Mac New Game Porting Toolkit is Wine

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
I skimmed through it and it seemed very confused to me.

He says Apple needs people play games on the Mac so that game devs make games for the Mac and the best way to do that is open source GPTK which will cause game companies to open source their game source code which the “community” will then use to port the games….hmmmm.

I’ll watch it properly to see if I’m being unfair but uhhhh

I mean, if he doesn't immediately turn around and say that's totally unrealistic and will never happen, he's naive at best.

Naive is what I think. In what I've watched (not this video but others), Tsai seems like a young guy who's taught himself the surface level of game porting by playing around with the GPTK, and he's really excited about it. He was able to lead the way on "porting" a bunch of games! Of course he hopes it can become a community thing so he can be part of this fun new hobby. He probably doesn't understand that things get 1000x more difficult and technical in the post-GPTK phase of making a native port, so nearly all "community" members (probably including himself) might not have the ability to meaningfully contribute. He also doesn't have any feel for how the owners of game intellectual property think if he believes they'll just open source their stuff.
So I just watched it and it is a little more nuanced. Basically he’s proposing to at least further opening the licensing of GPTK/D3Dmetal so companies can formally use it for releasing non-native ports and then he also proposes open sourcing GPTK/D3Dmetal which would ensure more rapid improvement of the tools and wider compatibility.

He does acknowledge that native port development could take a backseat if developers think they don’t have to bother with native development, just use compatibility layers and they’re done. However, he believes once a Mac gaming market is proven to exist then that will itself encourage developers to make native ports. He does not think the open source community would aid in the development of native ports or that companies would open source their engines.
 

Jimmyjames

Site Champ
Posts
590
Reaction score
637
So I just watched it and it is a little more nuanced. Basically he’s proposing to at least further opening the licensing of GPTK/D3Dmetal so companies can formally use it for releasing non-native ports and then he also proposes open sourcing GPTK/D3Dmetal which would ensure more rapid improvement of the tools and wider compatibility.

He does acknowledge that native port development could take a backseat if developers think they don’t have to bother with native development, just use compatibility layers and they’re done. However, he believes once a Mac gaming market is proven to exist then that will itself encourage developers to make native ports. He does not think the open source community would aid in the development of native ports or that companies would open source their engines.
Fair enough. Good to know.

Apart from my disagreements with that strategy, which we discussed before, I think the problem is GPTK has ok performance, but nowhere near as good as Proton on Linux, due in no small part to the architectural differences between Apple Silicon and traditional PCs. I don’t think ports with really mediocre fps is a great look for Apple.
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
Fair enough. Good to know.

Apart from my disagreements with that strategy, which we discussed before, I think the problem is GPTK has ok performance, but nowhere near as good as Proton on Linux, due in no small part to the architectural differences between Apple Silicon and traditional PCs. I don’t think ports with really mediocre fps is a great look for Apple.
That’s true the Steam Deck doesn’t have to deal with x86 to ARM translation so even non-native (non Metal) ports would at least need to be recompiled for ARM to avoid that additional penalty, especially in CPU limited scenarios. However if ARM Windows chips from Nuvia/Nvidia/AMD take off that might be less of a concern anyway in the future.

The title is clickbait but unfortunately as discussed on here multiple times the YouTube algorithm necessitates that somewhat.

I’m not sure if Apple needs to take Steam’s approach which essentially what he’s advocating for but I don’t think it would be the worst thing if they did. I think Apple’s current approach could, if you’ll forgive the expression, bear fruit. Either way it’ll be a slow process and no one should expect an overnight deluge of native Mac games. Some patience is in order.
 

Jimmyjames

Site Champ
Posts
590
Reaction score
637
That’s true the Steam Deck doesn’t have to deal with x86 to ARM translation so even non-native (non Metal) ports would at least need to be recompiled for ARM to avoid that additional penalty, especially in CPU limited scenarios. However if ARM Windows chips from Nuvia/Nvidia/AMD take off that might be less of a concern anyway in the future.

The title is clickbait but unfortunately as discussed on here multiple times the YouTube algorithm necessitates that somewhat.

I’m not sure if Apple needs to take Steam’s approach which essentially what he’s advocating for but I don’t think it would be the worst thing if they did. I think Apple’s current approach could, if you’ll forgive the expression, bear fruit. Either it’ll be a slow process and no one should expect an overnight deluge of native Mac games.
Oh I think it will be a few years. I’m not expecting it to improve massively in a year or two. I’m looking to see what happens this year. Last year we had what…around 15 to 20 good games. 2022 saw about 4 or 5. If we get 25 to 30 this year, I’d consider that an upward trend. Ultimately it depends on how games like RE4 are selling.
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
Oh I think it will be a few years. I’m not expecting it to improve massively in a year or two. I’m looking to see what happens this year. Last year we had what…around 15 to 20 good games. 2022 saw about 4 or 5. If we get 25 to 30 this year, I’d consider that an upward trend. Ultimately it depends on how games like RE4 are selling.
Yeah his video definitely seemed a bit impatient to me. Regardless of what Apple does, this is going to take awhile. I suppose from his perspective though at least more games would be playable sooner even if they aren’t at their full native potential. I dunno. I feel like for those of who can use the current tools to get these games working, great! But let’s see if Apple can attract more native development right off the bat and the Proton-solution becomes a viable fallback if that doesn’t work.
 

Jimmyjames

Site Champ
Posts
590
Reaction score
637
Yeah his video definitely seemed a bit impatient to me. Regardless of what Apple does, this is going to take awhile. I suppose from his perspective though at least more games would be playable sooner even if they aren’t at their full native potential. I dunno. I feel like for those of who can use the current tools to get these games working, great! But let’s see if Apple can attract more native development right off the bat and the Proton-solution becomes a viable fallback if that doesn’t work.
I’m not sure how I understand how releasing gptk as open source helps gaming as opposed to letting users utilize Whisky/Crossover to play those games now. I’m not convinced these are technical problems. We currently have lots of apps and games where the devs could release their iOS app for the Mac by ticking a box…and they don’t do it. If they don’t think it’s worth their while financially to tick a box, why would they spend time modifying their game to optimize for gptk?

A bigger problem is this solution does nothing to help Apple’s strategy of getting AAA games on the iPhone/iPad. I believe Apple wants games not just on the Mac, but on all their devices. Iirc, they explicitly say the best way to port your game to the iPhone is by starting on the Mac. That’s the idea: use the popularity of the iPhone as a crafty way of getting games on the Mac as well. Proton does nothing for Apple’s most profitable platform.
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
I’m not sure how I understand how releasing gptk as open source helps gaming as opposed to letting users utilize Whisky/Crossover to play those games now.

The idea is to first grow the user base of people gaming on their Mac. His assertion is that releasing GPTK as open source would improve on the current situation because it’s basically a better Wine with Apple + Codeweavers + the community filling in the gaps and bug fixing. This would ensure a wider array of games, including newer games would work.

I’m not convinced these are technical problems. We currently have lots of apps and games where the devs could release their iOS app for the Mac by ticking a box…and they don’t do it. If they don’t think it’s worth their while financially to tick a box, why would they spend time modifying their game to optimize for gptk?
Oh he agrees that’s the problem. So his solution by going full proton is to rapidly increase the library of possible Mac games and increase the known Mac player base making it a more attractive target for developers. The problem of course is that Apple doesn’t operate a cross Windows Mac platform store like Steam so they can’t just go through and mark games as GPTK ready. Heck I remember when a bunch of iOS developers were pretty unhappy when Apple made the default for iOS apps to be made available to Apple Silicon Mac users. So again I’m not sold that this is the best first step for Apple to try.

A bigger problem is this solution does nothing to help Apple’s strategy of getting AAA games on the iPhone/iPad. I believe Apple wants games not just on the Mac, but on all their devices. Iirc, they explicitly say the best way to port your game to the iPhone is by starting on the Mac. That’s the idea: use the popularity of the iPhone as a crafty way of getting games on the Mac as well. Proton does nothing for Apple’s most profitable platform.

By virtue of being the far more profitable platform the iPad/iOS needs less help being attractive to develop for. If you can make the Mac attractive to develop for, developers will want their stuff work on any iPhones/iPads that can run their stuff. Basically the proton solution is to use Proton as a stop gap to grow the user base to then make native development attractive. It’s the chicken and egg problem: developers don’t want to develop for a user base that’s too small and users won’t game on a system with too few games. So jumpstart the process with great hardware and non native games.

Personally I think the middle of the road approach that Apple is taking enticing developers with money, support, and advertising combined with GTPK to help fill in gaps for more advanced users/gamers is probably the right approach with going full proton as a backup option because, as I said, Apple doesn’t operate a massive store with Windows games, like Steam, something Andrew overlooked.
 

Jimmyjames

Site Champ
Posts
590
Reaction score
637
The idea is to first grow the user base of people gaming on their Mac. His assertion is that releasing GPTK as open source would improve on the current situation because it’s basically a better Wine with Apple + Codeweavers + the community filling in the gaps and bug fixing. This would ensure a wider array of games, including newer games would work.
I don’t understand how open sourcing GPTK grows the user base more than the current situation does. How does the community fix bugs for games they don’t have the source code for?
Oh he agrees that’s the problem. So his solution by going full proton is to rapidly increase the library of possible Mac games and increase the known Mac player base making it a more attractive target for developers. The problem of course is that Apple doesn’t operate a cross Windows Mac platform store like Steam so they can’t just go through and mark games as GPTK ready. Heck I remember when a bunch of iOS developers were pretty unhappy when Apple made the default for iOS apps to be made available to Apple Silicon Mac users. So again I’m not sold that this is the best first step for Apple to try.
I think we’ve discussed this before and I still don’t see how this gets native games on the Mac. According to steam Linux has more gamers than the Mac and yet no one is making games for Linux.
By virtue of being the far more profitable platform the iPad/iOS needs less help being attractive to develop for. If you can make the Mac attractive to develop for, developers will want their stuff work on any iPhones/iPads that can run their stuff. Basically the proton solution is to use Proton as a stop gap to grow the user base to then make native development attractive. It’s the chicken and egg problem: developers don’t want to develop for a user base that’s too small and users won’t game on a system with too few games. So jumpstart the process with great hardware and non native games.

i don’t believe that grows the base as much as making the iPhone and iPad possible targets for AAA games.
Personally I think the middle of the road approach that Apple is taking enticing developers with money, support, and advertising combined with GTPK to help fill in gaps for more advanced users/gamers is probably the right approach with going full proton as a backup option because, as I said, Apple doesn’t operate a massive store with Windows games, like Steam, something Andrew overlooked.
That seems reasonable.
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
I don’t understand how open sourcing GPTK grows the user base more than the current situation does. How does the community fix bugs for games they don’t have the source code for?

It’s not fixing bugs in the games or at least not those kinds of bugs. Basically as I understand it a lot of big AAA titles, especially with custom engines, ship fundamentally breaking the graphics API in some way. The day0 patches from AMD/Nvidia is to provide bandaids in the drivers to work around the games’ issues. Wine and GTPK have all sorts of tools and game specific code to get things workings in the same way. Basically the more eyeballs looking through the compatibility layer’s code and the more people able test different games, the more likely and the more quickly solutions are to be adopted to ensure Wine/GTPK is able to run the games without glitches that don’t happen on the Windows side. Hell for some older titles it’s actually less of a headache to run the games through Wine than modern Windows.

I think we’ve discussed this before and I still don’t see how this gets native games on the Mac. According to steam Linux has more gamers than the Mac and yet no one is making games for Linux.

Again early days and it’s still small. Steam deck was released Feb 2022. It’ll take years for either Valve or Apple to see large movement of developers if it’s going to happen, especially if we’re going to measure by native ports actually released.

i don’t believe that grows the base as much as making the iPhone and iPad possible targets for AAA games.

Oh I agree. That’s why I don’t think Apple’s approach should be the same as Valve’s. I think Valve is doing the best approach for their business model while I think that should be Apple’s back up plan. Apple already has the draw of iOS and iPadOS but doesn’t have the developer relationships and huge Windows games store of Valve. Different platforms, different business models, different business strengths and history, different business plans.

That seems reasonable.

Well I’m a very reasonable guy. 🙃 All the voices in my head agree … 😬
 
Last edited:

quarkysg

Power User
Posts
64
Reaction score
36
Apple’s main objective is to sell Macs and grow macOS with GPTK. Regardless of how well it is able to run DX11/12 games, nobody in their right mind will be buying Macs just to run Windows games. Why not just take Apple’s words as it is in that GPTK is there to help with games porting.

Looks like Andrew Tsai is letting his enthusiasm cloud his analysis,
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
A relevant article:


About half way through it, seems good so far.

==========

On a separate note: One hardware thing Apple can improve for gaming is its displays. Now Apple’s displays are top notch (no pun intended) in most respects but pixel response times is an exception. They are apparently quite poor in that respect relative to other displays, especially gaming displays, and for certain types of games like fast FPS the motion blur can be quite bad.
 
Posts
11
Reaction score
24
Location
Brisbane, Australia
A relevant article:


About half way through it, seems good so far.

==========

On a separate note: One hardware thing Apple can improve for gaming is its displays. Now Apple’s displays are top notch (no pun intended) in most respects but pixel response times is an exception. They are apparently quite poor in that respect relative to other displays, especially gaming displays, and for certain types of games like fast FPS the motion blur can be quite bad.
The slow response time is a byproduct of the local dimming/keeping colour accuracy over the entire brightness range.
When they switch the MacBook panels to OLED, that issue should go away.
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
The slow response time is a byproduct of the local dimming/keeping colour accuracy over the entire brightness range.
When they switch the MacBook panels to OLED, that issue should go away.
Oh? Do you have a link for that? Not that I disagree it’s simply not something I know much about. Wouldn’t mind reading more.
 

mr_roboto

Site Champ
Posts
262
Reaction score
418
Oh? Do you have a link for that? Not that I disagree it’s simply not something I know much about. Wouldn’t mind reading more.
No link, but there's an inherent slowness to LCD technology. To change the color of a pixel, first a different electrical field is applied to the pixel site, then the liquid crystal material needs some time to reorient itself to the new field. There are various techniques for making this physical movement faster, but this is not a technology where you can have your cake and eat it too. Apple has traditionally prioritized color spectrum, accuracy, and off-axis viewing in their LCD panel designs, so they end up with relatively poor switching speeds.

OLED is an emissive technology: run current through a LED and it emits photons. This is essentially instantaneous, so it's inherently very very fast. But...

When they switch the MacBook panels to OLED, that issue should go away.
I'm not sure Apple's in a rush to go OLED for anything bigger than iPhone. OLED still has fundamental problems - organic LED materials degrade too fast with use compared to inorganic LED chemistries. OLEDs also aren't as power efficient as inorganic. These problems can be managed on watches and phones, but are difficult at notebook panel sizes with higher expected duty cycles, especially if you want to push peak brightness way up for high-gamut HDR.

Apple's strategy so far has been to invest in developing their own mini-LED displays for notebooks and tablets. Because mini-LED backlighting for a LCD panel uses inorganic LEDs, they get more light for the same power (albeit then have to lose some of it to the inefficiency of LCD), and are able to drive the LEDs hard for HDR without nearly as much worry about longevity.

There's been all sorts of leaks and rumors about Apple working on their own micro-LED displays for the future, supposedly for the Apple Watch first. If this pans out I expect them to roll it out across the entire product line; if they can make micro-LED work it would be a game changer. (This is a technology in which you manufacture individual "micro" AKA pixel-sized inorganic LEDs, and assemble a lot of them into a display panel. Needless to say, there are challenges in economically assembling millions of LEDs into a big display panel.)
 

dada_dave

Elite Member
Posts
1,975
Reaction score
1,962
No link, but there's an inherent slowness to LCD technology. To change the color of a pixel, first a different electrical field is applied to the pixel site, then the liquid crystal material needs some time to reorient itself to the new field. There are various techniques for making this physical movement faster, but this is not a technology where you can have your cake and eat it too. Apple has traditionally prioritized color spectrum, accuracy, and off-axis viewing in their LCD panel designs, so they end up with relatively poor switching speeds.

OLED is an emissive technology: run current through a LED and it emits photons. This is essentially instantaneous, so it's inherently very very fast. But...

Thanks!
I'm not sure Apple's in a rush to go OLED for anything bigger than iPhone.



There's been all sorts of leaks and rumors about Apple working on their own micro-LED displays for the future, supposedly for the Apple Watch first. If this pans out I expect them to roll it out across the entire product line; if they can make micro-LED work it would be a game changer.

This is the impression I get from the rumors as well. However, it’s been awhile and I’m unsure what the current state of affairs is.
 
Top Bottom
1 2