@vlads, please accept my apology for causing any offense. I really, very much, appreciate you, Rob and the entire Corona SDK community.
I know you are extra busy, so please take your time in responding.
Hi Troy,
This really is all explained very clearly on the blog and in this thread, but…
Porting to Metal is not necessarily a big hurdle - and not the dead end people make it out to be
Corona will change into a program that run locally to build as well as simulate. At the moment simulate is local, but build is on the Corona servers. Since these servers will stop you will be able to build on your own computer. This is not ‘native’, it’s just building local instead of building in the cloud.
@troylondon
you should take some time and read all the posts of this thread and the truth will reveal itself :P.
Depending on Apple’s decisions Corona should go Metal sooner or later. Vlad can and probably will do it in time.
Offline builds are coming soon.
Long story short:
There are many people with business depending on Corona, many hobbyists who like Corona and many people who make their apps with Corona and don’t want to change platform. All of them will support financially Vlad and Rob to keep working.
It will be like Phase 1 of MCU. In Phase 2 more developers will come, new users and Switch support :P.
This scenario sounds too good to be true but who knows?? 
Edit:
thomas6 beat me to it. What he says too :).
Glad to see you back on the topic.
Sadly I’m very short on time atm - but I still like to offer some infos/ideas/urls just in case it helps in any way.
You very probably know about https://github.com/bkaradzic/bgfx
But have you evaluated it as a basis instead of Vulkan/Molten? Advantage would be that someone else already did all the painful thinking about unifying lowlevel rendering frameworks.
What about simple SDL? It’s used as core platform and rendering abstraction, including consoles (hint Switch etc.).
Next one would be Sokol https://floooh.github.io/archive/ which may not be stable enough as a basis but given Andre seems to be a very pragmatic dev there’s a decent chance to get some valuable info from his last few blog postings.
And in case it’s still going to be Vulkan in the end, here’s a posting that might be of some value too https://zeux.io/2020/02/27/writing-an-efficient-vulkan-renderer/
Currently everything is on OpenGL ES. Switch seems to be fine with it. It just Apple throwing feats.
On a side note Micheal mentioned something a few weeks ago (3 weeks) that I didn’t know existed but reminded me of something that I did know (and very similar to what Micheal mentioned). Google Angle. I am familiar with Angle and I spent a few hours over that weekend seeing if I can make it work with Corona for iOS Metal specifically. I got very close to having it working (most of the time was spent compiling the beast). I had some reference errors that I ended up fixing with a terrible hack that would need to be cleaned up. I would still prefer a native implementation of Metal or Vulcan, but Angle is possible.
@Rob, count me in for a monthly Patreon pledge. And I’m so glad to hear Vlad is still moving forward. Will you be continuing on with Vlad as well?
Might I recommend you change the message at the top of this thread? Perhaps start it with…
“The future of Corona SDK is changing. CoronaLabs is going away as a company, but not until we finish the open-source tools. Further, the developer and I will be supported by developers like you through a subscription service (i.e. Patreon) - details we are still working out. We have big plans for the future of Corona SDK, but now the burden for our support will make us dependants upon those who rely upon Corona SDK and seek to have it continually grow. For information about supporting us, email me at ____”.
This, IMHO, would be received better than the current post and reduce the potential damage caused to the community by developers seeking to migrate away from Corona SDK.
Also, I’d recommend sending out an email in this more positive light to all your Corona SDK developers. You may be surprised by the financial support offers you receive.
Just my two cents.
@agromonte, will you consider starting a thread about Angle and how it is a potential solution to support Metal for iOS?
I wouldn’t know what to say, but of course you can start one. It is similar to what Micheal linked a while back. It is abstraction layer.
Here is all I know:
Angle is the one that chromium uses or use to use to run on almost anything. It is in most cases a OpenGL drop in replacement. And that is all I did in my test. It took me a few hours to compile (compiled the wrong version first) . Then I searched in the corona code for all the OpenGL calls (basically 3 projects if I remember correctly) and I replaced them with the similar or exact angle calls. As I said before ran into some problems with linking because of a reference to a library inside a library that I couldn’t find (enter the ugly hack). Then I ran a test with my apps to see if it would work and that it could capture metal frames (so I know it was running in metal). I also downloaded the skateboard free template from PonyWolf (figured it would be more taxing than my crappy casual games). Added a counter on the top to display frame rates and ran it twice. Once from the simulator build (openGL) and once from the testapp folder build (angle) on my iphone 7. Now you know as much as I do and the extent of my testing and hacking.
Wow, this looks great! I recognize the username from a couple GitHub projects, but hadn’t dug deeper into his stuff.
I did stumble across Sokol. Philip Rideout cited it as an inspiration, though I forget where. I soon got distracted by all his embeddable 8-bit computer emulators, though. :) (On that note, I think porting this from SDL might be a fun stream idea. A correct implementation of this sort of library involves non-file-based audio, lack of which has stymied a few ideas, so I’m thinking I’d have a go at that and keep these on hand for testing. Speaking of SDL, the audio bits do offer “rwops”, so this might even be straightforward.)
Sascha Willems is another on my radar whom I’ve yet to give much attention.
bgfx is one of those projects I keep meaning to really explore. I’ve never been able to get a feel for it by just perusing the samples. It probably wouldn’t hurt to give multiple backends a try.
Also, no worries about time and all the rest. Ideas and information are very much appreciated. Best of luck on everything else. :)
@ agramonte One version of it is actually in the Corona codebase. The comments suggest it was used for Windows Phone.
Can we move this discussion about Vulkan, Metal, etc. to a new thread? Let’s keep this thread focused please.
Rob
@Rob, we pay $55/mo for the Adobe Creative Cloud suite of products. I’d be happy to spend that same amount to help ensure a positive future for Corona SDK. I would consider paying even more, considering the amount of money keeping Corona up-to-date will save us as our apps get continually updated for the next few years. With Patreon, are you limited to whatever people are willing to pay, or can you require or suggest a monthly support amount? Is this the right place to discuss this or do you want to start another thread? Just let me know. I appreciate you all.
Same here. 50 euros a month, no problem!
I wonder: Is there a specific ceiling? …
How much do they need for a single year?
Does vlads and Rob need someone else’s inclusion?
@w967 what do you mean “specific ceiling”? If you mean to the GitHub match? Yes, it’s something like $5,000 which, based on current pledges, is close to being hit. However the cost to continue Corona development is much higher.
I would suggest those studios that are big enough contributes more. I know that many of the tools out there - eg. analytics, attribution, backend services, etc easily cost over a thousand a month. I would argue that Corona is more important than most of these tools, thus should be valued more.
@akao, analytics is free, attribution is free for organic installs (appsFlyer does charge for paid installs) and most backend services have a free indie tier that is suitable for most small devs.
[quote name=“Rob Miracle” post=“408330” timestamp=“1584654342”]However the cost to continue Corona development is much higher. [/quote] I think this is one of things people are trying to understand. What exactly is much higher? Are there going to be specific fundraising goals? Fundraising to an unknown number is a bit hard.
There are 11 pages prior to this to wade through so excuse me if any of this is redundant:
Every thing should be in same place, or accessible through same palaces. Simulator builds would still work etc. site and forums would remain, but current decking etc would be in new place, with links on coronalabs.com clearly visible