Ubiquity lets you browse the social web by typing, not pointing. In 2008, Mozilla Labs launched an ambitious project: a natural language interface for the web 2.0. It is now no longer in active development, but is still maintained to be compatible with the latest Firefox. So you can still enjoy this mature piece of research.

Every few years, natural language processing sees a little survival. In recent years we’ve seen Wolfram Alpha and Ubiquity. And with each iteration, the interfaces become better and better.

You can try out Ubiquity in Firefox right now.

Ubiquity does not aspire at understanding everything you type. It’s goals are much more humble and therefore much more attainable. It offers a sort of command line interface to the web. But the commands you type have been so refined it feels like you’re typing in natural English.

Its power lies in the huge amount of commands that are already available and its endless extensibility. If you know JavaScript or Python, you’ve got the tools to write your own commands.

In the past months, I’ve seen a lot of “social” browsers appear, like RockMeIt and Flock. All of them use a complicated graphical user interface to give you access to all your favorite social networks. Ubiquity offers the same flexibility, but in a very minimal interface. Some will like it for this, others will hate it.

Because of its command line style, it also looks like a great way to automate certain workflows. For instance, I’ve been looking at Ubiquity to optimize the way I do blog posts. Certain repetitive tasks, like image resizing and linking look like a great candidate for an Ubiquity command.

Although there’s no active Ubiquity development right now. I’m certain we’ll be seeing more of this. Maybe it will be a new version or maybe it’ll just be a completely new project. I’m certainly interested in finding out what’s next for natural language user interfaces.

(image credit)

Gravity Bone could be the indie game equivalent of Portal. If only it was a little longer. 2 years ago Brendon Chung’s Blendo Games released Gravity Bone. More a concept than an actual game. It’s still worth playing today.
Just before I finished this little game, I thought I was onto a great action puzzler to haul me over until Portal 2 comes around next year. Sadly, just after that thought the game ended.

If you haven’t played the game, I’m not going to spoil it. You step into the role of some kind of hitman in a stylish cube-faced world. That should be enough to make you go out and grab the game.

Judging by the docs, I got the impression that Brendon had material for a much longer game that would consist of using a number of tools in increasingly complex and imaginative situations. For some reason though (money?) it appears he never got around to it.

And that’s a real shame.

Maybe we should create an indie games fund. There are a few sites out there that allow musicians to let people buy a share of their music productions. Do we need something similar for indie games?

Dion and Ben would like to see a “WebStore” for websites. Basically what Yahoo! once was, a curated list of only the best websites. They think this is possible because with the advent of HTML5 and all of its related technologies, there really is no reason why a website should be any less interactive or any less refined than a native app.

They might have a point. Creating a HTML5 web application has the added advantage that, if done right, it’s also cross platform. If you don’t use proprietary extensions, the same web app will work on iPhone, Android and other smartphones. It’ll even work on a normal browser. Although I suppose the difference in screen resolution will be a little too huge to get a nice looking site on both extremes.

I did a little research to see what’s out there for mobile web app development. I have a few ideas to make something myself, so who knows, my first iPhone app might be coming soon. Until then, here are some good resources to get started :

• jQTouch is a jQuery plugin for mobile development. On my iPod Touch, the demo site was very enjoyable to navigate. I’m going to have to load it in the Android emulator soon to see how it looks there.
• Just like jQTouch, the Appcelerator Titanium platform was suggested by Dion and Ben. From the little I’ve read, I’ve gathered that the platform offers a cross-compiler that can compile one web application to many different platforms. It looks very professional and is open source so this might be the first platform I’m going to try myself.
• iWebKit is focused on the iPhone platform. The advantage is that it looks better than jQTouch, the disadvantage is that it probably won’t look as nice on Android phones. GigaOm has a nice tutorial.

If you know any more frameworks/libraries/platforms, feel free to share them in the comments. There’s always room for some more choices.

Not all web applications are created equally. Some attract millions of users, others only a handful. One major differentiator is how data and features are represented visually and how the user interacts with them. Some chase users away by their terse, technical and random screens. Others lure in even the most uninterested user by engaging them every step of the way.

But how do you get from that boring spreadsheet to an interactive user magnet?

The Design with Intent Toolkit helps by asking key questions on user interaction and design in general.

The toolkit is a pack of about 100 cards that ask questions on how you could improve your design in order to influence the user to do exactly what needs to be done. Each card also offers a possible practical answer to the question by showing an example.

The cards have influences from a wide range of design disciplines. Ranging from architecture to web site design. Some may not be applicable to your specific situation, but a surprisingly large number are.

It’s a great game: If you want to improve the design of your application, pick a card and see what answers you can think of.

For a while I’ve been adding a few features to my blogposts through WordPress plugins. Most importantly, I’m using and loving the Outbrain rating plugin and the Sociable social bookmarking and sharing plugin. One thing I didn’t like was the way those plugins integrated on the webpage. They just add a bunch of stuff to the end of the post that you have no control over.

Until now.

This is a fairly advanced topic, but not that complicated once you get the hang of it. Plugins add HTML code to your site by adding filters to the WordPress rendering. For instance, by adding a filter to the “the_content” execution, it is possible to influence the output of the blog post details.

Outbrain and Sociable work exactly like this. When the content of the post is rendered they add a filter to it that adds their own widgets. What I wanted to achieve was to have more influence over the placement of those plugins.

If you go to a post detail now, you’ll notice that both widgets are now inside their own div (to which I added a border). It’s something that is not possible if you respect the plugins standard placement.

So how do you go about it? Well there’s three steps:

1. First you need to figure out the name of the filter that the plugin uses. You’ll need to open up the plugin source code and find a line that looks like “add_filter(“the_content”, “xxxxxx”);”. For Outbrain it’s “outbrain_display”; for Sociable it’s “sociable_display_hook”. You might need to experiment a little with your plugin.
2. Next you remove the filter in your functions.php
3. And finally you add your own filter, exactly where you want to display the plugin. Or you could execute the function where you want it to. There are a wealth of possibilities here. This also happens in functions.php.

For instance, if I wanted to move the Sociable plugin from the bottom of the post content area to one of Thematic’s designated areas, called “thematic_post_footer”, I’d add the following code to my themes functions.php:

1
2
3
4
5
6
7
8
9

function remove_filter_from_the_content() {
  remove_filter('the_content', 'sociable_display_hook');
}
add_action('init', 'remove_filter_from_the_content');
 
function my_postfooter() {
  echo '<div id="social_footer">' . sociable_display_hook('') . '</div>';
}	
add_filter('thematic_postfooter','my_postfooter');

This will remove the Sociable executions from the content area, move it to Thematic’s post footer and wrap it in a div that can be used to style it.

Feel free to share your WordPress hacks in the comments.

Basically Stylish applies custom CSS style sheets to the websites you specify. For some reason I thought the option to specify personal styles had always existed in browser (and Firefox in particular), but I couldn’t find much information on it. In any case, Stylish makes adding new CSS sheets extremely easy. It also has a few style management functions that allow you to select if and when CSS files should be active.

Two styles I’m currently trying out and which I really like are:

• Gawker Blogs – Minus Garbage Plus Width: If you’re following my Google Reader shared items, you might have noticed I’m a pretty faithful Gizmodo reader. While I like many of their articles (especially the longer, more thought out ones), I hate their design. This custom style is a godsend if I want to go and check out the comments on the site.
• Google Reader simple and clean: Google Reader is already a fairly minimalist application, but this style goes even further. Only try this if you like white and pure content.

Feel free to share your favorites in the comments. I’d love to discover some other ones.

I’ve been looking into mail clients lately. At work I use Outlook, and for my other mailing needs I have Gmail. Neither of these options are bad, but I’m also not too excited about both. So I’ve been looking around and found a few options and I’m now using a few. I can’t give you any definite judgment just yet, but here’s  a list of what I found and liked.

Most e-mail clients have very similar functionality. I’m looking for a few things:

• It should be able to handle large amounts of mail
• Search and archiving should be as easy and quick as possible
• I’d prefer it to be free
• If possible, make it visually attractive

Based on those criteria, I’ve selected two candidates, Postbox is the one I’m currently trying out and I’ll switch to Thunderbird next week.

Postbox

Postbox has its roots on the Mac and it shows in the interface. It’s very clean and functional. I also liked the almost automatic configuration, you fill in your e-mail address and password and it figures out your mail settings. I suppose this won’t work for all providers, but it worked flawlessly for my accounts. If you’re used to Gmail’s threading, don’t fear, Postbox has it too.

Postbox has both a free and a payable version. I’m currently trying the full version, but I’m not sure if they are worth the price.

One thing I think I’d find useful is a global inbox view. I couldn’t find this kind of view, but it might be lurking somewhere. Otherwise I haven’t really found much wrong with Postbox, so I’ll see how the trial continues.

Thunderbird

Thunderbird is Mozilla’s e-mail client. I’ve tried it a long time ago and wasn’t too impressed, but I have the feeling it deserves a new look. It took me some searching, but it appears Thunderbird also supports threading, so it seems to have very similar features as Postbox.

Thunderbird does have add-ons, which might be nice, I’ll have to check them out.

update: since I wrote this, I’ve been getting into Postbox (and liking it) and it looks like it is build on top of Thunderbird. Which means that it also supports the exact same add-ons that Thunderbird supports. Although this isn’t prominently displayed on their page, there is a section for compatible add-ons.

If you have a different favorite offline e-mail client, please let me know and share your choice in the comments. I’d love to have some other options.

(image credit)

Two weeks ago I discovered that Wesabe will shut down at the end of July. Wesabe was the tool I used to track and budget my expenses. Without as much as an e-mail, the Wesabe owners decided to give us a one month notice. Some promises for more export tools were made, but have not yet materialized. Considering the value of the data that they kept, I must say I am thoroughly disappointed. Obviously, if you run out of money, there’s not much you can do, but still, a business plan might have been a wise investment.

Anyway, I’m not here to discuss the potential pitfalls of cloud services. I’ve been there and others have done it better. I’m here to show you the way out that I took. I exported everything to Excel. I used the PearBudget spreadsheet and used the Excel to set everything up. Sadly I haven’t yet figured out an easy way to import existing data, but at least I now have a backup of everything that was on Wesabe.

First things first, PearBudget is now an online tool, but it started as an extremely useful budgeting spreadsheet that can still be downloaded (and is even maintained from time to time). You can download it from the PearBudget site. All the documentation you need is on the first worksheet.

Exporting from Wesabe

To get started, I exported all my data from Wesabe to CSV: there’s a “download your data” link on your profile page. Importing in Excel 2007 was a little different than it used to be:

• Create a new Excel document
• Click on “data” and import “from text”
• In the popup, you can select the file you downloaded from Wesabe
• Check “delimited” and in the second step choose “comma” as delimiter
• Click finish and your data will show up on the worksheet

Extracting expenses categories

On Wesabe, I used to tag all my expenses with the category to which they belonged. This allowed me to create budget. I don’t think every one used Wesabe like that, but I presume it was the most popular way of budgeting.

So I wanted to extract the unique tags in order to set up my expense categories:

• Copy the tags column to a different worksheet
• Remove the “Tags” header
• Select the column and in “data”, I choose “Advanced filter”
• Select the “Unique records only” checkbox
• and you’ve got a list of unique tags

Those can now be used as a basis for the expense categories. I used this chance to tweak them a little.

Importing data, todo

I’m not yet sure if I will be importing the data, because it’s not always suited for the way PearBudget works. For instance, I couldn’t import my Visa bills into Wesabe, so I never had any details on those transactions. So it’s still an open issue how I’m going to deal with that.

One more thing I’d like to do is create a program to parse my monthly bank statements to an easy format. Seems like a great way to test-drive Python.

(image credit)

Previously, I mentioned my love for Camtasia. What attracted me was the extensive video editor that was part of the package. What turned me off was the fairly steep price. Probably just right for professionals, but a little too high for many amateurs. Now, I think I’ve found the solution in the open source Kdenlive.

Kdenlive offers at least the options that Camtasia Studio, and even more. I’m only missing a quick speech bubbles, but you can reach a similar result, although it’s a little more involving. The interface is intuitive and similar to many of the best commercial video editing software out there, so you’ll get used to it really quickly. And if you need that little more, there are about 20 video tutorials available, many more articles and a user forum.

If that very short introduction enticed you to go try out Kdenlive, don’t be disappointed by the fact that it’s a Linux program. If you have a Mac, there’s a MacPort available. If you have any system that runs VirtualBox, there are images. And best of all, there is also a live DVD available that you can burn and boot.

There really is no excuse not to try it out.

A big thank you to NixiePixel for introducing me to Kdenlive!

(image credit)

The Codex of Alchemical Engineering is a strange little game. Some will find this the greatest game to come along in a long time. Others will not understand its attraction and will probably find playing it a chore. Don’t worry if you’re in either category, tastes differ.

The Codex is all about programming robotic arms. The hardcore way. This is not some kind of user friendly simulation, it is how factories work. Well, sort of, but it gets close. Your goal is to produce molecules. You’ve got a few sources of simple atoms and have to put them in different devices to create the desired end result.

It all sounds like chemistry, but it really isn’t. It’s robotics and it will keep you busy for the rest of the night.

Click if you’re up for a real challenge.