Posts Tagged ‘mozconcept’

Firefox New Tab: Visual Update

Monday, April 13th, 2009

All has been quiet on the new tab front for the last couple of weeks. We’ve been up to two things in the process of getting the new tab ready for potential uplift into Firefox. The first is we’ve been working on an overview of how the add-on was designed including performance and security. If you’ve ever wanted to get a guided tour of how something like the new tab is implemented, check it out. The second thing we’ve been working on is finding a visual style that blends in with Firefox.

Inspired by the horizontal styling of the thumbnails in Chris Stone’s answer to the call for participation, we’ve got some new designs that incorporate the learnings from the last 36 revisions while finally making it feel much more Firefox-y. Unfortunately, we haven’t implemented the new style yet so for the time being, it’s see only.

(more…)

TaskFox Mouse: Iteration 0

Monday, April 13th, 2009

Two of TaskFox’s goals are to: Work with existing workflows, not against them, and to help people complete tasks rather than be forced to laboriously jump through hoops.

We’ve been looking at what TaskFox looks like in the URL bar. It’s an interface that excels when you want to start a new navigation task or do something somewhat unrelated to what you are currently looking at. If you see an street address on a page, an artist you are interested in, or text in a undecipherable language, you should be able to start using TaskFox right there.

(more…)

Firefox.next: Tabs on the side?

Friday, April 10th, 2009

Inspired by numerous tabs-on-the-side extensions (in particular Tree Tabs by Piro-san), hall-way conversations, and Oliver Reichenstein’s recent blog post, we’ve been thinking more about the possibilities and ramifications of putting “tabs” on the side of the browser.

(more…)

Taskfox Prototype: Ubiquity in Firefox

Wednesday, April 8th, 2009

As a user experience exploration, Ubiquity has been incredibly successful. Over a million downloads have highlighted the need for the web to be connected more tightly with by the power of task-based interfaces. Due to the passion of users, the user tutorial has been translated into ten languages. Similarly, the thousands of commands written for Ubiquity illustrate a latent desire to be able to write tiny amounts of code that enhance the web in fundamental ways.

We are currently working on bringing some of that power to Firefox. For a more detailed look at some of the directions we’ve been thinking about, check out the mockups page of the project wiki.

(more…)

Firefox New Tab: In-Line Search

Friday, March 27th, 2009

Over the last couple of weeks, we’ve been iterating on the idea & design of what a Firefox new tab could look like. All told, we’ve now gone through 36 different versions of the page, with thousands of particpants helping test and provide daily feedback & new ideas.

(more…)

New Tab for Firefox: Iterations

Thursday, March 5th, 2009

We put up a post on the iterating on the new tab page for Firefox on the Mozilla Labs site today. It builds on feedback from the last two rounds of new tab concepts, we know that the page needs to load instantly (even a small wait breaks user experience); that it shouldn’t be visually distracting; and that it should be a launch point into your daily activities. One level higher, the distilled design themes are: No configuration, Streamlined, and Polite.

(more…)

Ubiquity In the Firefox: Round 2

Thursday, February 19th, 2009

We’ve been iterating hard on ideas to bring the power of Ubiquity to Firefox main. The two places it makes sense to surface Ubiquity-like power are (a) in situ with content when we are trying to manipulate, and (b) in the location bar, where we already type to perform navigation tasks. This post focuses on the second use case.

The three design goals, in shorten form, from round 1 were:

(1) Don’t force new work flows.
(2) It must be localizable.
(3) It should feel like Firefox.

We’ve added a new design goal, as a subset of not forcing new work flows: discoverability. The interfaces we design should be self-learnable. In this case, that doesn’t mean ever piece of functionality is immediately obvious, but that over time the system can teach you — step by step — how to use more and more of itself.

Note that all of these mockups are sketches. They don’t imply anything about the final visual style. From an interaction standpoint, they focus on tight feedback loops, as well as putting contextual autocomplete as close to the text being entered as possible.

Mockup 1

The Ubiquity-esque actions appear in the Awesome Bar results, and are subject to the same ranking algorithms as everything else.

The inset image on the right is an alternative way of accessing verbs: instead of having them appear in the awesome bar results, they appear as autocorrect-style text above what you’ve typed. The benefit is that you can always hit tab to quickly get to the action you want (as opposed to using the arrow keys for navigating the awesome bar results). It can also be unified with methods of structured modifiers (see later mockups). The detriment is that it is yet another mechanism and is visually noisy.

Other thoughts: The background of the url bar can change colors to add a visual key that an action is taking place. We can also unify the keyword mechanism, so that if you type “g ” it automatically gets expanded to a “Google” action.

Mockup 2

(more…)

Scaling Ubiquity to 60+ Languages: We Need Your Help

Tuesday, February 17th, 2009

The year is 1953. Robert Floyd, the man most cited in Donald Knuth’s The Art of Computer Programing, takes the stage. His topic is on using English as a programming language:

“Each word must be implemented by a procedure which somehow contains its meaning and consistently interlocks with other procedures for other words. I can’t think of any task, intellectual or not, which has ever been carried out which approaches this magnitude.”

Over half a century later, those words ring prophetically true. Instructing a computer to do what you want with natural language is astronomically hard. That’s why Ubiquity cheats left and right to do it. It embraces ambiguity, uses tight feedback loops, and a restricted vocabulary plus grammar to give the appearance of something human. And it can do a lot better.

As we think about Ubiquity uplift into Firefox, we aren’t just bound to English. We need to have an interface that can potentially work in the 60+ languages that Firefox has been localized too. For that we need your help.

I’m pleased to announce that Mitcho Erlewine, a linguist-coder, will be leading the charge in helping us understand how to bring conversational computing to the Firefox scale. His first blog post is on How natural should Ubiquity really be? Although he speaks four languages fluently (English, Japanese, French, Chinese) and is a gifted linguist, he can’t do it alone. Especially if your native tongue is not English we need you to get involved in blogging, thinking, and mocking up ideas for how Ubiquity in Firefox’s Awesome Bar can work in your language. Put your mockups on Flickr and tag them with ubiquity and mozconcept.

Question: What are the greatest difficulties in bringing Ubiquity to your language?