Hi2all,

Thanks for sending me e-mail, haven't noticed there is some interest about it....

Here is some explanation about what is going on and roadmap...
Originally I was doing embedding API's which would allow to share common functionality between different ports of Native Mobile Firefox (Android, Win8, IOS, Maemo et.c.), also allow to build "quick" browser using these embedding API's and make it works out of the box with IPC or OMTC(multi-thread rendering).

I did both in this patch queue: http://hg.mozilla.org/users/romaxa_g...e/file/default
embedhelper - OMTC
ipcembedding - IPC - patches
also here is test Qt embedding library which allow to embed this new embedding particularly into Qt/Qml scene.
http://hg.mozilla.org/users/romaxa_gmail.com/qtmozembed

Next step is to merge IPC and OMTC API's and make them similar, so qmozemedtest will work in IPC or OMTC mode just by pref change.

Two big tasks TODO:
1) Make this embedding API as part of mozilla native firefox browser (so it will be supported as part of the product)
2) Find and adopt native/QML UI which would work on top of these API's

This would be useful for Nokia N9, Qt environment and probably for future Jolla phone.

I've managed to make default Nokia N9 browser work with these IPC Lite API, and it actually works, but Default browser UI and source code is closed, so I need to find something working in opensource, or write new Qt/QML UI from scratch (too much nitky-pitky work need to be done).

(Now I'm traveling and we'll be back at end of September), sorry for bad response if such.

Br, Oleg

May be plasma active devs will be interested in working on such kind of UI as well? Having Firefox engine enabled browser in Qt mobile OSes (including Plasma Active and Jolla) with responsive UI will be a real killer product.



But I think it's good to reproduce current Fennec mobile UI (which has a handset and tablet flavors). Mozilla did a great job on design, so Qt based Firefox shouldn't be any inferior if it to be comparable with other versions. May be we can make a community effort of reproducing it. I'm interested in participating.

@romaxa - both me and zeq tried to follow your build instructions here. Unfortunately both of us had no success, the patches don't apply. Mind to have a look when you have some spare time?

QT5 snowshoe browser is opensource and already runs on N9, maybe you can reuse the UI (which seems to be great).
http://snowshoe.qtlabs.org.br/?about
https://github.com/snowshoe/snowshoe


Or HeliumReborn UI, hint: the author of it worked for Nokia and is now working at Jolla
http://veskuh.blogspot.de/2011/12/he...-for-nemo.html
https://github.com/veskuh/HeliumReborn

I think it'd be preferable somehow to unify the UI design paths for Mozilla, in order to avoid totally different approaches. I.e. their design development for "native" or XUL Fennec as applied to the Qt based embedded one. Unless they'll go too much into Android specifics (I'm not familiar with that). It requires some thinking (if it's even possible at all). Ideally, in the future Mozilla can switch to using embedded IPC even on Android, so having a unified design somehow formalized is a plus. Though it'll probably require some declarative translation between some metaspec and resulting native variants. Just an idea. Right now they are too engrossed in Android, but it's bad. Since more native platforms will come up, they need some shared metadata to define the UI if they want it to be similar across platforms.

I think the native UI is based on the Australis UX which will be used in future versions of Desktop Firefox (afaik) --> https://wiki.mozilla.org/Tab_Strip_Visual_Redesign

Here is the Mozilla WIKI site about Mobile native UI:
https://wiki.mozilla.org/Fennec/NativeUI/UserExperience

I personally would much rather see an update of the XUL-based version using OMTC and a UI that looks more similar to the Firefox for Android UI.
Just because Android sucks enough that it made it impossible to do a decent versuon using the glorious web technologies for UI, it doesn't mean that's true everywhere.

Bringing the XUL UI up to date is an option, but will the XUL build automatically benefit from OMTC when it'll be fully enabled in the code, or it'll require some modifications? If responsiveness of the XUL UI will significantly improve - there won't be a need for Qt one, since XUL is naturally portable already. But right now it's a real pain to use (I'm testing it on N950). Constant freezing and unresponsive behavior are just bugging the whole experience. I'm not sure though if it's Harmattan's specific problem or not. I didn't test it on Nemo yet.

Oh, I've been working on different branch.... now it is in default, you can try again

Well XUL was slow on Android too. I'm sure it's XUL's fault. Even all desktop Firefox versions are being slown down by XUL. Thats the main reason why everyone thinks Chrome is fast, because it's UI is fast.

In my opinion they should have gone all-in for QT in mobile and desktop Firefox versions long time ago. You got all major OS targets nowadays in Qt (Win / MacOS / Linux / Android / QNX / MeeGO).