odilitime

irc.rizon.net #/g/netrunner

I'm working on a threaded-prototype of netrunner where messages are passed between each other and could even be move out to IPC or even the network.

Also why I'm thinking this project is more like electron. And a browser could be built on top of that.

We've had a couple webkit developers tell us how hard it was to hijack some seemingly simple behavior. Interesting to hear that Firefox is the same way.

I'd love to compare diagrams.

- a ton of dependencies - randomly hangs - can't handle suspending - cursor blinks when not in insert mode - doesn't work - Python not C - QT - Lacked security features

I'm not sure if all of these are true but this list was their complaints and you'd have to ask them for more detail. A lot of them do sound like they're fixable or just personal preference.

4chan has been discussing options since mid-June and probably has around 50 threads revolving around Netrunner project. And the current devs are not the same people that started the initiative, so the project has evolved quite a bit.

Netsurf, is another project that comes up quite a bit: http://www.netsurf-browser.org/ As well as various forks of the major browsers such as palemoon, icecat, unChromium, ungoogled-chromium, waterfox, opera 12.x, vivaldi, and otter.

I actually have an offline repo with multithreading started but put on hold to focus on other issues. I'd like to implement threading sooner than later as it will be easier to do on a smaller project than a large one.

It's all about the architecture. We want to get that right first and are willing to throw away our current code to make sure we have this nailed. We are focusing on the renderer using a flow-programming style (only update what needs to change) as much as possible. So we're yielded to the OS unless there are pending events (using glfw). So no render() 60 times per second (unless there was something pumping events as such)

We're actually going to be converting HTML/CSS into JavaScript and the renderer will be handling a dynamic DOM. So we're not falling into trap of expecting things to be static. A couple of us are ex-game devs and have worked on HTML 5 games.

Thanks for the kind words.

My goals are a little different. I want this to be a browsers by devs for devs. And I'd rather learn how to make one by exploration instead of copying.

No expectations of this being a good browser, just hoping for the most dev friendly and we'll see how far we can take it.

The goal now isn't a browser but a framework that is easily modified, well documented and can be embedded in other projects.

We realize how much work a real browser is but if you look at the source of WebKit, blink or servo, it's a huge mess. We're trying to be low dependency and multiple interface, perfect for scripting.

Our web page targets are simple just 4chan/g and stackoverflow. But we also have plans where this framework can also be used for document viewing and editing.

I'm thinking of the project as more of an electron replacement. But we'll see where the future takes us.

Tabbed browsing by the end of this month is on track.