No this is not a 'the screen sucks' or 'it's too thick'-thread.
It's an honest post about what I feel is still troublesome and, for me, maybe one of the main reasons why I can't see myself buying an N900.
You don't have to agree at all as 'concerns' are almost per definition how you subjectively perceive something so I'm sure different people view these 'concerns' differently depending on your own paradigm.

I have an N810. I think it's a really nice piece of hardware and it is often even valuable to me (quick email checks, instant messaging, some VoiP stuff, etc) - but mostly it's a really neat toy. ( Note: For me. You may be using and viewing yours differently). Applications seem to be in a constant state of beta or stagnation as developers move on to something else (new app, new OS version, new-whatever). I know that the idea is that having the source-code made available means that someone else can/will pick it up and continue the devopment - but in most cases this doesn't happen.

I really like the concept of open-source and I think it works well in the more common distros like Ubuntu.
Maemo is very small distro in comparison, a particular OS version like Diablo, Mer and Fremantle is an even smaller subset. There are simply not enough developers that want to both support abandoned applications ontop of developing their own, new shiny stuff.

Taking into consideration that Fremantle is already a known dead-end OS version, to be replaced by Harmattan on future devices it gives me even more concerns that Fremantle, again, will be another OS with applications that gets 80% there.

If I would be willing to carry both a smartphone with 'stable'/built-out features AND an N900 I would do so simply because I think the N900 will be great fun to play with and try out new stuff in with a rather excited community. And/or if I was a college student who didn't have to have the boring stability that people in many professions require.
Sad thing is I can't see mysel carrying two phones.

Maybe I'm the only one who feels this way but in the case I am not it does make me wonder how Nokia will address such concerns in the future from professionals who wouldn't mind an open, open-sourced smart-phone as their primary phone.

Well, you question is legitimate. What i don't understand is what's keeping apps written for maemo5 from working on earlier OS versions. Are they that different? I've heard voices about harmattan being backwards-compatibile as far as hardware is concerned. If it is, this should fix the problem for a pretty long while

In a very generic and broad overview of the situation, there were many changes to Fremantle from Diablo (the 3D UI one of them) which makes Fremantle apps not backwards compatible. However, it is to be expected that Harmattan apps will run in Fremantle, because Fremantle will support Qt, which is the standard Nokia is pushing for Harmattan and Symbian.

Well written, but I don't think your concerns are quite in the right place. The way I'm looking at the evolution of the platform, Maemo 5 (while not at the 'end of the road'), is looking increasingly polished and ready for everyday use.

The much bigger interest in the N900 vs any of the prior devices (c.f. Google trends, etc) is also a good indicator that the 'splash' this makes will be a lot bigger and longer lasting, possibly attracting some real clout to third party application development.

Of course, yes, Harmattan is still down the road, but really, this is the way technology is as we know it - new stuff obsoletes old at a rather breathtaking pace, even in more mature markets like desktop PC land. (Mostly on a hardware level now, the maturation of the software stacks there happened a few decades ago - yet people managed to use them in business situations, regardless of the (lack of) stability). [edit: and yes, the encouragement of Qt applications in Fremantle will make the transition to Harmattan a lot less painful, I think]

In what way, really, are you concerned about Maemo 5 as a development platform being an issue? That would be my real question to come out of this. If you're concerned about it as a development platform, then you might have some merit - but as an end user, I think you're likely to get quite a good experience.

Fremantle is quite a different beast from its predecessors as far as I know. Not entirely reinvented, but certainly not a walk in the park. Even just looking at things like the user interface changes (and the low level application changes required to use them) are going to be a bit fun for applications looking at supporting both.

I suspect the N900 will have a much bigger community as a device with much broader appeal. The previous tablets were nice and everything but quite niche.

The *majority* of geek developers are really only interested in working on 'cool' stuff or learning something new. Things such as complete QA, maintaincement, documentation are secondary.

In my opinion, the only way to achieve a good variation in polished apps is with a paid app store. Otherwise, you're just going to get many developers writing the same type of app. Developer egos are generally big so of course THEIR app will be best.

Nokia needs to attract the professional developers who have 6 figure day jobs to work on paid apps in their free-time. And since the development is in C/C++ and most college grads are taught in Java these days, the good C developers are generally the ones already with several years experience.

Open-source ,Fremantle i dont have a clue so i am out of here

Having done a lot of OSS development, and a lot of paid development, I'd strongly disagree with this.

The only difference between paid and unpaid development is that paid development (usually) has leadership which is hopefully in a fairly strong position already to be able to support developers, so, they have enough direction to be able to make things like proper QA happen.

Hobbyist developers, on the other hand, either work in groups of peer development without many formal processes, or solo - and don't have a lot of project management experience.

Both of these brushes can't be tarred across the whole spectrum, though, there are *plenty* of exceptions to every rule. It only takes high profile disasters like Windows ME - and to give a counter example, OpenSSL (and the like) to see it's possible to have it totally different from what I've outlined above.

Getting back to my point, though: what a project needs is direction. If it lacks it, all the money in the world isn't going to save it.

I must be running with the 'older' crowd, all the developers I know have bills to pay, mouths to feed. The last thing they want to do is go home and spend their free time on something that is given away for free UNLESS it enhances their professional development (learning new techniques, APIs)

Maybe its different for developers with only a few years experience or those with rich parents.