Sorry but the insight about UI in [this] community, and software design in general (not only FOSS), is seriously lacking. Thankfully, we have HIG for that. There is one person who IMO stands out. Who made refreshing, insightful contributions to UI albeit with feedback from others. That person is wazd.

I am rather glad some vocal people here don't have the UI their way because else it'd be a mess. Or they simply refuse to understand implications of their request, do not understand why their compare is invalid, such as with universal rotation/portrait mode discussion. The reason they don't get their way is because Nokia's UI team has a spine and stands behind their product, and because they are the experts within their field. Not the user! The user did not design or know the HIG as if its his bible!

Note this is something different than saying user's opinion is worth nada!! The customer input is feedback as-is where a Nokia UI designer may or may not put value in. The customer is not always right! The customer is perhaps made to believe he is always right, but those people who make him believe that are hired to do so and not honest in their motive, or able to express in a more tactful manner. It is also important to understand the background of the user! In every usability study I've read this was taken into account! PalmOS user, Symbian user, iPhone user, and even the desktop is what is relevant in such contexts.

Important reason for layers between engineers and end-users for unless mutual respect is given they don't mix too well. So, you let layers inbetween weed out the conclusions or useful tidbits. Those layers are qgil, peter, community council, and perhaps on behalf of UI team ragnar. On something like FOSDEM its different, there you can get direct interaction, because its developers among developers instead of 'pesky' end users who are indirectly paying; the paycheck comes from ie. Nokia! Its much akin to Obama not writing his own speech, disclaimer on corporate blogs, or spokesperson of VIP dealing with blog.

Tabs, including in browsers, were added to solve task bar cluttering allowing further cluttering the context of browser stayed in browser, and various browser windows can define various contexts albeit without official tagging support; ie. it is virtual; user is context-aware, OS or browser is not.

Now, let us assume MicroB in Fremantle has tab support. OK...

First problem: how is user going to switch back and forth inbetween these tabs? Look at how Fennec does that. It isn't easy to implement this well. The previews are rather small, and Fennec sacrifices its swipe right to left for this functionality. (Personally, I'm not sure what I prefer... yet.)

Next, this gives Dashboard potentially more windows while also rendering Dashboard somewhat useless because it has no support for tabs. So some windows would be previewed and even live rendered whereas others would not. While this could be added for example by allowing the user to select the window and browse through the tabs (in an elegant way perhaps a pseudo Dashboard... which MobileSafari has (without live rendering)), it may lead to user not able to find their tabs at all anymore, and it is exception which breaks HIG. _If_ you are going to implement such in browser there will be other tasks where it is appropriate and people will say '..but he does it too' hence welcome new addition to HIG.

Otherwise, the difference between tabs and windows is rather minimal, especially if you lack a taskbar containing your windows. It may be something to consider perhaps later. Hence, it becomes 'leave as is' and 'stick to HIG'.

And now the reason why the browser should not have tabs: the user should not have 4 windows with 4 tabs open. It is a waste of memory and space contrary to the way the user is meant to interact with browser for it has bookmarks, awesome bar, history. The point is very much getting things done and be done with them instead of keeping too much open.

They are arriving at new platform with new UI paradigms. Another reason is that device is 'also a phone'. Symbian was from ground up an OS for phones, and its phone related features ofcourse became polished after time. In Symbian on non-touchscreen device, it makes a lot of sense to quickly browse through tabs. Size of checkbox is not important. In a stylus/touch-based device perhaps as well... but in a finger/touch-based device this is more difficult.

Different filters do make sense, but why the rush? See my comment below about respect!

By insisting wanting their previous UI paradigm without trying out new they are not giving N900/Maemo 5 a fair chance, and are wasting not only their own time but also hours. What you want for feedback is something of substance: for example, A) a very well documented user experience where a user sat down and wrote down her exact experience ([1], [2], [3]) or it was recorded ([4]), or B) where person used software & hardware for a while and accepted and embraced the way they work and work from that starting point addressing its strengths and weaknesses. C) I'm not aware of any C! And, doing A is either expensive or requires a lot of time and a good setup. I loved the format, thoroughness, and style of [1] but the links are dead. If anyone can revive the links I'd very much appreciate!!

It makes no sense complaining your Lamborghini isn't easy to handle in middle of Amsterdam because its just so much easier to use a Mini in centre. But if you're used to a Lamborghini, driving a Mini is very different, and immediately complaining about its shortcomings without trying it out for a while understanding its strengths and work from what it is rather than what others are is what is required. This is respecting the hard work of those who contributed to the end product!.

Maybe someone else should listen and provide dept. (in this case UI team) the conclusions of feedback.

Users are very unrealistic and believe in abacadabra. They, sometimes literally, tend to believe developers are wizards. The developers are the parents, the users the children. Children are bright at time, and one should not ignore their expression, but given the parent (and teachers etc etc) are the wise with the experience and knowledge children do need guidance of their experts, and accept them as such.

[1] GNOME 2.6, KDE 3.3 (links dead...)
[2] KDE 3.2
[3] SUN GNOME 2001
[4] Example prices of usability studies...