QtWebEngine and Falkon are now finally available in the IPFS repo,I downloaded and tried it.
Falkon runs great,it’s even way faster than Firefox on Linux (Maybe that’s because Haiku runs on an i7 and Linux on a Raspberry Pi,but who cares?).
There are still some bugs but I didn’t expect to have none,I think they’ll be fixed rather quickly now that the biggest work has been done.
Sending this from Falkon on Haiku btw,I don’t have to use Firefox to use the Forum anymore 
Here are the HTML5Test scores of the four different browser engines on Haiku:
As you can see,Falkon achieves by far the most points.
8 Likes
For instance, having LibreOffice or Firefox on another obscure platform would make me consider switching
Well, the converse can also be considered to be true.
If there’s no LibreOffice or Firefox, it would make me consider switching.
I think the answer should be:
1 Like
It’s just placeholder names.
we have QTWebEngine_bin in HaikuDepot
Would ut not make Sense at this point to adopt the qtwrbkit/engine etc as the official backend for Haiku.
Is it just to much for haiku developers to try to keep up ???and then port webpositive to use the qt backend.
Just wondering if ut would be less work for development to maintain a qt port versus a native backend ???
I port Qt applications because it gives me a chance to get results here and now. In any case we should make native applications whenever possible. They will run faster and interact better with the operating system. But I use Haiku as my main and only operating system on my computer, so I need an environment where I can do the operations I need right now.
17 Likes
I agree native is best, but it’s looking very very difficult for developers to get webkit working fully and implement feature’s.
If haiku had a big developer team with 25 dedicated guys working in just broseer, services kit etc, yeah, more doable for sure
No, as was discussed countless times before,
I don’t understand this sentiment at all, thete is only a couple of devs working sometimes on haikuwebkit, and pretty much any time there is work done it becomes much better…
I would honestly say webkit is in a very good state right now, each release is rendering better, crashing less etc, It would be nice if more people helped out instead of claiming that porting webkit is somehow an impossible task.
4 Likes
It’s not a impossible task, but it’s like walking in a race where everyone is jogging, and has a 100 mile headstart
It’s more like someone has tripped and instead of helping them get up everyone around is discussing how hard it is to help and would surely fail if one tried.
5 Likes
I don’t know if you have a grasp of how far behind the native webkit port is ???
I have a pretty good idea, beeing one of the people to occasionally work on it.
4 Likes
No, for two main reasons:
- QtWebKit is dead upstream, has been for years at this point. HaikuWebKit uses a newer WebKit version.
- WebKit is realistically the best web engine to use long-term, since it can be integrated with Haiku APIs. For now though, many people will need a web browser that works with most websites. Falkon and QtWebEngine can become that in the short-term, if they mature.
Not everyone who works on either WebKit or Blink (via QtWebEngine) is also working on the other too, so the development resource split is rather minimal. IMO it is a good thing that there are now two major web browser engines on Haiku, since deficiencies in both can be found through comparisons. It is also likely that improvements made to Haiku itself as a result of both efforts will benefit other software, even each other.
6 Likes
Microphone support for QtWebEngine:
25 Likes
One of the core MorphOS developers is single-handedly building a new Webkit-based browser for their system https://wayfarer.icu/ And that‘s in addition to developing a new email client and working on MorphOS itself. So one person is quite enough to build a modern browser.
1 Like
What is the point? If Haiku’s goal become to just run Qt applications, I have no interest in that. You can do that from Linux already, and do it much better there. We can’t compete on that field.
What makes Haiku work well is the fact that everything is designed to fit together, from the kernel all the way up to the web browser. You are suggesting to replace the middle part of this stack. Which means, we would have to force Qt to do what we want it to, and have it interact with our Kernel (not designed for it) and do things the way we want on the UI side (also not designed for it). Essentially throwing away the one thing that makes Haiku a relevant project. If the apps are written in Qt, there is no reason to write our own kernel. And there is also no reason to write our own apps, because that would just be Yet Another Linux Distribution And Desktop, and there are far too many of those already.
Also personally I find the Qt APIs quirky and confusing (mainly because I spent far too much time writing apps using the BeAPI and it’s hard to relearn things). Which means I would be even less interested.
16 Likes
Web WhatsApp - record and play audio messages
26 Likes
Keep in mind, however, that MorphOS is a single-threaded multitasker. The development time to debug a multicore-capable browser engine is more difficult. I’ve used Wayfarer on my MorphOS box and it is pretty cool that a single-core G4 MacMini can still browse the web, even if just barely.
1 Like
Yes, it is. I see it uses Cairo (for rendering) and Curl (for network access). In the Haiku port we went the (probably crazier) way to implement our own libraries for HTTP and use app_server for rendering. So it is a bit more work.
Still, it may be manageable by one person, if it was not me (already busy with a full-time job, and various other projects to take care of in my time off work).
3 Likes
I see your point, so, if fund raising was a goal, and you could roughly define the estimated hours for a core developer to hop in full time alongside waddlsplash, what would that funding requirement look like.
I’m asking because goal driven fund raising is fruitfull