Okay, so now is another serious blog about UI design.  As an architecture graduate and also a ‘Head’ UI designer, I keep in touch with what’s going on around the UI world.  A few heroes of mine in this genre is Apple Design and Jensen Harris of the Office 12 team.

Apple designs are often worshipped for their simplicity, and their ability to convince the engineers that removing features is actually a good thing.  With intelligent solutions, Apple’s remote control for Front Row only has 6 buttons, and it pretty does all the things the Windows Media Center remote does.

Believe it or not, cutting down features is not easy.  On one hand, it is a lot of design decision and intelligent guesses (since we don’t have user data unlike rich companies such as Microsoft).  On the other hand, it’s like telling the developers to cut an arm off their babies.  Thirdly, the geekiest crowd will cry, no matter what.

This happened in the 2.0 to 3.0 era of Trillian, when we were faced with an increasingly complex Preferences window and context menus.  There were always a desire to simplify Trillian, but there was never the guts to do so.  Rethinking, reorganizing and removing is the process, and knowing what should be removed and which goes where was a difficult process.  There were also technical limitations as well that would dictate the position of certain items, which I have to design around and justify them.

It was pretty much like cleaning up your room.  However, instead of tucking away all the junk in boxes that you will never find again, you are building a whole new shelving system that houses everything and lets you access their quickly without creating a visual mess.

Most of our new features in Trillian Astra do not occur in the Preferences window, however, and instead they would be mostly present in the most often used windows such as the contact list and chat windows.  It presents a whole new problem as this is a software that presents changes in real-time, and a hierachical tabbed model in Preferences window does not always work because certain real-time information needs to be visible at all times.

Since Trillian Pro 1.0, the UI has adopted the default Windows-like Menu+Toolbar model, a common way to lay out a software, in attempt to lure commercial users since the previous UI in .7x was considered too wacky.  Having a standardized UI similar to the rest of the OS makes it ‘easy’ for new users to learn Trillian because menus and toolbars behave in the way they expect them to be.  However, the problem is that most of the menu items were forced – Trillian does not really have that many menu items and tools to choose from.  Back then, since we had limited functionalities in SkinXML, the menu bar becomes a place for items that would be given too much focus if shown as a toolbar button but too little focus and out of context if shown as a context menu.  A lot of items are duplicates of toolbar buttons as well, e.g. the edit bar button.  Therefore, the menu bar, especially the ones in chat window, is more of a filler to create a sense of familiarity for users (or for us so we are confident to sell it?).

Trillian Astra faces the problem on both sides: On one hand, we got a lot of fillers for the menu bar; but on the other hand, we have a lot of new features to fill up the windows.  Can’t we just put all the new features in menus?  No, because doing that just buries the new features and requires users to click a few times in order to get it.  So instead of trying to fit these new items into the current menu and toolbar model, why not just do away with the menus and toolbars?

That was the ‘radical’ decision made more than a year ago in August, when Microsoft has their new Windows Vista new UI waving around with no menu bars.  I was skeptical of their decision, but felt good because I could finally find a plausible excuse to break away from the Trillian Pro UI and “return to form”.  I don’t understand necessarily when Vista removed all their menu bars because they do have a lot of features need to be accessed, but I certainly agree that Trillian does not need it after all, since we don’t have much.  Office 12 demos came out a few months later, and confirmed my theory.

Even with a new vision, new technology was needed in Trillian to execute the new UI.  Back in 2.0, we could hardly make anything exciting because we were only limited to buttons, and the buttons could not change any other aspects of the window.  3.0 got worse as we had video and audio chat, tons of new buttons flooded in with panels and stuffs, making it impossible to make any freeform skins.

Fortunately, SkinXML in Trillian Astra, like (the better part of) 3.0, will continue its curve to introduce more variables, thus the ability for a much more flexible UI.  It allows us to create a richer user experience by categorizing new features in tabs (similar to Office 12′s “Ribbon” Menu Tabs but they are actual tabs as well).  I surfed the wave of these new UI trends and took advantage of it, creating the new re-organized default skin to be known as Trillian Cordonata.

With that said, it doesn’t mean we were trying to follow trends, but the trends allowed us to rethink what we are and should be.  There were other reasons for a brand new skin, which I will talk about later.