I had a bad exerience with compiling some very easy softwares on-device (the SDK could not be downloaded, it failed every time). The reason was the dependencies. Or else it was just follow the instructions and you have a peice of software for your device. If our favorite devs rebase Maemo with new Debians, I don't know what else could be achieved.

One question though, if the dependencies are not available on device, they won't be on the SDK either?

These gents have succinctly put the problems at hand.
Esp. sulu's last post.
I get you are new...and having fresh perspective to some long standing hurdles is always good.
I also get the angle you are coming from concerning a new / diff. os for the devices. I've played around with enough on the nxx0's to enjoy them for what they are...love 'em all...the trick is though..the kind of os maemo is...it is perfectly optimized for the screen size of the devices. whether it is the 770, n8x0 or n900.
To get the hurdles facing the n900 ...
look to the n8x0's ...pretty much the same probs.
as succinctly stated by sulu...the proprietary bits (convenient at the time for Nokia) got in the way eventually.
And where alt's like Mer, Nemo and the like hit their "wall" was not just in the proprietary end of things. But they also realized they were attempting to do an alt. os for devices that were losing their numbers of users in favour of the "latest shiny" . Not encouraging for starters. Not having enough brains or time focused on a problem then can have an impact.
Then such endeavours as Mer and Nemo can also result in the devel's great and honest question "We could be making a living wage doing something else..."
There are quite a few alt. os's for the n900 that work decent. Some not so good.
For the n8x0 the total list of os's tentatively attempted,functioning, ones that need work, or hit "the wall" is epic.
Angstrom - Arch - Atari800 - Deblet - Dos - Easy Deb - Garnet - Gentoo - Funtoo - Kali / Backtrack - KDE - XFCE - Basilisk (mac) - Meego - Mer - MSX - Openwrt - Windowmaker - Ubuntu - Opensuse - Mamona - Poky / Yocto - Nitdroid / Android - Fedora - Inferno..
that's off the top of my head that I can immediately remem.
I keep feeling like I'm missing some.
Looking at what the many with n8x0's worked at and the hurdles I see the similarities to the same kinda probs facing the n900...
trick is this
I don't think it is deja vu ...
'tween backporting and finding methods to have a more "open" os on a more open nxx0 device I think is the difference today with trying to solve said probs.

You probably like Debian for its (mostly) well-working package management. And although Maemo uses the same tools, it's a totally different story here. Debian's ease and stability doesn't come with the tools (alone). It comes with the strict policies of when a new package is accepted and when not.
Fremantle's community repos may have the extras/testing/devel segmentation that may remind you of Debian's stable/testing/unstable, but it's not comparable - mostly due to the lack of proficiency that almost ineviteably comes with a very small community of volunteers.

Fremantle is a fork of Debian from a time when Squeeze was testing. It's essentially a Lenny/Squeeze mix, with some parts even dating back to Etch, which were apparently carried over from an earlier Maemo version.
In every Debian community you'll be asked to clear that mess up before you ask for any support. Here you'll just have to live with it.

Are you actually interested in a new OS on the N900 phone or are you just unsatisfied with userland development options of the N900 as a pocket computer?
If it's the latter, then seriously, have a look at DebiaN900! If you can live with the fact that the battery runtime won't be much better than on your laptop then this is a good way to go.
If it's the former, well then I have no real idea, except for maybe switching to a more mainstream phone.
If it's a mix, Easy Debian [1] might be a compromise. It's a Debian chroot (with X) on top of Maemo. But it's mostly geared towards desktop use (iceweasel, libreoffice etc.). Development should be possible, but I gave that up due to the N900's limited performance and it's watchdog, which will never become my friend. I even went as far as getting a Cubieboard 2 to create the images.


[1] http://talk.maemo.org/showthread.php?t=34550

I was about to mention "Why not create an OS that can run on devices other than the N900 so you bring in a wider community", but I just realised how difficult that would be. So scrap that.

Re: the mobile layout. That wouldn't be too difficult to achieve with some learning, and practice. And a quick Google search revealed mobile window managers such as: https://www.enlightenment.org/p.php?p=about&l=en

I might be completely off base here. I'm just thinking logically. I'm the sort of person that sees the N900 as a computer, independent of any OS and sort of ignorantly looks past potential problems. That's why I'm thinking, something must be possible here.

I mean, we're a small niche really, with a passion for the devices that are no longer in production.

It's actually kind of frustrating lol.

I meant Linux in general, I actually switch between Arch and Debian on my two Thinkpads.

Ah, I assumed it was Squeeze, but jeez thats old.

A new OS yes. But not for any purpose other than I love the device, I use it mainly for pen testing, but also for programming on the move. Challenge myself to puzzles when out and about. As I said previously, I'm just seeing this small "computer" (not phone or tablet. I've never put a simcard in mine) that shouldn't be locked down like it is. Again this is completely personal. The challenge is what makes me want to do it, and the fact I'm going to learn a **** load of useful stuff that will aid my career/hobby along the way.

Android is "Linux" too.

You want DebiaN900! It's as simple as that.
I still use my N900 as a phone, which is why I still use Maemo. But if I wouldn't, I would have already replaced it with DebiaN900.

If you're also looking for similar small computers, have a look at the Dragonbox Pyra [1]! Just as the Neo900 it's sheduled to be released this year.


[1] http://pyra-handheld.com/

enlightenment is grand.
meant to work on older hardware / low resources.
love it on the n8x0's ...
love it on my laps.
Bodhi rulz.
more d.e -ish. than distro -ish for the n900...
hm...look to illume on the freerunner for a good working example of it on a telephony enabled device...maybe you'll see something plausible / viable I immediately can't.
hm actually you could use e17 on your n900 as your d.e. in Easy deb or 'buntu ...
In fact if memory serves ..didn't the guy running bodhi os have a n900 and wasn't he trying / quasi-successful in getting it to run?

For me the issue is outdated libraries, plain and simple. It is preventing ports, making others far more difficult and causing me to have to test a number of different work arounds for each package. Not to mention Microbe, my Qt4 embedlite based replacement for microb, having to go on hold because upstream switched to Qt5.

Council is looking into suggestion I made about getting GCC in autobuilder/extras that is complete, ie I don't need to use gfortran from 4.6 because 4.2 and 4.7 are not included as part of the toolchain. This will help a bit, especially if we can get all targets (Extras, CSSU, CSSU-Testing, CSSU-Thumb) using the same compiler suite, simplifying packaging for features missing prior to 4.4. The other issue I keep facing at the moment is previously defined functions which means forcing it to use C89.

My project to create an updated version/clone/whatever of Maemo is miles off, at a snails pace and a seemingly impossible goal at the moment. Hell, I haven't even built one armel/armhf package yet as I'm still using amd64 chroot to test. The realisation is that the number of people who may be interested in it is not going to be sufficient to build and maintain all of the OS, which is why I've opted for a litttle....help. By making use of some of the Mer/Nemo software, all I have to do is repackage it for Debian. Then I can build around that, in a similar way Sailfish does. Even with all of this "help", there still remains the issue of bringing up to date everything that makes it Maemo.

A little example: Matchbox2, the underpinning window manager for hildon-desktop, uses a Nokia customised version of clutter-0.8 and Tidy as a UI toolkit. We have clutter-1.2.something IIRC in Jessie, which doesn't have Tidy updated for it. Luckily there was Cordia, clutter-1.x port with Mx toolit. Er.. Mx was originally used in Moblin, then Meego, then nothing development wise since 2013. The only suitable solution I've found so far, as in actively maintained, is ST (Shell Toolkit) from gnome-shell. So now I have to look at porting to that.

But (sorry, I know it's not good English to start sentence with but) as I think I've said in the DebiaN900 thread, the work done by Aapo, Wicket and myself with Debain and the FPTF work on Fremantle will all benefit each other. As xfce-4.12 shows, gtk2 isn't going away just yet, so maybe someone will port maemo changes to the latest version of gtk2 and knock up a version of Hildon using that, along with clutter-1.x and ST from my project. Maybe my work porting mce etc from Nemo will help Wicket get some better power management.

To sum up, there are some looking to maintain, there are some trying to remove the closed-binary obstructions and there are some trying to take this and look at the options for future development. There's a lot of people who like Maemo so I don't see why it can't continue in one way or another.

polishing qt5 compiling pyotherside as a headsup , idon know
wiki.maemo.org/Qt5-Maemo5
talk.maemo.org/showthread.php?t=84782

...because we are the Maemo Community and we are all here to keep Maemo going.

The way I see it if all the closed parts of Maemo are replaced and everything is open w can then see how to move forward. The FPTF are working towards porting Maemo to mainly the Neo900, when this happens we have a base to start from.