After the recent work to clone MCE, more of the major hardware bits are falling into place
Main hardware bits remaining are audio, cellular modem, FMTX, GPS and possibly OMAP DSP (I dont know how compatible the DSP in the N900 SoC and the DSP in the Neo900 SoC are)

Both are C64+,100% compatible. And the driver is upstreamed (albeit in staging)

I would very much like to help out on the Fremantle Porting project, but I just don't know where to jump in. I'm a competent C programmer, I've made a few apps with Hildon widgets, I've got the Maemo 5.0 SDK running on a laptop, but I've never done anything like try to make a replacement library, so I just don't know how to get started on something like this.

I've not seen anything in this thread that indicates how this project is being coordinated. Is anyone assigning tasks? If not, how do you prevent manpower from being wasted by duplication of effort? Has the list of closed-source components that need to be replaced been prioritized?

Are there any tasks remaining that would be suitable for a newbie, or has all the low-hanging fruit been picked, and all remaining tasks are quite challenging?

Has anyone described somewhere the process by which he reverse engineered one of these libraries? Do you start by looking for API documentation, or do you dump the library symbols? Do you disassemble?

I've got time to work on this project, and I'd love to contribute, but the initial learning curve looks pretty steep, even for someone fairly familliar with Linux (I've written device drivers, and ported Linux to a previously unsupported PowerPC VME card) and C. How can I get started?

Does anybody have an idea which web browser can be used on the Neo900 or more generally on Fremantle in the future at all?
All the free (as in speech) mobile linux (i.e. non-dalvik) browsers seem to be discontinued and the mobile version of Opera for Meego is proprietary, which doesn't seem like a good default for a system that strives to be free.

Would it be feasible to free some Android browser from the dalvik stuff?
I've just tested almost all browsers available in Debian and none of them seems to be suitable for a mobile device with a 800x480 px display.

Hi!
please /join #maemo-ssu on freenode IRC! right now FatPhil, Freemangordon, jonwil, Pali, are the guys most active with FPTF, and coordination entirely happens on IRC (if at all ;-P).
Please join that channel and holler, eventually one of the guys will talk you thru all the initial fog.

@fmg, pali: it maybe would be a good idea to post a mini-howto_get-started-FPTF here, and occasionally give a status report here as well. ;-)

cheers
jOERG

I tinkered a bit with several browsers in some x86 virtual machines and came to the following conclusions (so far):

Ubuntu Phone's webbrowser-app [1]:
+ UI works mostly fine on a 800x480 desktop
+ autohiding url bar is easily accessible by dragging it up from the lower screen edge
- uses experimental QML api
- needs multitouch to zoom (pinch to zoom)
- no tabs, no way to open links in new windows, therefore strictly linear browsing
- no support for HTML 5

KDE plasma active's webbrowser [2]:
+ seems to work well on touchscreens (from what I can tell without having one)
- UI needs too much space to be useful on small displays
- depends on patched KDE libs, therefore incompatible with Debian (unless one is willing to replace major parts of Debians KDE)
- huge list of dependencies (when installing only the browser with its depends the list doesn't seem too long, but although the browser runs it only displays a black box instead of websites - installing the whole plasma-active DE fixes this but pulls much more dependencies, maybe 1GB)
- needs multitouch to zoom (pinch to zoom)

Chromium:
+ can be made to act mostly like a mobile browser by adjusting the agent string and enabling "Enable touch events" in about:flags (which enables drag page scrolling).
- url bar and tab bar to small and apparently no way to change that (this is a mayor show stopper on tiny displays)

Firefox:
+ can be made pretty much compatible with small touch screens by using the following addons (maybe there are better ones):
# Hide Tabbar (becomes visible by sliding over the region where it should be or pressing a hotkey - camera button?)
# Hide Navigation Bar (becomes visible by pressing a hotkey - camera button?)
# Big Buttons (bigger tool bar buttons - only one size available)
# Grab and Drag (scrolling)
# User Agent Switcher (comes with an "iPhone 3" setting)
# Make Address Bar Font Size Bigger (does what it says, unfortunately it doesn't increase the url bar height and places an icon in the bar that can ot be removed permanently)
# not an addon but setting this in userChrome.css increases the tab bar height so it can be used without a stylus:
- the result still looks crude (might be adressed by better choice of addons)
- when Hide Tabbar and Hide Navigation Bar use the same hotkey they sometimes don't respond to it at the same time
- url bar and tab close buttons still too small


[1] http://packages.ubuntu.com/trusty/webbrowser-app
[2] http://packages.ubuntu.com/trusty/pl...ive-webbrowser

IMO one option for browser on Neo900 Fremantle is to keep microb and upgrade microb-engine to a newer Gecko version.

What about Alopex? http://talk.maemo.org/showthread.php?t=89506
I never used it myself, because of the thumb only, but what I heard, it is really god.

isn't browser rather an app than an OS component? (OK, arguable on maemo ;-P)
Anyway the purpose of FPTF is to allow apps to run unchanged on new hw-platforms.

If that's feasible...

I never heard of it up to now, but according to the thread it seems pretty much dead.

I guess on any smartphone - which despite its other qualities the N(eo)900 still is - the web browser might be one of the most used applications available. And it's one of the most prominent potential security risks.
So not using some ancient browser is a major security concern to the whole platform.

Maybe I got that wrong, but wasn't the purpose of FPTF also to improve Maemo's software situation in general, starting with a recent kernel and eliminating some of the proprietary blobs?
Afaik these are the major hurdles that always made it hard to impossible to keep up with the rest of the GNU/Linux progress.
So if these hurdles could be removed by FPTF then a new browser based on recent technology should more or less come naturally, assuming there is already an existing suitable browser out there.

If this thread is not the right place to discuss this, I understand and will happily join the choir of those requesting a Neo900 sub-forum.