Are you talking about the relatively new popup that shows up after a sustained mouse click or about something further back?
I was thinking of the original dialog box, it became smoother as time went on and version numbers up IIRC. I have no clue what is being done now, haven't had a clean install in years. The point was that unless you were reinstalling a system or something similar, you would not intentionally make a mouse gesture the first time around you triggered that dialog. Basically you got a tool tip at a fairly random point in your browsing history telling users about this great new feature.
In a manner of speaking. Now, I don't consider that the epitome of usability.
User makes a spastic move, e.g. grabbing the mouse whilst falling off the chair
Opera chirping: I saw that you made a mouse gesture. Repeat your action next time you want to clone a tab!
But the mouse gestures were discoverable. The reason was that Trond, the designer of mouse gestures, also was in the group shaping the UI. Other features were actively non-discoverable. A modal dialog box for each feature "discovered" while in the process of doing something else, like trying to avoid breaking a bone, would be a horrible UI. There are better ways of doing this.
Regarding your point 1, I always wonder how a user can become more advanced if everything's hidden. Windows Vista and up hide the menu bar, so the file associations feature might be hard to find.
I argued for discoverability, the system starts out simple and as you use it you discover new possibilities that you can take advantage of or ignore. Over time the system will adapt to your needs. There is a great number of hurdles to overcome to manage this well, but Opera was in a position where they could actually pull this off, since many of these hurdles had already been passed more or less successfully to handle other problems.
I used to describe Opera as a configuration engine that could read web pages.
The forum system is filled with gremlins, there is a reason why they have tried to rewrite it for half a decade. I was sceptical to the "Eek! It is written in Perl (+SQL+XSLT). Let's rewrite it from scratch in Python!" approach, but it is too bad it didn't work out, that's after all the reason we are here. My Opera has occasional synchronisation issues, less nowadays, but that is the new normal. Try Facebook for a system that really suffers from it.
I tried to make a joke about it on the forums, but I actually think it's a serious case - again. Some of these alleged autobans just don't smell right. If autoban is really so trigger happy, then why, for example, am I still a member?
You're probably grandfathered in.
Don't blame on malice what you can blame on badly configured software. I can read those threads, they are not particularly embarrassing.
I suspect Facebook is like many other buyout rumours the last decade, manufactured for stock manipulation purposes. My impression of the Oslo Stock Exchange is that it is more Wild West than stock exchanges at large, and investors there have definitely not been beyond a little rumour spreading from time to time. During my time they had a stellar performance record of never matching reality.
I thought the rumour of a Microsoft buyout was particularly rich. (At least Facebook would kind of make sense.)
String's sleuthing skills struck a fumble. "Group started: Sep 2005"
I'm not sure of that. Most posters that comes from D&D have a certain experience and things don't usually goes too far in terms of friction but new ones can be a different matter. People needs a cold warn when too hot...
New users aren't that much of a challenge. You are nice to new users, show them the ropes, and that there is enough for them to hang themselves with. Then you ban them. Sometimes they then reincarnate into a troll, then you show them the sun, exerminate their messages and ban them. There was this particularly tenacious guy who did this about a dozen times, flooding the forums with porn or worse, but each time his creations were gone within minutes, and he gave up after a month or so. He, Grumpy Old Man, and ultimately Bantay were the only ban on sights, the rest got a chance to break the rules at least once before a ban.
You have those that generally behave sensibly unless provoked for some (often trivial) reason, then they break as many rules in as short a time span as humanly possible. Ultimately they all ended up banned, we didn't do anger management classes.
You have those that generally behave sensibly unless in the same thread as someone else who also generally behave sensibly, but together they celebrate the critical mass. Most of these eventually ended up banned.
Trolls and internet warriors can be a serious pain. We spent incredible effort in trying to reform the irreformable and Grumpy, until and actually after the ban from the land policy.
All that notwithstanding hell hath no fury like a spam storm. Spam is the real concern.
The change wasn't just for usability or aesthetics, but also for security. There were too many ways to spoof web site addresses. This approach, and the padlock and the rest, made it easier to see the important information, and harder to spoof it.
If you look at a longer URL it has some useful information, and a lot of gunk, some which is of interest to the site owner and not the user (e.g. which Facebook page that referred to it, or for searches that Opera should get percentages of search income). There is nothing implicit in a URL (outside the host address) that say that this piece of information is useful and that is not. Nothing that can't be subverted anyway.
I think there is some cleverness unused in this field, but the problem with cleverness is that one day it is going to bite you. One such cleverness would be to take advantage of history, the parts of the URL that change when you go from one page to the next are more likely to be useful than the parts of the URL that don't.
Most of it anyway, it hit the 20,000 limit. I suspect that is based on UTF-8 and 20,000 bytes, as I don't think there were actually 20,000 characters in the quoted text.