org/ [1] https://steamcommunity.com/app/221410/discussions/0/16386755...
I have a 6 year old imac 5K. I don't game a lot but I always had a weak spot for x-plane. X-plane 11 came out shortly after I got this machine and I've been running it throughout. It has improved a lot over the years and they just released a beta with metal support. It's quite amazing that I can get decent performance with this version in 2020 that is better than what I got in 2015 with the same hardware. I run it with number of objects maxed out at a low resolution and mostly can keep the fps above 25-30, which is the minimum for an acceptable experience in a flightsim. The GPU is a still pretty decent AMD R9 M295X with 4GB ram.
However, my Steam library looks pretty depressing. As of Catalina, only a handful of games actually still work. I wasn't really playing most of them but I did spend money on this stuff at some point. I've actually been considering switching to Linux recently specifically because of Steam and having access to a bit wider selection of games. Apparently a lot of stuff works pretty well these days; certainly much more than what works on macs.
IMHO if Apple ever does a move on the gaming front, they have a good shot at domininating this market. If they can manage to mass market a good enough product, the numbers start adding up. That's why the IOS store is such a money maker. I've always been puzzled why they never made more of an effort to position the Apple TV as a gaming console. They tried, a little, but it never added up to much.
Apple making their now widely rumored move to their own CPU/GPU could unlock this. IMHO any VR/AR plans they may have are probably dependent on that. I doubt they'll make life easy for the likes of Steam to run in such an environment though. Apple has never liked middlemen.
Linux is very backwards compatible. I've run an OpenGL application I wrote 10+ years ago without any trouble and have run non 3D applications that were over 20 years old without a need to recompile. This is because the Linux kernel has a very stable ABI.
Unfortunatly, this isn't the case if you ship with dependencies on dynamic libraries but that's pretty easy to avoid. Static link everything and only dynamically depend on SDL or something.
You could also go the windows route and ship all the DLLs. This is what steam does under Linux, they ship their own Ubuntu 14.04
I followed a bunch of online guides to do everything I can to disable the acceleration and everything else, but for some reason when I try to play a game it still feels terrible. Even when the game has a high framerate the mouse never feels smooth.
Tried multiple MBPs. Multiple gaming mice from multiple manufacturers, etc.
Has anyone ever had any luck making a mouse feel as smooth as it does on Windows? If so, what did you do?
I'd get rid of my Windows computer in an instant if it wasn't for the poor support Mac has for gaming. Windows is fine but I enjoy the Mac workflow more.
Regarding the comment above, it is definitely possible there are subtle configuration differences that cause the described behavior (different acceleration curves, polling rate/DPI settings, display refresh rates, etc). FWIW, my “gaming” mouse works fine with my MBP, but it takes a moment to readjust to the different sensitivity and acceleration behavior.
Even if you use the command line to actually disable it for some reason the mouse still feels off.
If you google, for example "counterstrike mac os raw input" or something similar, you'll see lots of posts about people having the same issue. The raw input setting doesn't work properly. Granted the posts are from a couple years ago. The last time I tried to play a game on a Mac was probably around 2018. I tried every workaround I could find on the internet and it was still terrible.
What I described has nothing to do with a difference of sensitivity or polling rate/dpi. Although if a person is unfamiliar with those things they could have an issue with them.
A well-designed game would do that, anyway. I wouldn’t be surprised if the counter strike port for macOS doesn’t qualify as such.
Can you name an online FPS with a decent sized community with proper raw input on Mac OS? I'd be willing to give it a shot and see if its different. There are plenty of games that aren't CS that people have complained about over the years.
Also, raw input or not I was talking about multiple issues. Every mouse I have ever used on the 2 MBP's I've had felt terrible and nowhere near as smooth as the same mice on a PC even for general desktop use. Even Apple's mouse feels pretty bad.
I don't really know why, their touchpads are amazingly smooth and responsive.
Regarding Apple's mice, the Magic Mouse has a rather unusual polling rate of 90 Hz (likely to save energy), which probably explains why it feels awful to use.
Downside is it's a one-shot sensitivity which can't be adjusted, owing to the hacky nature of it, which was semi-found-by-chance to boot. After lots of searching on threads, seems this is one of the only hacks that works, so doesn't seem to be much support from macOS on this stuff.
Used to have an app format of this so I didn't have to run the console command every time at login, but need to find an updated solution for Catalina.
I priced out VR recently and I was unimpressed with the price per value available, with most systems adding $500-$1,000 to my gaming system’s cost, and some requiring that I modify my room to work properly. And that’s not even accounting for the high cost of a VR capable gaming rig, if you don’t own one. This is a pretty steep financial and logistical hill to climb given the current set of titles available.
I'm dead serious though the experience you get in a good immersive VR rig is completely unrivaled. Well worth the hassle and overhead. This will get better as costs come down and developers continue to innovate.
I think one of the problems of VR is one of cost and credibility. Even if I knew you personally and trusted you, there is no way in hell I’m going to float thousands of dollars getting a new gaming PC, VR headset, and customizing my room for the full setup just because someone said it’s amazing. The risk/reward ratio for that is just insane, and not in a good way.
Heck, even $500 for an entry level setup is a pretty dicey proposition, assuming my PC can lift it. $500 can buy me a lot of non-VR gaming satisfaction with effectively 0 risk.
If it was $200, I’d consider it.
There are options available to try high-end VR before you buy, and ~$200 low-end versions if you prefer going in blind.
I know not everyone has a friend with a high spec rig, but that is improving over time as well.
For a try-before-you-by thing, I'd recommend doing 2 weeks. For the first few days, it was, "OMG, the future!" Then for a few days it was, "Not overwhelming, but still cool!" By the end of the 2 weeks, though, we were happy enough to return it.
But yeah. It was like $3000 in total or something ‘~’
I find $60 for AAA titles a marginal proposition.
Wait 10 years and then ask.
It's brand new, genuinely computationally intensive, tech - it's going to be expensive. The Original PlayStation cost more in today's money than I bought my Vive for.
Also, if you can get a used Vive or similar there is almost no risk. If you genuinely don't like VR then maybe its not for you but I've played a lot different games and I can't think of anything that remotely compares with VR. Super Hot with a mouse is brilliant, Super Hot in VR makes me genuinely excited that I'm going to be around to see the near future. Beat Saber is very good fun, too (quite hard to describe but even in a vive the graphics are very crisp and smooth).
> If it was $200, I’d consider it.
Apparently half life alyx is enjoyable in a windows mixed reality headset. There is also the Oculus Go at that price point although I've never touched one because I have a proper VR headset.
ymmv but I've been able to buy used VR systems like the Vive Pro for $375 on ebay. Just to use them for either driving or flying simulators, VR is already worth it.
That said, there's still plenty to complain about for VR but that doesn't mean it doesn't have mass market potential.
I often said ‘when 3D TV’s don’t require glasses, I’ll buy one.’ I guess that was the general opinion, and when that didn’t happen quickly enough, it died.
Beat Saber is the most fun game I have ever played. No qualifiers. No shifty language. Nothing else is even in the same order of magnitude as Beat Saber. I dare to even call it a video game, because very few other games even possess the language to compete with it. The combat in FF7R is visceral, but you're not actually swinging Cloud's sword. The loop in Animal Crossing is entrancing, but it doesn't become your entire world like when you put a headset on. Even other VR games, which have that language they can use to tell their gameplay, often feel like the VR experience is bolted onto something which could, and maybe should, be non-VR. You often fight with the game; the game transplants you into a world, not just on a screen, but every degree surrounding you, yet you're limited in how you move, you're limited in how the game allows you to interact, nothing feels natural, and you soon adapt your actions with what the game wants.
I am solidly in the camp that, honestly, VR will not become a household item like a game console is. Even with headsets priced about at the level of a game console, like the Rift, the platform has too many downfalls.
Startlingly few games work well in VR. Even an experience like Half Life Alyx, the pinnacle of VR technology, feels like a shell of a game like Half Life 2. Its different. It can't tell a strong story, because players don't generally spend more than 30-90 minutes in VR at a time. It can't really innovate in puzzle design, because movement and viewport control is so heavy that it becomes tiring if you ask the player to do too much. The mechanics feel incredible. No one enters HL:A for the first time, loads that pistol, and doesn't have this HUGE GRIN on their face. But, its also a novelty; it wears off far too quickly once you begin realizing how many fundamentally core things Valve had to sacrifice to make the game work at all.
Nearly every game in VR is, bluntly, a tech demo. The platform is still figuring out if it is viable. Of the top selling VR games on Steam, #7 is a sim game which models a hyper-realistic anime girl, allowing you to mess with her clothes in a predictable anime way, and #10 is Skyrim. Of course, HL:A and Beat Saber are #1 and #2, but roughly this state is where VR was six years ago. It hasn't really evolved, and I'm not convinced HL:A was the catalyst I think Valve is hoping it is.
But, maybe I'm wrong. I will say this: There is no price you can pay for a VR headset that does not make Beat Saber worth it (within your means, of course). It is that good. I have an Index ($1000); beyond spending an hour or two in as many games as I can, 100% of my time is spent in Beat Saber, and I do not regret the purchase. Everything else VR has to offer is just icing. But this, a platform, does not make.
Space Pirate Trainer is another game that did a really good job of this. Before Beat Saber, that was easily my favorite, though wow Beat Saber is _so good_. Blarp is another game that did a great job of immersing you into a single room and making the most of the mechanics.
More "full" games don't work as well, because they have to figure out some way to transport you around the world without breaking immersion or causing nausea. And I don't think anyone has pulled it off well.
I am confident we will have more games that stand up to Beat Saber. It's a platform that requires discarding a lot of the strengths that video games have in order to work well, and developers have to re-learn a lot more from scratch than I think anyone realized when the platforms started.
I haven't tried the new Half Life, so I'm wondering if you're right.
(low-level spoilers)
There's one moment early on where you're asked to solve a vorgon puzzle in a drainage ditch. Every person I've asked who has been through this area disliked it, and I haven't found anyone who solved it legitimately. Its confusing. It has parallax/perspective elements which ask you to move around a large area, position yourself perfectly, and possibly even crane your neck in the right way to get an angle on what is the intended solution.
But, movement sucks in any non-room-style VR game. You can't use the joystick to just move your character, like you would in a non-VR game, because it causes unreal motion sickness in, like, 100% of users (this is available as option. its an experience that i think everyone should try; I had no idea I could get motion sickness until I did, but I was literally falling over). Instead, you press the joystick in the direction you want, a targeting beam pops up, and you "teleport" to the desired location. This sucks. Its imprecise, its slow, and it makes you feel, to some degree, like you're actually watching a powerpoint presentation of the game.
And, head yaw can also suck in VR. This may be surprising; its the entire point of VR. But, depending on whether your headset has a tether, and how big your play area is, many VR players are stuck stationary and with, maybe, 180 degrees of freedom in front of them. In one of these setups, you cant naturally look behind you; you have to rotate in-game, which calls back to all of the negatives that movement itself has. But even the in-game rotation is in-precise; in HL:A, it jumps by roughly 60 degree increments. Imagine there's something 45 degrees to your left you want to center; you can either turn your head 45 degrees left to center it (uncomfortable), rotate your body 45 degrees left (scary), or rotate in-game 60 degrees left, followed by rotating your head 15 degrees right or body 15 degrees right. In small play areas, the body rotations are a slippery slope that eventually ends in you being turned around, contacting furniture, or getting tangled in the headset tether. This is, at least for me, always front of mind; I can't escape the reality that my play area isn't large. So, you end up with these nearly (only nearly) subconscious calculations involving how many times you need to rotate in-game to make neck rotations less uncomfortable.
This is the problem space that any non-room VR game has, and will have, to solve. It is unsolved. I believe it is unsolvable. Even if you can get the player/object interactions down to perfection (this is insanely difficult, and different for every game; HL:A is the first "complex" VR game I've seen get close), even if you can craft a story and world that is both compelling, and works well with the intrinsic movement limitations VR has (no one has done this yet; again, HL:A gets close), even if you can remove slideshow-teleportation movement without motion sickness (I believe this is impossible without something like an omni-directional treadmill); you still run up against most players just not having the space for room-scale VR & N-degree freedom of movement.
VR is filled with technical challenges, and every technical challenge developers have to solve limits their ability to be artists. This is why arcade-style games are the showcase piece for VR. And, in my opinion, will remain that way for a good, long while.
Aren't tethered headsets a thing of the past now? I have a Quest and never had that issue.
The Quest does two things that are interesting, and rather unique from a full market perspective; local rendering, and inside-out tracking.
Local rendering is probably the future for VR platforms. Unfortunately, there's really only one company on the planet with a mobile SoC design capable of driving the performance necessary for truly current-generation VR experiences, and that company is not interested in sharing their technology (Apple). The Quest, with its android smartphone-class specs, is not, in my mind, a current-gen VR headset from a performance perspective. It has other (huge!) advantages which make it an awesome purchase (generally, the best purchase for most people), but performance is not one of them.
I think the jury is still out on inside-out tracking. It may approach a level where it is good enough for most use-cases, while still being worse than outside-in tracking is today. Arguably, its there today. Fundamentally, if your environment is dark, or lacks the many visual intelligence landmarks necessary for the AI to operate, its going to struggle, not to even mention the challenges it has in tracking hand positions behind your back (which is very common in FPS-type games, where you often reach into your backpack to grab more ammo).
Plus, the local rendering, the inside-out tracking, and internal battery, all add weight to the headset, which already cheaps out on the circumcranial support straps to properly distribute weight (many people swap their Quest straps with, say, those from a Rift S or even an Index/Vive, because they're so much better). The Quest is both one of the most and least advanced headsets on the market.
I've heard streaming to the Oculus Quest is basically like owning a Rift. I haven't tried it myself though.
My basic take is that the technology is really impressive; I agree with proponents that the Quest finally nails it technically. However, I don't think it's enough better than other options to make it worth the significant drawbacks. Something being immersive is ultimately a function of the player experience, not the hardware. And as is demonstrated to me whenever I try to get the kids to turn off almost any game, non-VR hardware is definitely immersive.
If anything, facehugger VR seems less immersive, because the stuff is way less comfortable than other gaming hardware. The kid who loved rhythm games would play Beat Saber less than non-VR games because of neck discomfort and face irritation. Battery life was also lower than the Switch, and of course infinitely lower than the Playstation. Which also cut into immersion.
I did really appreciate the new 6-axis controllers; they were very cool, and gave a great sense of spatial freedom. But unfortunately, that freedom is a lie; my sense of immersion quickly ended every time I hit a virtual boundary. To successfully play Superhot, I found I had to continuously maintain two spatial frames of reference: VR and reality. So it ultimately felt less immersive to me than a console game, where I could just plop down on the couch and forget my body entirely.
Same experience (Vive Pro) with niece, nephew and their friends. They tried Beat Saber and all other age-appropriate VR games. After those few times, not once did they ask to try again. They would much rather play Fortnite and The Sims, with a mouse and keyboard.
The entire concept of the game (time only progresses while you're moving) fits _so well_ with VR. It makes you feel like a complete badass as you're able to perform ninja-like feats of dexterity and dodge bullets all the while being a just challenging-enough puzzle and providing a decent workout.
Everything else I've played was either basically a tech demo or felt like our old games built within the limitations of VR (what really differentiates Beat Saber from DDR, Guitar Hero, etc?). SUPERHOT was one of the few games I played that really felt like a proper VR game.
“Go spend a few hundred to prove the value” is a hard sell.
But for VR? It is a niche, for sure, but I know a few people really enjoy it.
It can be a niche forever (like high end sound equipment) or it can grow into mainstream (like TV), I have no idea, but I doubt it will fade away.
Apparently Quest sales figures aren't public, but I see estimates that they sold a million units last year. Nintendo sold that many Switches last month. And the month before that. And the month before that. (And in December it was something like 5 million/month.)
However much niche users enjoy it, I have a hard time seeing how they'll sustain enough of a developer ecosystem to keep them entertained. High end sound equipment can draft off of every music player in the world for content. But good VR is very much VR-specific.
Valve, like a lot of other companies, hopes that VR will become a thing, so they're making a headset. Because they'd like to sell headsets, they took a major property and made a VR version. If they lose money on their VR version of Half Life, they're fine with that.
Some game makers will surely try VR versions as well. This could be a big wave, and they want to be ready, so it's ok to lose money. But if unit sales aren't there, they're not going to keep doing it. In which case, things could well end up just like every other attempt at VR going back to Nintendo's 1995 effort. Or like the various waves of 3D movies and TV, where everybody's really excited for a hot minute and then people stop caring.
I tried VR back then, and I tried VR now. It sounds like you haven't tried it now? I've seen people crying because they couldn't believe how crazy it was. Just give it a try!
I have tried it a number of times, including the current generation. https://news.ycombinator.com/item?id=23048814
but 3D TV and movies were pretty bad... It was always a gimmick. The experience you get in VR is nothing like that.
Quite bullish that the next generation will be absolutely incredible. I bought an upgraded dev machine to try out VR dev because of the Quest -- it seems like it will be the next mass-computing interface.
That being said, I use my VR headset probably even less than my 3d printer. As mindblowing as it is when I put the headset on (and truly, it never gets old), I realized it's just not for me.
I usually play games on my sofa, with a steam controller and the screen on my wall through a projector. I always play turn by turn or low intensity games rather than fast-paced action games. I have my laptop on front of me on my coffee table, and constantly check my mailbox, rss feeds and many sources of news, on the laptop between playing two turns in the games or message with friends on my phone.
This is a funny thing, because I didn't realized that before I bought a VR headset : basically, I'm always multitasking during my leisure time, and you just can't do that in VR. Each time I think "hey, let's chill in No Man's Sky", my immediate reaction is "oh wait, but if I do that, I can do _only_ that", and I play a classic game instead.
Now, this is a very personal experience, maybe that won't be a problem for anyone else. But as far as I'm concerned, VR fullfills all its promises, and yet it's not for me.
EDIT : fun fact, the time when VR is really useful for me is during week-end during the day when I don't feel like working on a side project or learning a new thing. I can't use the projector with daylight, so in those cases I'll just put the VR headset on rather than closing shutters.
For me it's the multiplayer games that I like, more of my friends are getting VR headsets right now and I suspect that as more of them get into it the more we'll end up playing VR together.
People underestimate the amount of applications for VR. But Mark my word, in a few years your comment will have poorly aged as every household will have a VR headset the same way every household had a printer or a desktop PC 20 years ago.
But yeah, VR requires my full and undivided attention, and this will just not happen :)
It's the same for me and gaming in general, I have very little patience except if I'm playing warzone or rocketleague with friends. It's just not the same :)
I wrote about my mind-blowing experience playing the Catan board game in VR: https://p1x3l.com/story/239/social-virtual-reality-and-the-o...
Not quite. From Nvidia: "Apple fully control drivers for Mac OS. But if Apple allows, our engineers are ready and eager to help Apple deliver great drivers for Mac OS 10.14 (Mojave)...Unfortunately, Nvidia currently cannot release a driver unless it is approved by Apple." — https://en.wikipedia.org/wiki/Nvidia#Apple/Nvidia_web_driver...
Apple is preventing Nvidia from releasing Mac Pro and eGPU driver support.
I’m kind of curious though: what value do their web drivers have to them now that they couldn’t just release them as OSS and allow a determined group to rewrite them as Metal.
Nvidia's eGPU web drivers already supported Metal in High Sierra [0]. Apple refuses to sign their drivers for Mojave and up. Without Apple's signature, GPU drivers will not run on macOS, OSS or not.
[0] https://www.insanelymac.com/forum/uploads/monthly_2020_03/84...
I know there are some older Nvidia cards that do work on Mac, but most of the recent GTX/RTX/etc. ones don't. Are they similarly Metal (and just missing a signature)?
Because whenever I was actually keeping up with this, the stalemate was over Nvidia not leveraging Metal.
Also, to answer your question: no. Nvidia is in the GPU selling business. They wouldn’t mind having their high priced GPUs in high priced laptops. Apple keeps control of what hardware they sell, which dictates what software gets developed. Why would Nvidia or Apple spend resources to make software for a computer that no one is selling? It’s Apple who refuses to buy rather than Nvidia being unwilling to sell.
Game consoles never had OpenGL support, other than a timid attempt from Sony with GL ES 1.0 using Cg as shading language.
That’s the problem they’re trying to solve, it’s not about cross platform.
You might ask, ‘Then why is OpenGL deprecated?’. The answer is that Apple is just deprecating it’s own support since there’s a more low level API, OpenGL should be implemented on top of Metal, which (for example) MoltenGL is.
I'm not going to spend valuable time and resources on a non-standard graphics API for a platform with anaemic graphics hardware. Instead, I'm going to target Windows and Linux with Vulkan or OpenGL. If I do support MacOS, it will be via MoltenVK, and it will be a second tier platform.
From my point of view, by dropping standard OpenGL they have turned their platform into a second-class citizen.
OpenGL and Vulkan only happen to work on Windows via the ICD driver model introduced in OpenGL 1.1, which Microsoft hasn't yet bothered to remove.
Instead they aren't supporting it on UWP and Win32 sandbox model, which only allow DirectX.
Microsoft's way forward is also to build OpenGL wrappers on top of DirectX like ANGLE.
https://github.com/microsoft/angle/wiki
https://devblogs.microsoft.com/directx/in-the-works-opencl-a...
And then there are the game consoles.
Khronos has just messed up providing SDKs that are at the same productivity level as commercial APIs.
I also don't care about Microsoft's UWP and sandboxing stuff.
They might have a desire to take the platform in that direction. But I have no obligation to follow, and if they want to retain application developers they won't be able to simply cut us off. We've always got the option of abandoning Windows entirely if they make it too inconvenient and troublesome to develop for. It's not like we don't have a precedent for abandoning MacOS X as a top tier platform for exactly those reasons.
"How Windows 10X runs UWP and Win32 apps"
https://www.youtube.com/watch?v=ztrmrIlgbIc&list=PLWZJrkeLOr...
MSIX
https://docs.microsoft.com/en-us/windows/msix/overview
> Apps that are packaged using MSIX run in a lightweight app container.
Yes, Win32 can still choose between both models in current versions, depending on Windows 10X reception, I bet it might eventually change.
Sandboxing resistant developers are always free to keep their Windows XP or 7 copy, I guess.
The problem is that it's unique and different. The problem is that I (as an open source developer for a popular gaming-related project with an OSX port) can't realistically develop and support a metal backend without having to buy an expensive mac with a good modern GPU.
And with OpenGL being deprecated on Mac, with old and (sometimes) broken drivers that are missing 10 year old features, along with apple breaking things and adding new signing/notarising requirements, it's hard to justify continuing to support mac.
Same kind of thing goes for Nvidia/Apple. It doesn't matter who's fault it is for starting that feud, it's still hurting the users.
> The problem space Vulkan and Metal solves is not high level cross platform graphics (which OpenGL tries to solve), it’s low level control over the GPU. That’s the problem they’re trying to solve, it’s not about cross platform. You might ask, ‘Then why is OpenGL deprecated?’. The answer is that Apple is just deprecating it’s own support since there’s a more low level API, OpenGL should be implemented as a library on top of Metal, which (for example) libraries like MoltenGL is.
iOS is an exception due to it's position in the market but I've been seeing less and less interest in mobile overtime too
Pretty much Mac owners are on laptops. Even in the PC world few laptops can power VR. Those that do tend to be bulky for cooling and have nVidia GPUs; both things which MacBooks seem unwilling to have.
Having a extensible desktop at 1k that supports nVidia would probably go a long way.
You're not going to get very far without drivers an you'll be stuck on OpenGL 4.1 and Metal. That and you'll have the problem that there are less games built for MacOS than Windows an Linux + Proton.
Investing in a GPU for MacOS for gaming would be a huge waste of money IMO.
Already a terrible experience.
Really, this experience actually sounds pretty pleasant.
----
¹: https://www.asus.com/Graphics-Cards-Accessories/ROG-XG-STATI...
From Apple's perspective, that's a machine with razor-thin margins. Why would they want to sell that when all of their other products have enviable margins?
Moreover, such a machine would cannibalize their high-end Mac Pro sales. After all the money they spent to develop it, having your pro customers scoop up the thin margin gaming machines seems like a poor strategy.
The market for mid- to high-end desktops is indeed lower margin, but it's a much bigger market overall. The old Mac Pro was at the high end of this market. Expensive, but justifiable if you needed it. If Apple had the will, they could be profitable in this space. They have chosen to exclude themselves from it.
Back in the mid '00s, I saw Mac Pros in use in scientific research settings, because they allowed use of lots of RAM compared to equivalent PC mainboards of the time, could take lots of storage, and had lots of horsepower to throw at bioinformatics and modelling problems. We had compute clusters as well, but for some problems these were a better fit. Today, Macs are banned. The academic funding bodies don't consider them good value for money. And they are not wrong in that assessment.
Don’t start with system requirements first and make a game targeting that because you want to show everyone how elite a hacker you are.
If Sony didn’t require games achieve 60FPS at 1080p on PlayStation 4 before they’d allow them on the platform, most of the developers would be targeting PS 4 Pro because they want to show off. I expect that’s mostly the case anyway, and these developers have to be dragged by their corporate masters kicking and screaming back to baseline PS 4 support.
Game developers who want to reach a broad market and are willing to meet their deployment platforms halfway do not say this. They just port and ship.
What kind of impact will that have on the ecosystem?