WebKit based browser initiative

Because there is a lot of discussion about a new browser that we highly need for AmigaOS 4, we had a good discussion on discord today, and I was thinking that there is no better time to start working on it than now.

I strongly believe that this should be worked by a small team of people who can work together to make the first steps to the top of this mountain.

So, I am forming a team of people who would like to contribute. You don't need to be a developer to do so. This project has so many aspects that people can contribute everywhere.

If you believe you can share your expertise, you are willing to work in a team and put effort into this project, please contact me with an email, a PM, on twitter, on discord, wherever you want.

Together we will see what is the potential, what we want to achieve and how.

@walkero

Kudos to the effort and all the best of luck.

@walkeroQuote:

Great to hear that! I wish you luck with this project, also I will support it much how I can. I am not developer, but if you need someone for testing when time comes for that you can count on me. :D Also donate button for this project would be nice to have ;(

@walkero

I not a fan of the build systems, but if we can get it to build natively on AmigaOS4 without too much hazel, I might be able contribute a bit.

Maybe you can compile it natively on OS4.. But it will took 4 years..

@walkero

I guess it goes without saying that if you can pull this off *everyone* in the community will support in one way or the other...may it be time for testing, giving feedback, translating, moral support or whatever he/she/it can afford...

I can't code either, but whatever else I can, I will gladly try to not break

Please make an package with required files for people to easy install. So maybe more can try to help. Not all are good at finding, downloading, installing all kind of strange files to get things started.

@walkero

If I can help somehow I'm in.

Hello George. Count me in as a beta tester.
Sam440ep/AOS4.1FE - PEGASOS II/AOS4.1upd6

@walkero

I have not been following the discussion, so I don't know if you are targeting high end machines or all OS4 systems. When the time comes, I would be happy to try out the new browser on my MicroA1 and provide feedback.

Just for reference, AWeb, IBrowse, NetSurf, Origyn and Odyssey all worked ok on my system. TimberWolf did run, but it hogged all avaiable RAM and crashed on exit.

My system has 256 MB RAM, ATI Radeon 7000 and 32 MB graphics RAM.

Meanwhile, all the best,

redfox

@walkero

Our community is so small that I think every programmer have to volunteer in this kind of effort.

What jacadcaps said that you need some 12 core machine with 32 gb memory to even compile WebKit. I don't have even that heavy Windows machines.

@TSK

That sounds stupid… why do we need to recompile every file on the planet?

Should be as easy as:

Gcc myweb.cpp -o myweb.elf -ljavascript -lwebkit -lgui

Maybe we can turn webkit upside down, and make GUI and other parts we want to reapply change into Amiga libraries, like JavaScript engine.

Another stupid thing, is why does not Amiga libraries work if they compiled with G++, instead of GCC, that’s another question I have, is there problem with identifying symbol in G++, that’s issue for OS developers, but might really nice to compile amiga libraries with G++.

@TSK

Quote:
When/where did he say that? You got a link?

I'm sure a 12 core machine with plenty of memory would help speed up building WebKit, but doubt it's a requirement.

Others suggest (link) that the build machine needs ~2GB per vcore for a full parallel build. If you don't have enough RAM, you can reduce the number of build threads. Fewer build threads means that the build will take longer, but it shouldn't fail.

Hans

@Hans
The number of required cores and memory is mentioned at
https://github.com/jacadcaps/webkitty/blob/master/PORTING.md

But I don't believe that's a problem. There is plenty of power in the world, that can be used. Using CI/CD and the necessary logic you can create a cloud server, run the build and destroy it afterwards quite easily.

I would like to see different parts of webkit be built in parallel by different servers, which I think is doable. Also, I believe it is doable when you have a part compiled to not need to recompile everything, but only the changed parts.

This reminds me a lot of the old discussions we had here before the porting of the first Origyn Web Browser
As demonstrated by kas1e there is NO need to form a large team of developers nor to collect a huge sum of money, all we need is mostly a single motivated person and ALL the necessary tools available for starting the port, and then a lot, really a lot of annoying betatesting...
So as a starting point first of all we need someone to implement TLS support on Exec, then at least start from there, then the rest....
So the question now is, is AEon available to start implementing TLS on Exec, and above all to release a public kernel update in a reasonable decent time (aka not 10 years but 6 months max) so that we can start thinking about start the port of Wayfarer to AmigaOS4?

@samo79
I agree with the majority of what you say. Although I have to disagree with the TLS part, because the port is not blocked by the TLS, as much as I understand. There is going to be a huge performance impact, sure, but it is not blocked.

And as much as I understand, the TLS problem would be easier to be fixed if we had something like a WebKit to test it with. I see that as the chicken-egg problem, and from my side, the way to start is by starting the port of the WebKit.

About who is going to work on it, as proved the last few years we discuss it, no one started it so far. And the developers that have the ability to do it on their own, are occupied with other projects. As I stated at the start of this thread, I am not the most experienced and the most capable person to do it, but I am willing to start working on it and dedicate the rest of the year to it. I might end up with nothing or with something. At the end of the year, I will evaluate what happened and if it makes sense to continue or not.

But anything that comes up from this work is going to be available for everyone to see, and if I fail, my amount of work might be useful for someone else.

But also it doesn't make sense for me to work alone. People with different expertise can help and already a few volunteered. And for a start, we are going to keep the team as small and flexible as possible.

And remember, this is still the first step. If someone feels he would like to contribute please contact me.

@walkero

Yes, you won't need to rebuild the whole thing every time; it does have incremental build. Having more cores (and fast hard-drives) speeds things up when you have to rebuild parts frequently.

I'm grateful that you decided to not wait for kernel TLS support, and to just give it a go.

Hans

@all
The work already started on the port:

- already formed a small and flexible team of 4 people who will help on the WebKit port, and established a channel of communication.

- Jacek's webkitty forked
https://github.com/walkero-gr/webkitty/tree/amigaos_2.36.8

- A project is created with the work that is needed to be done and I am working on the first tests, based on Jacek's guidelines
https://github.com/users/walkero-gr/projects/1

Remember, we are still at the beginning.

@walkero

Who are involved?
Good luck guys ✌️

@walkero

Congrulations. It will be very exciting to monitor the progress :)

I guess the port will be clib2 based ?