Odyssey 1.23 progress: r5 RC2

@radzik

here there is the last webkit patches from Tobias Netzel and are about R600.8.2 version for Leopard PPC .
http://sourceforge.net/projects/leopard-webkit/files/600/Sources/

In any way i say Teen4Fox is Geko and Geko working great on PowerPc ... Firefox is on ppc the version is on lnx ppc the 43.03

@tlosm
You should read the 104Fox development blog more:
http://tenfourfox.blogspot.hu/2015/12 ... and-maybe-45-as-swan.html

@BSzili
all is related the macports gcc48 and gcc49 that are crappy i was reported months ago to macports mantainers ... lets hope something will be fixed

@tlosm
GCC is the least of our problems. I should have TL;DR-ed out the most important part: Mozilla has been working on its own functional programming language called Rust. Parts of Firefox are already in Rust, and the new layout engine will be written entirely in it. This means no Firefox, unless someone is willing to maintain a Rust compiler for OS4.

@BSzili
This means we have no other open alternatives?

I think for now is the best fix by us WebKit, but what about future? NG systems without moder web browser??

@radzik
It mean that we need to find someone for current webkit, he will do the job, and then, we will use that browser for 3-5 years. And if we all will be here after 3-5 years, we can think what to do then. But for now will be enough to fix current version.

I heard something that Bigfoot can do it?

@kas1e

I hope the webkit that we have will work for next 5 y.... the inet is running really fast

You can take that 5 year with a bit of salt. You will see pages that just stop working. For example like google when they update their pages with "old browsers, please update". Like now youtube, google drive, spreedsheats etc. At least microsoft when they have "old browsers, please update" we can continue to the page anyway like for hotmail.

But a good thing is pages that migrates from flash to html5 as we lack flash.

BSzili
Quote:
Thanks for the highlight.
If they develop a new language, does it mean C is not as good as everyone says?

Maybe they should have tried AmigaE!

@Elwood
C might not be very good, but it is ubiquitous. They are switching to Rust, because programs written in the functional paradigm are very easy to parallelize. AmigaE is still an imperative-only language :)
What really matters is that 43 ESR is the last Rust-free version of Firefox. Porting that would be an order of magnitude more work than fixing Odyssey, but you'd still end up with an outdated browser.

@BSzili

We're doomed! Doomed! DOOMED I TELL YA!

@Joeled

If retro-surfing ever becomes a fad, we will be the platform to beat 8)

On a more serious note, once and if the endian issues are solved in webkit, will it be easy to update (I use the term easy in relation to how difficult it is to solve the current endian issues)?

@BSzili

Yes of course.
It's still interesting to see the dependencies to build the Rust compiler are not that high https://github.com/rust-lang/rust
The following should be considered though: https://metajack.im/2013/12/11/building-rust-code--current-issues/

@Elwood
GCC's dependencies aren't that high either. This seems to be a perfect opportunity to get back to my original point. Who is going to port the latest Firefox and maintain a compiler for a completely new language, if you can't even find someone to fix JavaScriptCore?

Agreed, and if the world around us is abandonning PPC, how coud we stay up-to-date with the very little amount of users/developers we have?
This will become soon a crucial question.

But it belongs to another thread. Let's concentrate on Odyssey progress right now.

Just to throw a wild idea out there: Amithlon apparently had a modified GCC compiler that generated x86 code which behaved as if it had 68k endianess. i.e. It was modified so that every memory access had it's endianess swapped. I wonder if something similar could be done for PPC, and then SOMEHOW used to just compile the JavaScript part of WebKit (e.g. javascript.o), such that the endian issues were resolved.

I'm sure there'd be a lot of difficulties trying to do something like that (e.g. how would the reverse-endian javascript code interact with the rest of WebKit running with normal endianess?), but I wonder if it might be less effort & more future-proof in the long run?

Well, just a crazy idea, feel free to ignore it

@Joeled

Quote:


-- eliyahu

@BSzili

How many developers we have? It's really confusing that nobody has will to fix WebKit. Or nobody cares about web browser on NG systems? Where is the problem?

"@Kas1e

At exit, when Odyssey window is just closed, I often have a requester saying something like "Wait until child end. Webcore file..."

Sorry, I wanted to grab the window but it closed itself.
This problem occurs with OS4Depot version or now with your latest build.

Thanks
"

I finally got it.