/tech/ - Technology★

No

No

No

>>850715 (OP)

>hey this task is too monumental for me to even comprehend, how about you guys do it for me?

>>850721

I'm all in if anyone else is in.

Embedded Servo.

>but muh rust

>but muh mozilla

Fuck you.

>>850823

isn't it just a CSS implementation alone?

It's called Gecko.

>How would one go about doing this?

Have millions of dollars and lobby W3C to stop coming up with bullshit, websites are now hard.

>Is it even possible?

Yes

>Is anyone here crazy enough to try it?

On the chans? No, literally impossible for one man unless they're unemployed genius autist Davis tier.

The best we have is Netsurf, is a browser composed of many portable C libraries.

Help them out on their HTML5 parser (Hubbub) and their CSS parser (LibCSS)

>>850715 (OP)

Do it yourself, ideas man.

>>850715 (OP)

We've tried this before.

https://8ch.net/zirconium/index.html

and again

https://netrunner.cc/

and again

https://github.com/OtterBrowser/otter-browser

and again

Making a new browser is literally a chan meme that typically goes nowhere, or at least progresses extremely slowly

>>850899

>https://8ch.net/zirconium/index.html

chromium fork

>https://netrunner.cc/

looks like a legitimate attempt

>https://github.com/OtterBrowser/otter-browser

Qt? digging around the source I can't tell if it's using qtwebengine or not, it doesn't look like it. might also be a legit attempt.

>>850911

not to shit on these projects, doing this from scratch is hard. you might as well write your own kernel. there's probably more lines of code in chromium/firefox then there is in the linux kernel.

>>850913

Thats just sad man.

>>850715 (OP)

the Borg method:

>download webkit

>build it

>build it, this time with superior lang embedded

>piecemeal-replace webkit with superior lang

>one (glorious) day webkit's original code is limited to some library wrappers and the important stuff is your own

requires good two-way communication between the host language and the new language, not simply an FFI.

in order for this to have any point, the superior language needs to be really superior. If you can just encode the old logic in the new language with no changes, then you'll end up having to fix the same security bugs in the new code.

language candidates:

1. ATS

2. ??? (dozens of languages with good enough C interop, but the languages themselves aren't good)

>but wait WebKit is written in C++

it's still possible

but have you considered replacing w3m instead?

>>851093

>but wait WebKit is written in C++

GNU/webkitgtk+ is C

Qtwebengine is C++

I was thinking the same thing though playing around with it. Build it with webkit first and then replace pieces if I feel like trying.

>>850862

I think that's Stylo, Servo is supposed to be the full engine.

Contribute to netsurf, niggers.

>>850747

No.

>goto hell;

First, I would build a Lisp machine on an FPGA which would increase productivity by at least 1,000,000 times. This should make it easy enough for one person to write a full Webkit replacement in a relatively short amount of time.