I admit that my knowledge of 3D graphics standards, such as OpenGL, of which MorphOS supports a small amount of with something called "TinyGL", is very weak, and my understanding of programming in general is not much better (or is it worse?), so forgive me if any of these questions don't make any sense, or are very naive.

I'll assume that supporting a subset of OpenGL was chosen because its documentation is public knowledge, and DirectX versions probably are not (please correct me if I am wrong). I'll also assume that TinyGL was created, because of some limitations of MorphOS which prevent, or make it difficult to completely support even some of the older versions of OpenGL. Or is TinyGL limited because the work to support a complete version of OpenGL just too much work for the one or two developers who probably created TinyGL?

Is there likely to be any updates to TinyGL, or is it more likely that new graphic improvements for MorphOS will come in the form of SDL updates (I know nothing about SDL either)?

Is there zero chance that MorphOS will ever get any type of DirectX support, that would allow old games to be ported to MorphOS (like DirectX v9, or even earlier versions)? The one game I currently enjoy playing uses DirectX v12 I believe, it's an old 2012 game that used an even older game engine I think.

I know that the programmer (Daniel, can't remember his last name atm)who has ported some recent games to AmigaOS4.x, AROS, and MorphOS, uses his own custom graphics tricks and tools, to help speed up and improve graphics display of ported games. Is there any chance that MorphOS will get any new tools, library's, translation layers, MUI classes, etc. that will improve our chances to more easily port games, and other graphical programs to MorphOS from Linux, or even Windows?

To put it more simply, what is the general outlook for porting games and programs to MorphOS that require 3D support to work, and is there anything that can be done to improve the current situation? What would it take to make porting newer stuff possible, and writing new games and programs natively for MorphOS able to use newer 3D standards?

If Bigfoot is going to spend his free coding time giving us better video card drivers, it sure would be nice to take advantage of the increased 3D power & speed, with newer games and programs, using newer 3D standards.

When you are using OpenGL with SDL you are still using TinyGL.

Hi BSzilli ! and others !

Any chance to view update for SDL2, SDL2_mixer etc... ?
with any 3D support (same as AOS 4)

Thanks

> I'll assume that supporting a subset of OpenGL was chosen
> because its documentation is public knowledge, and DirectX
> versions probably are not (please correct me if I am wrong).

Direct3D API is openly documented as well, else you couldn't easily write programs that use it. Reason for implementing (subset of) OpenGL over Direct3D was most likely that OpenGL has always been on all platforms including Windows whereas Direct3D is a Windows-only thing.

> Is there likely to be any updates to TinyGL

Latest TinyGL updates are as recent as MorphOS 3.10. It's not unlikely that MorphOS 3.12 will contain further updates.

> or is it more likely that new graphic improvements for MorphOS
> will come in the form of SDL updates [...]?

SDL does not contain a 3D graphics API so must be combined with an existing 3D API like OpenGL. Besides, MorphOS itself does not contain SDL (PowerSDL has been a 3rd party download so far).

> is there anything that can be done to improve the current situation? What
> would it take to make porting newer stuff possible, and writing new games
> and programs natively for MorphOS able to use newer 3D standards?

Simple: implement a recent OpenGL version ;-) AROS for instance has OpenGL 4.3 (through Mesa 12), and A-Eon's approach for OS4 (Warp3D Nova + OpenGL ES 2 + GL4ES/EGL, effectively giving OpenGL 2.1) is better than nothing.

> If Bigfoot is going to spend his free coding time giving us better video card
> drivers, it sure would be nice to take advantage of the increased 3D power
> & speed, with newer games and programs, using newer 3D standards.

Absolutely. Part of the problem may be kiero's limited availability.

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=12010&forum=3
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=11947&forum=49
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=12368&forum=3&start=18
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=11770&forum=32&start=33

Quote:

I'd be interested in updating it (I made small contributions to the OS4 SDL2 port), but only itix has the source code for PowerSDL2.

It will be excellent for MorphOS users to have SDL2!

BeWorld and others devs could be ports many news games!

Thanks for your proposition:

I hope Itix would like share his code.

Yes, there no more game in SDL 1.2.....

Itix if you hear us

[ Edité par beworld 04.03.2019 - 19:38 ]

Thank you Andreas_Wolf, for the simple and straight forward answers to many of my questions. Your answers cleared up a lot of misconceptions I had on this topic.

Just a heads up: SDL's use seems to have declined greatly in the last 10 years. Unity and libgdx are more likely to see FOSS games developed for them.

Quote:

SDL and Unity can't really be compared, and libGDX is for Java. As long as people develop their own engines in C/C++, there will be a demand for SDL. It's not like Unity will be ported to MorphOS, so a complete implementation of SDL2 will be useful to have.

Quote:




+1, We need SDL 2

A bounty for SDL2 is open on WArMUp website :

https://www.warmup-asso.org/sections.php?op=viewarticle&artid=33&prev=1

@Itix: can i attribute it to you ?

Quote:

>> Part of the problem may be kiero's limited availability.

> Mark did mention that he would want kiero's help for any
> serious upgrade to our OpenGL implementation (or the
> implementation of something to replace it).

Yes, my third link is to where you reported this in May 2018 :-)

Quote:

My apologies, master indexer.
I'm all for a better OpenGL implementation, as long as it isn't something like OpenGL ES.
DirectX isn't possible, its proprietary.

What other options might be possible?

> I'm all for a better OpenGL implementation, as long as it isn't something
> like OpenGL ES.

As outlined in comment #4, third parties offer OpenGL ES 2 and a wrapper for OS4, essentially giving OpenGL 2.1.

> DirectX isn't possible, its proprietary.

As outlined in comment #4, Microsoft's code is proprietary, but the API is public and can be reimplemented at will. There are various examples of this (mainly as wrappers to OpenGL):

https://wiki.winehq.org/FAQ#Does_Wine_support_DirectX.3F_Should_I_install_Microsoft.27s_DirectX_under_Wine.3F
https://www.fdossena.com/?p=wined3d/index.frag
https://sourceforge.net/projects/dxglwrap/
https://dxgl.org

Wrappers and reimplemented Microsoft APIs...how about just an updated OpenGL without the complications?

[ Edited by Jim 16.03.2019 - 03:55 ]

> how about just an updated OpenGL without the complications?

I'm all for it, but the hurdle seems to be the limited developer resources. And I think that updating to a more modern version of OpenGL by implementing OpenGL ES and using an existing wrapper on top of it is better than no updated OpenGL at all.

Quote:

Like HunoPPC's work? What makes OpenGL ES that much easier to implement? And how much performance would be lost using a wrapper?

The whole idea reeks of OS4.

Quote:

HyperionOS4 or CloAEONtoOS4 ????

*runs*

>> I think that updating to a more modern version of OpenGL
>> by implementing OpenGL ES and using an existing wrapper
>> on top of it is better than no updated OpenGL at all.

> Like HunoPPC's work?

Yes, but more like Daytona675x's, ptitSeb's and kas1e's work.

> What makes OpenGL ES that much easier to implement?

I don't know, but I guess there's a reason Daytona675x implemented OpenGL ES instead of OpenGL.

> how much performance would be lost using a wrapper?

Spectre660 reported some OGLES+GL4ES vs. MiniGL performance comparisons for Quake 3 in August 2018 here on MorphZone.

> The whole idea reeks of OS4.

Yes, I was explicitly using OS4 as reference in comments #4 and #16. And this way, OS4 (with 3rd party enhancements) has better OpenGL support than MorphOS.

Quote:

As Andreas_Wolf states in a post after the one I quoted above, yes, he was reffering to what is being done on OS4. Maybe something the same or similar could be done to improve the ability to port 3D software to MorphOS. Why would be doing something similar to the 3D work being done by 3rd parties for AmigaOS4 be a bad thing? I think having a compatible 3D support system for MorphOS and AmigaOS4 would be a great thing, as it would make it easier for the limited number of remaining programmers making ports, or original software for both AmigaOS4 and MorphOS, to create software content for both platforms. Hans DeRueter (sp) is doing great work from what I can tell, in expanding 3D support, and writing drivers for newer video cards for OS4. I don't see any reason for Mark and Hans to not collaborate with each other, if they can benefit each other's work.

Although MorphOS and AmigaOS4 share the same targeted users, I no longer see them as direct competition, the further apart they grow. When/If MorphOS NG for x64 is released, the two OSes will be miles apart and essentially traveling in almost opposite directions. It will almost be like the difference between AmigaOS4 and AmigaOS for 68k, and AmigaOS running via any version of UAE, in regard to users anyway. Many WinUAE users no longer use 68k hardware, and some users only run OS4 via WinUAE. Fewer and fewer users own and use 68k and PPC hardware regularly, and also use any version of UAE. I think that most users have made their choices between hardware or emulation, to reduce redundancy, even if many who use emulation primarily also own real hardware that is stored in a closet somewhere.

[ Edited by amigadave 17.03.2019 - 10:14 ]

What might be better is that if a newer GCC comes out like is happening on AROS, Mesa can be ported from there. The latest version supports Vulkan, OpenGL ES and legacy OpenGL.

By the way, if MorphOS gets multithreaded architecture OpenGL ES is more threadable anyway and OpenGL contexts will fade into the sunset.

From kas1e
Quote:

Source

OpenGL ES 2 version 2.1 now in testing. (this is ogles2.library)

Quote:

[ Edited by Spectre660 18.03.2019 - 07:41 ]