That is interesting. What would it mean to have that working out of development and cost?

@Capehill
Thanks to the info you shared, I profyled the two versions and got reports, which include the SDL methods. I created the following issue
https://github.com/AmigaPorts/SDL/issues/170

I agree 100% with all the points kas1e mentioned above, but I would like to mention a few more things about clib2 status:

- the code contains definitions that have to do with AmigaOS 3, Amiga 4 and MorphOS. But hardly any of the developers of these systems use it. MorphOS has been away from clib2 a long time ago, and AmigaOS 3 developers do not use it at all. So, maintaining any code that needs to keep support for all these systems is out of the question. One of the reasons why clib4 was created was to focus only on OS4.
- clib2 has a branch named development (https://github.com/adtools/clib2/tree/development), where Olaf is doing all his work on. His last commits were two years ago. As much as I know, they are not merged back to master because the latest changes are not tested enough by anyone. I wouldn't call it as active maintenance of a libc for our system.
- newlib is closed source and maintained by one person (salass) whenever he has time and needs.

Our goal for clib4 is not to bring confusion to the developers, but provide a better option and free them from any barriers and blocks. Instead of bitching about having to set a gcc argument for a different libc, we all should be grateful that a handful of people are doing what is needed for the development environments for our platform. They develop, they update, they test, and they benchmark as much as possible.

Currently, we have newer versions of tools available, newer and better libc free by any company and open-sourced, a better cross compiling environment for everyone to jump on, and thanks to Balaton, a cheap way to develop on any system using qemu, without the need to spend a few hundreds of money just to test a hello world code. A newer version of webkit browser would never be possible with newlib and clib2.

So, stop complaining, see the benefits and give us a hand to get into a better place. And most importantly, roll up your sleeves and start writing some new code. That's the only way to make this platform go a little bit further.

And if any of you do not like what clib4 stands for or what we develop, so be it. Try to develop with clib2 and newlib if that covers your needs. No one pushes you to use clib4, gcc11 or cmake. But others want newer tools, better editors, working debuggers and all these things that the modern development requires.

I would love hearing from one person in this community how we are going to bring Rust or Go in our systems. How will we be able to use better profiling and unit testing and more modern stuff. IMO, that's where we should focus, and not if the clib4 is a third libc or not.

Earlier today, we released a new add-on package for the OS4 SDK. The SDK Add-on package is meant to be installed on top of the AmigaOS 4 SDK. This includes new versions of necessary tools and compilers for native development on AmigaOS 4 systems. The package contains an installer that helps the user to proceed with the installation with just a few clicks.

We are working to add more tools in the future releases, that will help the developers to have a better experience for native development on AmigaOS 4. This package might include experimental software. If you experience any problems, feel free to open an issue so that we can work on a fix.

You can find it at
https://github.com/AmigaLabs/SDK-Addon/releases

@Capehill
Thanks a lot for the guide. I will test it and let you know.

What is the sdl3benchmark? Is there an sdl2benchmark? Do these benchmark SDL2 and SDL3 on various methods? Do you see any performance differences on OS4?

@Capehill
Thank you for your prompt reply. Should I be able to see how much time was spent in SDL libraries with profiling programs, like the Profyler I mentioned above? I am asking, because I tried with the non-debug version, and I couldn't get a clear usage of the SDL_ methods in the app list.

Do you have any other way or tool I should test, that you know would work better?

@Capehill
If I use the libSDL3_debug.a when I link it statically with my app, should I be able to profile the called SDL functions? Maybe using the Profyler (https://os4depot.net/?function=showfil ... opment/debug/profyler.lha)

Is it the same for libSDL3_debug.a, because I tried it and I couldn't get any info?

I am trying to understand if there are SDL slowdowns on the functions used by Lite XL, and if there is a difference between SDL2 and SDL3

@K-L
Thank you for your proposal, but afxgroup did the majority of the work. So, donating to his ko-fi page is totally fine.

@Capehill
The LITE_USE_SDL_RENDERER is for the darwin systems, so I do not use it in AmigaOS 4 port. I used the SDL_RENDER_DRIVER with ogles2, but there was no impact.

Most probably will be a problem from the latest code in Lite XL itself, which is something I am discussing with the original team.

By the way, I saw in your makefile you have the following comment

CATCOMP = catcomp # TODO: is there a cross-compiler version?

There is flexcat which can be used with cross-compilers
https://github.com/adtools/flexcat

@Capehill
Not really. Yesterday I ported the latest Lite XL 2.1.8, which is the first version the original team released, that uses SDL3, and it is a lot slower in scrolling and animations. Now, it might not be the SDL3 to blame here, which I am discussing with them. I just wanted to know if you have any information on performance benchmarks or something else, that show how well SDL3 is working on our system.

@Capehill
May I ask you if you have seen SDL3 being slower or faster than SDL2? Do you have any benchmark comparison between the two?

@kas1e
Awesome work man. It is great making native apps right?

@BillE
Have you considered using something like browservice?
https://github.com/ttalvitie/browservice

@kas1e
Thank you so much for bringing this utility to our systems. Great work pal.

I have the same issues with Linux and the USB device updates. So, it seems it doesn't have to do with the host OS.

As much as I tried to find a way to mount a folder as a hard disk, this was not possible. The other solution could be to use an ISO, but this will require much more work than just restart the emulation.

I am interested in a better solution as well.

Awesome post, as always. Thank you for creating it and keeping it up.

@Mr_byte
I can't agree more with you.

No, you don't. It seems that it hit the APi requests quota and YouTube forbids any more for today. Tomorrow it might work again at some point.

Unfortunately, the YouTube API has a really low limit, that's why I was using Invidious. But now Invidious API is down on every server they have, and this cannot be used any more.

So, the only way to find YT videos right now on AmigaOS is by using the m.youtube.com or an invidious website, and not through the aiostreams website, until I figure another way to get the needed data.

@pjs
Today, I released a new version of aiostreams that fixes the error with the search you mentioned. The reason why this was failing is that YT decided that if you search for videos through their API, they will return channels :O

So, there was a failure on finding some necessary data for these records.

@Maijestro
I am working on the page, but the aiostreams scripts do not need the page. They have parameters to search for videos in the shell. If you check the included amigaguide you will find them, or if you use the -h parameter, it gives information as well, i.e.