Maemo on the Pandora
 

Did not found a thread about this that really is about this topic, if there is one please shoot me :)

I wonder if there
a) are indeed a lot of ITT users who ordered the pandora would like to see maemo on the Pandora
b) if the Maemo Community Council would consider to propose to have the pandora as a device listed on maemo.org, which would allow the infrastructure to be used also for the Pandora.
c) If there are poeple with real deep "os skills" who would be interested in actually making a Maemo port to the pandora happening and make some kind of "base distro" for the pandora.
d) if the SDK would would need adaptation to support a "pandora" maemo version.

Re: Maemo on the Pandora
 

Let me know where and when it becomes available, so I can avoid that site like the black plague.

Re: Maemo on the Pandora
 

Can't say for sure, but I'd guess so.
Seems like putting the cart before the horse; as I understand it, it would make more sense to get Maemo running on Pandora, and then list it as a device...
Not sure how much deep os skills are needed. But if someone were to give me a Pandora (or if the N900 comes out amazingly early before Maemo is ported, and then I decide I like the Pandora better), I'd certainly work on it. You see, unlike Karel, I don't see Maemo as being a bad thing; I see certain Nokia-controlled elements of (currently) the only production Maemo OSes as a bad thing. While an open Maemo OS for Nokia devices should eventually be possible, porting to a non-Nokia device leaves us with a distribution guaranteed to be Nokia-free, and thus totally community-controlled.

Oh, and I think we have more software available than Angstrom, so it's also a big practical benefit.
I don't believe so. They're the same processor, so an off-device development environment would be identical, as would all non-core-OS elements of an on-device environment. So you do off-device development to get the device-specific OS working, then install the SDK straight onto it. (And then make sure everything still builds, in the name of self-hostingness.)

Re: Maemo on the Pandora
 

a.) I know I would
b.) Don't know
c.) There was a thread on gp32x.com talking about that.
d.) a little

A lot of pandora users are interested in the huge base of software that comes along with maemo and I think it would be beneficial to both communities...

However maemo is not 100% open source and some parts would have to be rewritten. Drivers would also have to be rewritten.

All in all with the coming OMAP3 NIT this would give us compatible software for the OMAP3 NIT before it is out (or closely compatible)

Re: Maemo on the Pandora
 

@ John
Yeah, forgot to mention that this was actually due a thread on the forum @ openpandora.org

@Benson, you're correct about point b) , but already asking (assuming maemo can be ported) to see if there are any (political? - not sure how much Nokia is involved in the actual maemo.org stuff) reasons.

I'm afraid I can't shuff out this month another 250 euro to buy you a pandora, sorry ;) my GF finds that I already spend enough at "toys" ...

Edit: is there somewhere a more or less detailed overview of what exactly is closed source (besides HW drivers?) that may need replacement and where there is none yet?

Re: Maemo on the Pandora
 

This isn't exactly the role of the council; but I wouldn't support listing a device which *could* run Maemo on there until it did run a feature-complete version which a normal end-user could install without problem.

As Benson said, this'd be putting the cart before the horse.

Re: Maemo on the Pandora
 

I clearly posed the Q in a wrong way, basically I would like to know if, assuming the Pandora could "run a feature-complete version which a normal end-user could install without problem.", there would be a problem for a non-nokia device. If so, then the "pandora" distro would need it's own repository/server/... Hence the Q.. :)

Re: Maemo on the Pandora
 

If Pandora wants something like Maemo, - the people should start putting interest into for example Ubuntu Mobile (see ARM port at mojo.handhelds.org) or Deblet (http://trac.tspre.org/projects/deblet). It should definately be possible to support most applications from Maemo when there's hildon libraries availiable.

For some work relating to this I've worked on recently, see screenshots at http://bsd.tspre.org/~stskeeps/nit-debian/hildlet6.png (Hildon+Advanced Backlight+xterm), http://bsd.tspre.org/~stskeeps/nit-debian/hildlet4.png (Network-manager + hildon), http://bsd.tspre.org/~stskeeps/nit-d...let-mobile.png (Ubuntu Mobile packages on Deblet).

Debian and Ubuntu are already ported to armel decently, so it's just a matter of people starting to help out with making packages, environments, etc for the tablets. But there's a lot of work to be done, power consumption problems, getting the stuff user friendly, and so on.

But I guess the new open devices will help this development..

Re: Maemo on the Pandora
 

I was just going to post that I think it would make more sense to see Ubuntu Mobile on Pandora, than Maemo.

Re: Maemo on the Pandora
 

Sorry to insist but I'm a bit confused now, seen Meamo is already (IMHO) rather "user friendly" and great with power consumption (on my N800 anyway) so why then not use this instead of trying to make those other distro's more "like meamo"?
Or do I miss something obvious?

Re: Maemo on the Pandora
 

Well, first off - Maemo doesn't really give the choice to run other environments other than Hildon. Two, Nokia has said that next version of the OS may not support the N8x0's - much to the dismay of the large userbase - so they'd be left in the cold regarding system updates. Three, porting to Maemo is notoriously difficult occasionally because of the fact they haven't bothered to follow up with updates Debian and deriatives have kept up with, and broken many APIs, which I believe harms the platform. Four, - because it's fun to work with and play with - a system you can completely modify to your will. Isn't that something?

Re: Maemo on the Pandora
 

maemo.org runs more and more independently as the community takes control. I can't answer for the council part, but be sure that Nokia has no problems with any kind of project attempting to run Maemo in other devices, as there is no problem with any projects attempting to run OSs different than Maemo in Nokia devices. In fact we like to see cool platform hacks around!

[/QUOTE]Edit: is there somewhere a more or less detailed overview of what exactly is closed source (besides HW drivers?) that may need replacement and where there is none yet?[/QUOTE]

https://wiki.maemo.org/Task:Components_and_packages and specifically https://garage.maemo.org/docman/view...s-20080725.ods

It would be really useful if someone would find specific and unavoidable showstoppers to run Maemo in OpenPandora or any other hardware.

It would also be really useful if someone would find specific packages (or versions of packages) becoming a pain, and the reasons why.

Bugs and enhancement requests in this direction would be welcome. A first step is also to ask why the affected packages are closed: http://wiki.maemo.org/Why_the_closed_packages

Re: Maemo on the Pandora
 

Thankx Qgil !
A quick glance at the "why" is already telling me that "power management stuff" is closed source , so basically any port of Maemo would need to rewrite that... And Is see also that "most of the OS2008 UI is closed source"
Looks like already a rather big hurdle to me and basically would not leave much left from the "maemo experiance" i have now on my N800...

Re: Maemo on the Pandora
 

But is the "power management stuff" even essential in a port? That's basically a kernel part, isn't it? OpenPandora would use its own kernel version I think. (I looked at the specs again.. 4000mAh battery! The power management part can probably wait :D))

Re: Maemo on the Pandora
 

Well, I know a few people have Maemo booting on the Beagle, so the issue is mostly finagling the hardware-specific stuff out and figuring out how to set up the initfs.

I did as much as using the TI reference kernel (2.6.22) to try to boot a backup of Diablo I had on the SD card and watching it fail miserably since the initfs wasn't there.

I'm pretty sure there's at least one person on #beagle who has done it, so you may ask around there.

Honestly, though, Stskeeps is right. You're better off just using Ubuntu or Debian (though neither of them are optimized for armv7), and simply setting up the Hildon desktop environment on there (since that stuff is all open source anyway). This approach is actually going to be even better moving forward, as even the tablet-browser-ui (or whatever actually replaces) it will be open in Fremantle.

Re: Maemo on the Pandora
 

Thank you all a lot for the information...
The message is clear :) It makes more sense for now to put effort behind Ubuntu Mobile or deblet or Angstrom...
And even if you get the initfs stuff working, from you're last comments it looks like it might even make a lot of sense to wait for Fremantle for the UI part...

Re: Maemo on the Pandora
 

Does that mean, that you don't intend to go further with maemo4 on pandora and instead wait for fremantle? From reading the roadmap that will take a fair while. Would it be possible to integrate hildon into an other distro to be able to use maemo progs(microB)?

Re: Maemo on the Pandora
 

Hildon is already packaged for Debian and Ubuntu (Ubuntu MID is using Hildon, also see johnx's beta2 release of Debian), it's just a matter of stitching everything together to get a nice, polished desktop environment like Nokia has done.

Of course you'll need to substitute in a lot of the stuff to replace Nokia closed-source UI-differentiation nonsense (various taskbar, statusbar, and home applets, etc.), but it gives you much better compatibility, and more versatility.

Take the good of Maemo without all the bad.

Re: Maemo on the Pandora
 

So a brief perusal of the spreadsheet linked above shows that maemo running (legally) on 3rd party hardware would be missing quite a few things that are included in OS2008:
- microb web browser
- image viewer
- media player
- notes application
- sketch application
- contacts application
- most of the control panel
- most of the status bar applets
- most of the desktop applets
- most of the themes
- various system dialogs and notifications
- others likely, this is just a short list.
Also remember that the following are closed source and likely legally tied to the Nokia tablets as well:
- skype
- gizmo
- flash

At that point you have to ask yourself whether you're better off spending time filling in those gaps with open source stuff or just bringing over your favorite maemo apps to another desktop.

-John

Re: Maemo on the Pandora
 

The important thing for me would be to have microB on pandora, because I think there is no faster handheld-browser out there(yet). I own a N800 and never would sell it, because I would still use it for certain tasks. So I think, I would have the right to use the closed software on every device capable of running it, as long as I have the license, which I've purchased along with the device itself.

Re: Maemo on the Pandora
 

Depends how much stock you put in EULAs, fair use, copyright law in general, etc.; you certainly won't catch much grief from Nokia doing that, but if someone were to distribute whatever parts of Microb are proprietary (I'd been under the impression it was open, BTW), there'd be a lawyer fight over it, even if there was a disclaimer telling you not to grab that package unless you own an N8x0... In practice, this means it can't be distributed.

As for it being the fastest, I'd give it credit as the fastest commercially-backed one, maybe. But check out a lightweight webkit-based one like Midori or Tear (alpha warning!), and you might change your mind, especially if predicting forward to when Maemo-on-Pandora would happen.

Re: Maemo on the Pandora
 

Be careful with your terminology. MicroB is free and open source software, but MicroB is only the engine, it's not the whole browser. tablet-browser-ui (or the front-end part of the 'Mozilla based browser for Maemo'), on the other hand, is closed (but will be replaced with an open version in Fremantle).

Re: Maemo on the Pandora
 

That's the same front-end as was used for Opera in OS2007, right?

Re: Maemo on the Pandora
 

... or requesting relicensing of the stuff that is really interesting to you. The current licensing of the components developed by Nokia responds to the current reality and demand. If you challenge that reality and come up with solid demands we will consider all requests.

Also about the Nokia owned components, I'm not a lawyer but is just common sense that the company won't sue you for experimenting with Maemo components in other platforms. Personally I would see it as a proof of the interest component X is having out of the Nokia devices or the Maemo platform and a strong argument to consider the relicensing.

Again, we are interested seeing people experimenting with Maemo and its compatible devices. Licenses are there to avoid legal or business misuse, but within the terms of experimentation and fair play they shouldn't be an obstacle.

Re: Maemo on the Pandora
 

Not entirely, but closely related.

Re: Maemo on the Pandora
 

From Nokia's point of view, they want to see an example of what is possible before they invest time and money in opening up code and development processes.

From my point of view (and I assume the view of other hackers/devs), I don't want to sink a lot of time into a project that might be doomed from the start. My three big concerns are:
- licensing issues
- not having a say in the direction development goes
- having no recourse if Nokia stops maintaining said closed source code.

The situation really breaks my heart in a way, but it seems like we're in a catch-22.

Re: Maemo on the Pandora
 

Come near little guinea pig, have you some time to spare :)

Sorry your avatar was calling to me :D

Re: Maemo on the Pandora
 

Working for one of the biggest software company's around I know that there is a big difference between what "tech people" "feel" and what the Legal department "feels". I'm not doubting your words, but I would not feel comfortable unless there is a clear message from nokia inc (tm) that they grant a licence to use any nokia owned Meamo libs etc on non-Nokia hardware. Of course there might be restrictions like not for commercial purpose/no resell or that it cannot be used preloaded or <insert your favorite restriction here>. But I have a (healthy AFIAK) distrust in any legal thing seen I don't known enough of that gibberish stuff.

And it would then also be a bonus to know what/if there are things in maemo closed source that NOT nokia owned but licence by nokia from others to use in Meamo on the nokia devices. As this is not nokia to give...

I simply try to sum up what would go against this IMHO, like Johnx did:

* Nokia has parts of the Meamo stack as closed source because they (and it's their right - I'm not debating that) consider this as the "competitive edge" for the nokia devices. Fair enough, but part of this seams to be in a area where the real advantage of using Meamo would be like in power management or ARM optimisation, so there is not much incentive to actually use the current maemo stack without those parts over other currently non-optimised distro's. Unless Nokia would put engineer resources in helping out with problems on their closed things on a non-nokia platform, which I frankly doubt.

* Parts that are now closed source (that gui-brower ui thingy part for example) will be replaced in a future version with a open sourced stack - so there is not much point in rewriting the current closed GUI part for example as it will be obsolete in the sense of not following meamo on nokia, however it's not clear when this will happen, what the roadmap is and what exactly this includes.

Bottom line is that IMHO if Nokia is really interested in having people messing around with maemo on other non-nokia devices licensing should be at least cleared up.

Also it would be nice if it was clear if Nokia would make any kind of commitment whatsoever to help out/adapt if needed (I do completely understand that Nokia would not "support" other devices,more thinking in the line of some kind of "best effort" or so) on the parts they control (or that this never would happen).

Or to open up the sources for closed parts that would be not be maintained when a new OS does not support the HW anymore. This would be also possibly tremendous beneficial for N800/N810 owners once these devices are out of the nokia "lifecycle"..
And would not really conflict with the "competitive edge" idea, seen older hw is by definition only a cost to support.

And then we still have a bit of a "roadmap" Q's of course :)

Just my 2cents of course and mainly thinking out loud...

Re: Maemo on the Pandora
 

There are many interesting points, all of them deserving a good discussion. Some evening ideas:

- Don't be too obsessed about "competitive edge" (others call it "differentiation"). It is only one of the many potential reasons to have a package closed - http://wiki.maemo.org/Why_the_closed_packages . I bet "legacy" is as heavy, and relicensing in those cases is almost a matter of priorities, time and resources.

- A lawyer will say what the license says. A lawyer won't say anything contradicticting a license. If you need a paper signed by a lawyer to start hacking then definitely you'd better invest your time in something else. However, investing it challenging the current Maemo platform you will contribute in its improvement and evolution. Even our lawyers understand that.

- Then again, in the short history of Maemo licenses have been... "challenged" sometimes by some good hackers for the sake of experimentation and research of alternatives. Ask them about the feedback or treatment they got from Nokia.

- Licenses are texts that can change when the context and priorities change. For instance, hackers interested porting your component to another platform means probably a different context and perhaps ven a diffrent priority.

- The Maemo SW team is a busy one and opening components takes a significant work even if you don't touch the code (which normally is not the case). We wouldn't go through a massive relicensing only to see if there is an interest. Most of the times is the other way around: first there is an external interest that can cause an internal reaction, the result of which might be the relicensing of what really matters.

Oh, and the spreadsheet at http://wiki.maemo.org/Task:Components_and_packages shows also the packages not owned by Nokia.

Re: Maemo on the Pandora
 

cheers, missed the 3the party stuff in the odo file...,
for the "law" matter, I rest my case for the moment, after adding a link to a certain blog about "key principles I hold dear when developing Linux based products around the maemo.org." of a Nokia VP to the discussion :)

Re: Maemo on the Pandora
 

Ari also said during his keynote at OSiM World that Nokia had no interest in licencing/working on getting Maemo onto non-Nokia devices. It's not yet-another-"standards"-body to prevent fragmentation: it's Nokia's mobile Linux stack.

As Quim says, that's not to say that hacking on a Maemo firmware release to get some bits running on the Pandora won't be quietly ignored. But I can't imagine, say, Adobe being quite so friendly if the Flash player gets extracted out of repository.maemo.org (the password's not hard to figure out) and run on another mobile device without them receiving any money.

Re: Maemo on the Pandora
 

Or better the Mojo folks have Ubuntu packages optimized for different ARM processors.

Re: Maemo on the Pandora
 

Which does not (yet) include armv7. . . .

Re: Maemo on the Pandora
 

But a improvement to the ARMv4 from Debian.

Re: Maemo on the Pandora
 

This looks like my answer to my question from another thread: could Maemo 5 be made to run on the Motorola Droid? I take it the answer is "no". Could any other Debian-based OS (Mer, etc.) provide good functionality (including phone functionality) on the Droid?

I had become convinced that the arguments over which OS is more open, Android or Maemo, had been decided in favor of Maemo. But I must say that there are stronger points than I had imaged in favor of Android, and this seems to be one example of such. Android runs on a lot of different devices by a lot of different manufacturers.

But still, for some unknown reason, I tend to just "feel" that Maemo is more open. Maybe because it is more like desktop Linux...

Re: Maemo on the Pandora
 

Depends on what you mean by "could".

Technically? Certainly. In that regard, the answer to "could it" is yes.

Realistically? Why would Nokia license Maemo out to be run on a competing device? Don't bet on it happening any time soon. In that regard, the answer to "could it" is no.

Mer? Sure. But Mer doesn't have phone functionality, last I checked. And the same for other Linux variants (Debian or not).