a)
Community have tried to make new OS 'from scratch'. Nemo/Mer... No success yet, but we can count these were basement for Sailfish OS.

b)
Community have tried to keep Maemo and replace outdated core from newer Debian (rebasing). No success yet. Many deep packages e.g. libc and glibc already work.

c)
Community have tried to start from newest Debian and bring Maemo top of that. Hildon is already running, but without GL-drivers it is not usable.

d)
There are several other non-debian distros running on N900. Arch, Gentoo.

e) ... (something?)


I vote for option C. Then Maemo would get much updates (and packages) from Debian and work will be 'only' hardware-related and hildon desktop environment.

Not so fast, it's not as difficult as it sounds. That is actually part of my plan and it should be quite feasible thanks to the excellent work of Stskeeps and co. I've had very little free time lately to spend on the project so I haven't even begun to look at the innards of libhybris although it is on my long to-do list.

On a related note, I'd be very grateful if someone can point me in the direction of libhybris documentation (I couldn't find it on GitHub, nor could I find it on the Mer wiki).

Reasons to ditch Maemo:
- Too many to list

Reasons to not ditch Maemo:
- Not many

So why not ditch Maemo?
1) Money
2) Time
3) Expertise
4) Hardware

#rant
But I tell you if I could go back in time to 2010, I would go back with the combination to a winning lottery number. In particular to those Super Lotterys of Europe with Hundreds of Millions of Dollars up for grabs. And if possible I'd also go back with a hard drive full of Linux innovations from the future. The result would be an OS and ecosystem that's as versatile as Android, but with the powerful Linux behind it. It'd have aesthetics matching or exceeding iOS 7/FlymeOS 3/Material Design. The only problems are fragmentation and having a SDK/Game Engine running. Perhaps a partnership with Valve to create a true SteamBox etc etc. Alas, I cannot change history.
#rant

regarding: "Why not create an OS that can run on devices other than the N900 so you bring in a wider community"
http://lists.openmoko.org/pipermail/...ch/069629.html
Collecting money for open software seems to be a good idea.
shr tried to craete a distro for multiple devices too, even the n900.

But in my opinion building a new distro doesnt work at the moment cause the tools arent there/ready yet (to build it better).

(In that sense maemo is a giant of its time.)

One main problem is a gui toolkit. The only one which works without blobs seems to be efl. But efl is one of the reasons why shr failed, cause shr/elementary wasnt/(isnt?) production ready. (vala wasnt ready too)
You can use qt, but qts money comes from proprietary software and so its design is "closed" to not expose property (by default/at all?).
gtk is more open thanks to gobject and it is getting better with vala and python, but has still (too?) many weak spots.

Why is Gui important especially in open software?
Cause it is a form of documentation. It can visualize your architecture and your use cases. One of the strenght of open software is reuse, so you want to make this as easy as possible. Ideally generate a debug view automatically, from your Modell to expose all your implemented functionality.
Reading source code isnt as effective as tracing a use case/example that fits your need that you in addition can live inspect.

All good Gui programms are scriptable (mc, emacs, blender, gimp, openoffice (even some fps are)) cause they give the user the most flexibility.
Part of the problem (or another one) is a good, fast, easy scripting language with a nice object modell is still missing.

Last point i want to mention creating a pim is hard. chandler failed.

The op question "why not ditch maemo and work on a new OS.?"

Baby with the bathwater.
Sure work on a new OS...but why ditch maemo?
Fremantle Porting Task Force believes it possible to make maemo what it should be.
Their work may make many avenues possible for innovation and not just for the neo900 but possibly also for the n900.

On other devices to bring in a wider audience?
well maemo is pretty close
https://www.youtube.com/watch?v=ySRZGjEs5y0
https://www.youtube.com/watch?v=F0-KQ6wPJ04
https://www.youtube.com/watch?v=RuIrRoRPufc
https://www.youtube.com/watch?v=Zt5bd2irgwM

I think once fremantle is opened up ...
the likelihood a standard computer...especially touch / penabled tablet / laptop would work quite well with it.
So yeah ...once the porting task force is done...
I think it'll only be a matter of time before I'm wandering through DistroWatch and see the Fremantle Distro.

well, as sulu in #7 and others said, the main issue is the need of these proprietary blobs, so we have to stick with an old kernel, an old userbase and so on. dependency hell is common on systems with that "dependency handling", but on old stuff this is more evident.

anyway, work can be done, at least, staying with linux
see
https://github.com/slonopotamus/n8x0.../modern-kernel
http://elinux.org/N900#Kernel_Status
http://plan9.stanleylieber.com/hardw...schematics.pdf
http://elinux.org/N800

some stuff can be reverse engineered too

our best bet is to "rebase" maemo on a relatively recent debian;
with hildon, gtk, and stuff already existing maemo programs should work

EDIT:
also, the above link for gentoo on n8x0 lists the TLV320AIC33 (which is for the wimax edition only btw), but there should be stuff for it
http://lxr.free-electrons.com/source.../tlv320aic3x.c

other useful links are
http://bues.ch/cms/hacking/n810-openwrt.html
http://talk.maemo.org/showthread.php?t=89389
http://www.natisbad.org/N810/index.html
https://gitorious.org/linux-n900/linux-n900
https://wiki.merproject.org/wiki/Adaptation/N8x0
https://wiki.merproject.org/wiki/Adaptation/N900
http://elinux.org/N800#User_space_interfaces_for_Maemo

porting OS2006/OS2008/Maemo5 DEs should be feasible, but using something else would be a lot easier, even if it could break some compatibility with programs that require the actual maemo-only stuff (which ones?)
i'm going to do some testing soon, and i'll tell you what happened then

pertaining to rebasing...
https://wiki.gnome.org/Attic/Hildon
https://wiki.gnome.org/Attic/Hildon/Building
https://wiki.gnome.org/Attic/Hildon/...ethod/Building

which I think ...if memory serves ..
the same methods of implementation are also in some part laid out in our wiki here somewhere...
So a great deal of work has already been done...in the direction of bringing hildon into gnome that is...and I think fremantle...to some degree...I can't remember...I may be wrong there.
Memory is a dangerous thing to trust... but in need ..a great friend to help