First, let me say, thank you for the wonderful post on what's going on with MeeGo. I'm happy to hear an update about where things are at right now (and a little about how it got there). I'd really wish there was a simple Wiki page somewhere for those of us that are tacitly following it, so we could see where it is without having to read a whole forum, and join mail lists, and/or have MeeGo community members bring up unknown changes in status in arguments here.

One thing that's noticeably lacking is any information about anything since the "summer release". Is there another release planned? It sounds like things are moving to 1.3? From the look of the wiki, you'd guess the project is dead, since it's not seen an update since before the last release. (In fact, finding a link to the summer release page isn't that easy...)

I agree with much of what you said, but have to say, not so much on the topic of the UI. The UI may just be candy to some, but it's quite important on a mobile device. It doesn't matter what a device core is capable of if you can't operate it because the UI is poor, missing, or lacking. Anyone who's used a first-gen device vs something more matured will tell you, UI is vital. The other components are important as well, but the UI is not just "fluff". If the UI isn't there, that core isn't going to get used.

That's not entirely true. Frankly, there's been little kernel level development in the past year or so that would impact most users of the N900, or developers for that matter. How does a new GPS stack help someone writing a QT based web widget for the N900? It doesn't. The existing GPS stack has a few bugs, but for the most part it works pretty well. Some new functionality (h.e.n., packet injection, battery module status, etc) is nice, but we wouldn't be "utterly screwed" without it.

That's not to say that kernel activity is unimportant. It's great that people are doing that, since it does open things up and brings new features and options. But it's not something that's going to make or break the N900. It is vitally important to MeeGo, since adapting an open driver to a new system will be easier. But for the N900 in general, it's just not that vital.

No... They deserve admiration and thanks. Respect is not something you can purchase or trade for, it's something earned through treating others in a sound and respectable way.

Programmers in particular confuse thanks and respect often. Generally you don't get respect for doing a particular bit of work. You get respect sometimes because to pull off X or Y you have to be socially pleasant enough to cooperate with others, coordinate an effort, and inspire others to do things. You can also get it by explaining how you did technical things so others understand it, without being rude or condescending. Through that some people do gain respect, which others (or even they) may attribute to doing X or Y, but it's more about the journey of getting there than about the act of work itself.

And that's where we disagree. This would be like saying "there's no need to target Windows XP any more, Vista is the future!" Sure, some companies do that. But reality is there are still lots of users running XP, and will be for a long while yet. Especially when "Vista" isn't ready for the average consumer.

As for having to redistribute full images, that's hogwash. There are plenty of closed parts replaced by CSSU at the package level, without images. I do get that there will be limits. But again, the current system can do a lot already, and there's a solid API for tacking in lots of new functionality.

I've seen completely open community efforts end just as poorly. The OpenMoko project springs to mind, along with it's derivatives and accomplices, none of which has seen standard adoption in anything yet. They put out to bits of hardware, and to this day none of the project images can do three simple tasks that my N900 does almost daily:

1. Connect to a tower, and stay connected.
2. Ring when a call comes in, and show a UI in time to answer it.
3. Route audio to and from a bluetooth headset for said call.

Most distros can do 1 & 2, but not 3. The few that can do 3, can't do 1 or 2 reliably. Even the android ports (there are 3 of them) fail to do all 3.

And least we think MeeGo is magical for having corporate sponsors, I ask: Where's Ofono today? You remember Ofono, right? The project Nokia and Intel started back in 08, an open source phone stack that got adopted into just about nothing? Well, nothing may be harsh... the FreeRunner and the N900 both have a distro or two that attempt to use it.

Just because it's got more open components doesn't mean people will flock to it. It has to actually run on something, in a stable way, in order to pick up steam. Having a busted (or no) UI means it will die just as quickly as the projects above. That's one advantage Maemo has: There's already a user base out there. Lots of people are actively using Maemo on their Nokia N**0 devices. That alone will keep things ticking here for a good long time, unless MeeGo actually gets to the point of being usable as a stable, every day distro. Until that happens, why discourage development here? On an active, stable platform with a good number of users...