Open source vs confidential products
 

Why Nokia keeps their products secret while pushing an open source approach in Maemo? The question has been raised in a way or another several times. And it has been answered also many times in thos threads.

The last time at http://www.internettablettalk.com/fo...&postcount=221 by eiffel. And the GA said:

Actually is not even that.

Open source is about software but most of the criticism towards lack of openness to Nokia in e.g. the thread linked above goes around hardware aka products. Nokia doesn't aim to translate the open source principles to product planning and marketing. The reason is clear: until now Nokia hasn't done bad selling products, and even if some competitors are selling also well openness seems not to be something key in their strategies.

Then there are projects that have taken an open approach when producing hardware (OLPC, OpenMoko, OpenPandora, what else). As interesting as these projects are or have been, their open approach hasn't brought them a striking success. Yes, there are many reason to that and any comparison might be unfair. But you see why the people in Nokia deciding how to invest the budget and plan the marketing and sales feel comfortable with the open source model for software development, but no for product planning and marketing.

And this is one of the reasons why Maemo is quite open (at least compared to direct competitors) when it comes to disclose and discuss about platform details relevant to developers, but less about end user features and even less about unannounced device products.

Thank you for your understanding.

Re: Open source vs confidential products
 

I think the problem is that it has been soooo long ago we had any hw news and that's frustrating for some people. :)

Re: Open source vs confidential products
 

This is exactly it. The community is stuck with a dead platform (Diablo), and obsolete hardware. You can only leave users with nothing to use and developers with nothing (the alpha SDKs don't count) to develop for before the frustration boils over into the endless bickering we've been seeing lately.

Sure, but let's not pretend that there aren't plenty of complaints to level at Nokia's behavior towards software. ;)

Yes, these are unfair comparisons. Clearly you can't bank on open source and open hardware as your only asset (OpenMoko certainly doesn't have much else going for it), but there's not enough evidence to support the validity of the approach one way or the other. I'd certainly love to see Nokia become one of the first big players to try it out, though.

Re: Open source vs confidential products
 

The way I see it, the NITs are pretty much an island, there is practically no alternative today if one wants truly pocketable openness (with android devices just emerging on the horizon). That's why many people (myself included) want the next generation so bad to succeed and at the same time fit whatever expectations we have. As yourself have pointed out, this time around the announcement will not be so much before the actual release. The extra speculation is just the flip side of that strategy (with the long announcement having it's own downsides, od course). Also, maemo is amassing more and more users who are not tech savvy enough (or simply don't have the time) to play with the SDK, but enjoy all the blessings that maemo and linux in general can bring - for them, the wait is even harder as they have nothing to do but sit around and wait. And idle hands are the devil's tools on forums :)

Re: Open source vs confidential products
 

not likely, the guys in suits running the show is probably trying to wrap their head around open source software. trying to be more open about hardware probably have them scared silly...

Re: Open source vs confidential products
 

Besides, the line between "software" and "hardware" is blurry at best - they're intimately tied to one another, and neither is useful in and of itself. Hence, releasing software details without reference to specific hardware paints a pretty vague picture. And as we've seen, a lot of people aren't encouraged by "vague".

That said, I suspect details will soon become clearer.

Re: Open source vs confidential products
 

indeed, from what i could see over on the other thread, most of the frustration came from details around inputs, something that relates to how one should expect to deal with interface design and so on.

not knowing fully what inputs will be available could lead to a whole lot of wasted effort in programming something that expects something that will not be available in the final product(s).

Re: Open source vs confidential products
 

Who said that open source has to be openly developed?

You might say it goes against openness and the community to do that and for certain projects I agree.

we both know from experience now what open community development does for a large project and its not pretty.

Gan, for a personal example, think about the nightmares you had about app categories, and multiply them 1000 times.

Being secretive and developing behind closed doors allows a group to focus on the vision and spec without ending up with a stinking pile of committee driven code.

sure, at the end of a cycle the code produced still might not satisfy a hungry crowd, but it fulfilled the original design objectives and tick all the right boxes.
The alpha and beta stages are meant to facilitate discussions about changes in direction.

I believe the same goes for hardware, but the "community" cannot really do most of the alpha and beta testing, but I'm sure some folks have done already and will continue to in the future.


as for the dead hardware, I certainly don't think its dead and as community council member for maemo.org general thats a pretty harsh thing to say.

Re: Open source vs confidential products
 

It take issue with that. My 770 is better than anything Nokia has released since, and, from everything I'm hearing, their future products are likely to be worse.

Re: Open source vs confidential products
 

There's not much point in it being open source if all you get is the occasional batch of code tossed over the wall.

Being open doesn't mean you're handing the keys over to any idiot with version control. It simply means you're talking about your plans and about what you're doing in a public place. It does not, as you seem to think, require you to sacrifice your creative vision to please the mob.

Re: Open source vs confidential products
 

There once was this project, you may have heard of it, called Linux. They somehow figured out how to do this.

We have some massively successful OSS projects out there, with a process that is amazingly effective. It's like Nokia looked at all of them and said, "Not Invented Here" and completely avoided learning anything from them.

He's right though. Neither as a whole nor any remaining component bugs will be fixed. MicroB, for example, has had some annoying input bugs since the release of Diablo. And how do you enjoy Modest's rapid pace of development (HTML messages anyone)? For a NIT these are significant portions of the whole devices purpose of existence.

Re: Open source vs confidential products
 

all in all, designed by committee may be bad if its a government committee, but open source ones seems to get along, at least to some degree ;)

Re: Open source vs confidential products
 

There are open source projects and there are open source projects. I've seen projects where the core developers literally went out on a limb to help not only fellow developers, but users (yes, even n00bs), too. On the other side, I've seen projects that managed to sink to the level of death threats :confused: Maemo, for better of worse, is somewhere between the two IMHO. So, there is no universal recipe for open source. The fact that a development approach (within open source) worked for one group does not mean it will work (equally well) for another. It's a very biological thing, it has to find the right mix of people, management and resources for a particular project to function as a whole.

Re: Open source vs confidential products
 

Linux was not developed initially by a community, it was grown by one person.

There are very few projects that go from inception to completion in the hands of the greater community offering patches and tweaks.

Re: Open source vs confidential products
 

It's been often said that it's not possible to give info about the devices in advance and it does seem to be the name of the game with all vendors, but I think it might also help people tolerate the situation better if somebody explained in detail why it has to be that way and what would happen if hardware would also be developed more openly. I have the impression that one key factor is that the competitors would have more time to react to new products, but I suspect there are other reasons at play as well?

Re: Open source vs confidential products
 

Huh? Not the Linux I know... it started out by one person, but he put up sources immediately and as far as I can tell, accepted code since version 0.02. It was *managed* by one person, but I seriously doubt he could have done it all on his own.

See also: Debian.

Both of these examples throw the code and binaries wide open, but have some very focused controls at the *official* release point.

I have good hopes for Mer, though.

Re: Open source vs confidential products
 

snoo, he wrote the kernel himself before announcing it.

as noted here from his own posting:

the project seed was in place, what took off beyond that was the backing of the greater community.

https://netfiles.uiuc.edu/rhasan/linux/

Re: Open source vs confidential products
 

Let me insist that this is not a thread to discuss how well Nokia understands or developes open source. Please open a new thread if you want to discuss that. This thread is to discuss why Nokia keeps using regular 'closed' strategies for planning and marketing device products also for the Maemo platform.

I'll tell you the same I tell to myself: this is easy to say when it's not your money and you are not accountable of the losses if the strategy doesn't work.

How many Nokia shareholders are in this thread? How many are managing investments of many zeros in your jobs? A device program is a complex and expensive game, and marketing a device produced might be even more tricky - specially nowadays. If none of the big companies have tried and the smaller that try go the way they go... would *you* really be the the first top manager taking the decision. It would be interesting to see the posters of this very same thread in that situation.

Re: Open source vs confidential products
 

Yes, I can understand that. It is also understandable that no matter what some people will fill any silence with assumptions of Nokia's wrongdoing, blindness, deafness and what not. :)

Probably what is sometimes shocking is the combination of love for the known products (most of us like the Internet Tablets and this is why we are here) combined with a consistent mistrust and fatalism to whatever is unclear about the future. You know this is a progression and the steps done with the 3 previous devices can be reasonably called successful. Well, you can expect by default success with the next step - unless you have tangible reasons to fear otherwise.

As an example there was that guy saying that if nobody of Nokia had commented on [fatal rumor X] it meant that the rumour could be confirmed. Interesting conclusion. Why not taking it the other way around: since September 2008 we have announced some features that show a trend and a future. Most of the relevant changes (and definitely most of the changes relevant to developers) have been announced already. If we haven't said anything about [feature X] is perhaps because there aren't changes worth mentioning, or no changes at all.

Going back to the core topic of device announcements. As many of you are saying in several places, announcements are just one part of the game. Sales starts, price points, countries covered, quality of the first software release, timing and frequency of software updates... All these are ingredients that might end up being more important even if they don't take so much buzz as the new pictures and demos.

Re: Open source vs confidential products
 

It is nobody's interest to waste developers' time and actually anybody's time. This is actually why I'm suggesting to the posters of the stylus' discussion to stop the speculation and panic trend.

I have answered a couple of times in that thread that we are supporting stylus friendly apps like liqbase, OSM2Go or NumptyPhysics as Fremantle Stars. Does anybody think that we want to waste their time as well? Maybe in 100 posts somebody will need to recall the same again. :)

Re: Open source vs confidential products
 

It's not that shocking, it's common human nature that lack of information breeds rumours.

And still the ONLY way to fight rumours is to take time to address them by giving enough credible information.

Of cource, any information that might be relevant to me as a potential buyer is important. And still what would happen if we get an announcement about the hardware, form factor, keyboard, stylus, screen, etc? Some "leaked blurry photos" and you can guarantee more positive rumours.

I don't think that in the future companies will countinue to work in that "cold war" style. ;)

Re: Open source vs confidential products
 

Uh, yes... I know that. In fact, that is exactly the reference I used. Go back and read his *whole post* not just that (out of context) snippet. Full source was available plus the post was an invitation to contribute.

If you mean that he made a project that basically compiled before releasing it, I don't see that that matters at all. That's common sense. Besides he remained the final point of release for all kernels anyway.

That's much different than Nokia's closed approach. GA said it best - "the occasional code tossed over the wall".

Anyway, I've said what I think and I'm just rehashing it now. This fight is unwinnable. I think any real progress is happening in the Mer and Ubuntu projects from now on though.

Re: Open source vs confidential products
 

Even that is overly optimistic; everything has to be re-written for the unique environment of Android.

Maemo's standardness is its main strength, the thing that sets the Maemo devices apart from all the rest. That's a standard Linux kernel under the hood, and everything is built on standard, open-source toolkits, so it is relatively painless to port stuff... Unlike the competition, which requires a complete development from scratch approach.

Nokia's hardware / product strategy is still pretty old-fashioned, though. I understand qgil's assertion that we're talking Real Money here, but his post sounded like an apology that nobody at Nokia is brave enough to risk Real Money on an unproven strategy, followed by a challenge along the lines of, "None of you big talkers would be brave enough either!"

In response, I'll say two things -- first, Nokia's still making a healthy profit, even in 'these times', and secondly, increased risk can lead to increased profit, sometimes.

But I think the risks inherent in the Maemo SW division are already high, and so we won't see any more risky behaviour, at least in the short term.

Re: Open source vs confidential products
 

The 770 is the finest machine I have ever seen. However, the attributes that make the 770 so cool have been/look clearly to continue to be systematically destroyed with the various successor/future successor units.

If you were still manufacturing 770s, then I could look forward to buying a new one when my unit eventually dies, but since that is not the case, it saddens me that after years and years and years of searching for the perfect mobile device, and then finding it, that it has been sh*tcanned so quickly.

Re: Open source vs confidential products
 

There is, I guess, a more than usual uncertainty about the future of portable computing.

Within a couple of years we will have ~10 or more hours of enough computing power for most people
very pocketable.
But not even (we) consumers know what kind of ergonomics we will want to carry with us.

Linux seems to be rapidly growing as a focus of interest.
And the ARM architecture has yet no real competition (?).
But what to put it into ?

Consider "old" times with paper and pencil, and the popular sizes of calendars :
Anything from a bit smaller than an n810 to a bit larger than a Filofax.

Really pocketable > smartphone size - (thin) NITs.
For photo albums > postcard size screen, or more ?
For keeping your diary > a really good (possibly thumb) keyboard.
Or a projector + a small screen ?
And how will business needs adapt ?

Will trousers have side pockets for light ~4*6*.5 inch devices ?
How about handbag fashion ?

Or (some time later) a sheet of fast colour-digital-paper containing computer and battery,
that curles up around your arm under the shirt/jacket sleeve, but snaps flat when you use it.


So, I guess Nokia would want maemo to take care of all kinds of input/output methods and sizes ASAP,
( software maturing slower than hardware, )

>> but I doubt that they (or the competition) have any good idea about the best strategy to achieve that without loosing a head start !


Nokia, anyway, made a brave start.
As to the difficulties they have had (and made) trying to merge with open source methods,
I think any large company would have had them.

It is difficult to make so different methods of developement cooperate efficiently.
I have seen that in other circumstances (http://www.internettablettalk.com/fo...5&postcount=91).

Re: Open source vs confidential products
 

This is clearly not true. Whatever else you may think about Nokia and its open source endeavors aside, Debian and Ubuntu's mobile efforts wouldn't be anywhere near where they are today without the work Nokia has been putting into mobile Linux, and Mer wouldn't even exist.

No, whether you'd like to admit it or not, Nokia is still a leader in the mobile Linux space and the larger, more established distributions are benefiting from it.

Re: Open source vs confidential products
 

My favourite example for proving this are Dan's elaborations on Novell's gnome-main-menu development (the technical details are outdated nowadays, but his points from 2006 are still valid: Bike shed, stop energy). See http://mail.gnome.org/archives/deskt.../msg00115.html .

Re: Open source vs confidential products
 

The problem here is that continued development in a lot of areas depends on having the information Nokia is withholding. Take the Application Manager. There's a fair bit of interest in improving it and even an idea or two about how to go about doing it, but because we don't really have much of a clue about what Nokia's planning for it in Fremantle, nobody's willing to spend any time working on it for fear of having that time turn into a total waste.

Thus, we're left treading water.

Re: Open source vs confidential products
 

I don't think that proves anything, except that Novell confused an increase in openness with a total loss of control.

Alan Cox's reply is particular relevant:

http://mail.gnome.org/archives/deskt.../msg00156.html

Yes, they'd've got a whole load of comments, but they didn't have to implement all the suggestions; or take every patch. That's where leadership comes in:

http://www.maemopeople.org/media/use...e-triangle.png
From maemo.org: what next?, by me exactly one year ago

However, by being more open they could get community involvement and buy-in. Many of the comments may have been dross, but some could have been good!

Re: Open source vs confidential products
 

IMHO, the application manager is one of the key end-user experience elements that Nokia wants to control themselves so that it fits into the overall device experience. This could be the reason why they keep it close/d.

Re: Open source vs confidential products
 

No, that's not it. In fact, it's basically been under open development since the beginning and everyone involved has responded quite positively to stuff like this.

Re: Open source vs confidential products
 

Perhaps these people would be better off to just DO something or slightly shift their focus into an adjacent area.
Waiting for a new device/OS version is a lame excuse for not starting something now (at least that's what I've begun to tell myself :-). There are more than enough things people could do to improve the Maemo environment and enduser exeprience that don't depend on knowing all about the latest API or hardware feature.
Patience is a virtue :-)

Re: Open source vs confidential products
 

Yeah, and while they read all the good and bad comments, the software gets written by itself. I'm completely with Novell on that one. Do it first, read comments later, add something if it is good. If there's no clear goal you'll end up with a big lump of unmaintainable patches glued together with spit.

Re: Open source vs confidential products
 

I've said it before and I'll keep saying it until the Nokia brass gets it:

The Nokia hardware product development process is broken.

It's bad enough with cell phones, where chronic (and easily avoidable) delays kept the sexy and high-potential N75 at the outer edges of AT&T's radar. Nokia consistently has failed to deliver promised product in working order on schedule.

I am not revealing anything proprietary here that will get what's left of my severance check pulled. This is public knowledge.

Now, if it's that bad for conventional product with a stable OS, imagine what it's like to deliver something on the bleeding edge.

I was on the front lines with the N800. I noticed hardware concerns months before launch (for which I had high responsibility) and when I expressed them, I was told by management that they were not an issue. But-- they became one several days before I had to ship 200 some-odd pristine, perfectly-functioning units to CES.

Those of us in product QA were constantly frustrated by silly development and release hang-ups. I'll have to avoid detailing that but surely people here can imagine.

Bottom line, if the maemo-to-hardware wedding hopes to result in a consumer honeymoon, PRODUCT RELEASE GAPS HAVE TO BE FILLED. Period. No gaps, but overlaps.

I know that many people in Nokia get this, and want it as much as my team did. But it's still broken. I'd love to come back and help fix it. :D

Re: Open source vs confidential products
 

Most developers aren't writing code 100% of the time. Most really good developers aren't writing code 66% of the time.

There's an obvious balance to strike, but the reductio ad absurdum retort to your argument is that a sole developer shouldn't even seek the opinion of a single peer, because that'd detract from the development.

I don't dispute that. However, I also know that a single developer will do something better after having solicited as much intelligent feedback possible as early as possible - as long as they know what the requirements are well enough to know to whom to listen.

It's sheer arrogance on the part of a company that they know better than the hundreds of people passionate enough about a product to be developing it for free.

Re: Open source vs confidential products
 

Looks like we're still trying to find that necessary balance...

Re: Open source vs confidential products
 

Ok I'll give you that. Let me rephrase it to "Any usable progress that is actually *available* is happening in the Mer and Ubuntu projects". I have no illusions that Nokia hasn't obsoleted the n800 and n810.

I personally don't think Fremantle will amount to anything, in critical mass. Unless it runs as distributed on older devices. Fortunately the trickledown will certainly help those other projects.

All my opinion of course. But I've been fooled once, and I'm not buying another NIT, as nice as my n800 is.

Re: Open source vs confidential products
 

Whilst at the same time, they open up its source code, accept patches and move the original developer on to other tasks? I don't think anything as specific as you suggest is going on here.

Hildon Application Manager is practically a shining example of Nokia open source; apart from:

• Fremantle plans aren't shared.
• Some new features were implemented without any discussion on its mailing list (although not many, because most open developers got involved after it had mostly been complete)
• Diablo patches were applied (and are shipping in Mer) to massively improve the user experience but Nokia doesn't have the processes in-place to ship them outside of a monolithic release.

With the clarified rule in Maemo 5 of "everything below UI layer open, differentiating applications closed" (the so-called 80/20); existing open applications like modest and h-a-m have, I think suffered - just when they were getting external open source developers involved. It's unfortunate as these end-user facing applications are the ones which make a good gateway in to lower-level hacking: people can see something which directly impacts them and make a small change to fix it.

Re: Open source vs confidential products
 

regarding openness in products from day one.

if nokia had said "we want to strip down and start completely from scratch in order to reboot the maemo product line"
whether or not it would be justified, practically every single person here would be up in arms and it would never get underway.
its not the sort of project they would undertake.

also, if nokia had just splashed out serious amounts of money on an unwritten project we would also think they were mad.

its a lose lose situation here for nokias commercial involvement in a hobbiest OS.
Sure, they have direction and can see certain elements hitting off, but on the whole it should be *US* the community who do the planning work and putting everything together.

Only once that work starts and gains momentum would I see it as beneficial for nokia to step up and support those projects - which surprisingly it has!

its no use us pontificating in here about projects that aren't yet under our microscope and instead we have been concentrating on whats to hand.

with that in mind, anyone is welcome to come and lend a hand in the liqbase playground (its more like a building site at the moment tho lol)

I have it up on an invite only git server and as long as one thing is adhered to I do not mind who in the community comes and has a look/gets involved:

there will not be a full release until its ready and stable.

I want the users to have the best experience possible and it should be completely functional and usable by my grandma.

it works nicely on 8x0, and I've spent the last few months taking the monolithic app and inverting it into a library to allow completely standalone apps to be created using the nice graphics and bits.
call into #liqbase on freenode

Re: Open source vs confidential products
 

That's the noblest goal I've seen here in a while - fantastic!

Maybe we can make that the rallying cry for Maemo overall...