lma | # 31 | 2010-08-12, 07:52 | Report

In this case they actually did, but never got around to releasing it. I would settle (source would be nice, but I'm not holding my breath) for a binary and some way of generating updated initfs images with it.

qgil | # 32 | 2010-08-12, 17:59 | Report

A sequence that might help you understand more the current situation:

1. At some point I explained in maemo.org that opening legacy software takes resources appointed elsewhere, and in some cases we would do it if someone would push and do the actual job.
2. The Maemo Community Council came up with a proposal for the role of the distmaster, proposing Carsten (Stskeeps) as the ideal guy backed by his work in Mer.
3. Tero and me did the numbers and agreed with Carsten to fund his role as maemo.org distmaster.
4. Carsten starts his work. He gets access to some old and fresh code, plus Nokia internal contacts and info. afair he gets involved in a report explaining the current situation, helps on the Diablo community updates initiative, on the closed graphics drivers for the N8*0 (stellar work done by him behind the doors), the licensing queue announced in this thread, etc.
5. Then MeeGo is launched, picking by surprise everybody. I can't remember now, but I believe Carsten was not aware at all.
6. Yet he is one of the most skilled guys around to start playing with the N900 port and the MeeGo OBS. He is offered more work and responsibilities within the MeeGo project - basically a full time job.
7. In the meantime both the core Mer guys and Nokia people like Tero or myself get convinced that the best technical and tactical way to address openness in the tablets and the N900 is to follow the MeeGo development mainline, adapting it to this hardware. Nokia is opening a lot of functionality for MeeGo Handset and the MeeGo project has chosen the N900 as official ARM platform. The current is favourable in that direction and Nokia is putting serious resources in that direction with the MeeGo N900 port.

This is the situation now, more or less. We are again in point 1, but with a big difference: now there is a MeeGo project officially backed by Nokia that is all about open development of open source components.

Tero and I could sit again and see if there is budget for a new maemo.org distmaster that would take care of legacy closed components... But honestly, if I would need to make that decision I would recommend to put those resources in someone helping to push faster and better the relicensing of closed components from MeeGo stakeholders, and also real and effective open development from the companies joining MeeGo development.

I understand this doesn't bring the short term benefit that fixing a Diablo/Fremantle binary blob brings, but the actions (specially when they involve money and resources) must be taken with a wider perspective.

inidrog | # 33 | 2010-08-12, 18:12 | Report

Aha. I see. The sooner i ditch and forget the N900 and 500$ and get a new MeeGo device the better. Because "this time" open = open...

Edit:

Just forgot this, sorry

http://tools.ietf.org/html/rfc1149

---

Last edited by inidrog; 2010-08-12 at 18:24.

SD69 | # 34 | 2010-08-12, 18:38 | Report

And at some point, I explained that Diablo and Fremantle were always presented to the community as being open source projects. Granted it is difficult and takes resource to open up closed source code, but the situation here should not be that difficult, because openness was an original goal, right?

The other thing that concerns me is how this openness approach got flipped to have a different rationale. The quintessential open source license is you publish the source, everyone is free to come and take it without getting special permission, and the only requirement is that they offer their source on similar terms. This is where open source communities shine - an organic process that lets people take the source and do amazing things with it that perhaps can't be imagined much less enunciated in advance.

At one point, starting with the "why the closed components" wiki page, Nokia attempted to impose requirements essentially saying that "if you want a component opened, you have to bring justifications of why we should do so and convince us that it is beneficial ." I can understand why you would do that for business reasons, but that is not an organic, open source approach and it is unfortunate that things got twisted that way. I am trying to say in a nice way that Nokia should try to set things right.

Thank you for considering this. I noticed with great interest the $50K donation to gnome. But they didn't seem to want to use it to support the effort of Conny and others to bring Hildon/GTK+ to MeeGo. That (and anything else that helps port Diablo and Fremantle apps to MeeGo) seems like a good way to move on, worthy of consideration, and consistent with the long term goals of the Diablo Community Project which I support. What do we do next?

Stskeeps | # 35 | 2010-08-13, 08:08 | Report

Just wanted to thank for this part - I'm pro getting a proper Hildon available for MeeGo and am frustrated to see nothing is getting done about it as it is entirely possible. With a budget for it that isn't getting used for anything, it might even inhibit others to do something about it as their work would be ignored once a paid worker starts doing something.

Matan | # 36 | 2010-08-13, 08:11 | Report

Can you please move MeeGo discussion to relevant thread, and use this thread for its original purpose: discussion of Nokia opening closed Maemo code?

Tomaszd | # 37 | 2010-08-13, 08:42 | Report

See, this is the place where the community (and the thread itself) goes its way and you go your own.

We are not addressing openness here, this is not the issue we have raised. Besides, what else do you want to open in the N900 profile for MeeGo? 3D graphics or battery charging, because that's the only two things left from what I understand. The rest will be opened, many people are working on this. That's fine and I'm happy about that.

We just want specific Maemo 5 and Maemo 4 components opened. I don't want to hear about MeeGo in this thread again. It's a smoke screen and I know this because nobody is going to convince me that hadn't MeeGo emerged and Maemo 6 been released instead that things would have been *any* different. As someone at IRC put it, it's "ADD marketing", but I like my way of calling it a "burned bridges strategy", also known as "oohh shiny!".
Please don't try to tell me that MeeGo's supreme openness prevents Maemo components from being opened, because that's how it sounds.

Carsten has been relegated to the N900 hardware adaptation team. I want MeeGo to be a great, fully functional platform for the N900. I'm far from blaiming him for anything here. I blame whoever decided it would be a good idea to not have a person that was doing what Carsten was doing previously any longer. That person is sorely missing.

MeeGo is nowhere near Maemo 5's functionality at this moment in time and surely for many months to come. And I'm not talking about the mythical MeeGo-Harmattan project, I'll believe it when I see it, what I saw was reference handset UX, which is one big placeholder.

Let's say I don't want to experiment with MeeGo, I just want a more open Maemo.

lcuk | # 38 | 2010-08-13, 10:42 | Report

Hey, my current role in the as Pre-Integrator in the Maemo Application Framework team puts me in a position to see both sides of the coin.
I have integrated and worked with the open and closed aspects of the system for a while now and have a high level of understanding of the complexities.

There is a crossroads coming in my current role since obviously the official teams are scaling down.

One of the principle things I have said over the time I have been here is that I do not mind closed source code as long as it is actively maintained.

I am passionate about Maemo and would be happy to discuss adding this feather to my cap and trying to open up as much of the stack as possible to allow the community to continue to maintain and build on our Maemo.

I do however understand that certain pieces of code will never be open, so on that basis would hope to continue to push the internal teams to offering important security fixes as they arise.

I do not want to just open the code and let it rot though.

It would be even better if the community improvements and advancements could be integrated into further "service packs" so that we gain the best of the best.
The community SSU created for the n810 shows that this can in practice work.

It has been shown already that variations and branches of the already open code is worthwhile and people in the community are happy to fix bugs and submit patches and there is activity in the Hildon components as well as many other areas - even as far as custom kernels.

Where do I apply?

konttori | # 39 | 2010-08-13, 11:50 | Report

I support Gary on this one 110%

lcuk | # 40 | 2010-08-13, 12:13 | Report

For a first test, I have submitted a request for opening the Sharing services framework

https://bugs.maemo.org/show_bug.cgi?id=11128

Bug 11128: Opening the source for Sharing services framework

I have heard from different people who have tried to implement extensions for this important aspect of the Maemo system for a while and recently albanc has gone to extreme lengths to actually implement a new button on the sharing dialog.

This kind of hacking that he did (as amazing as it is) should be totally unnecessary if the code was open.
People have at different times requested alterations and minor tweaks to this but have not had the skills.
Have a read of albans blog about how he did it in monorail

http://alban.apinc.org/blog/2010/07/...-on-your-n900/


you would be surprised at how much hard work adding one simple button would be!

Yet again, this shows open source will always find a way to do things, but I would like to make it as easy as possible for others.