A quick weekend project, building a site for my friend Cara using Indexhibit. It was easy enough to get started although while writing the plugin, I did have to add some extra stuff to the core. It did make me feel a little sad having to write PHP, but maybe this will inspire me to port it as a Rails/Sinatra app?

PS Cara is awesome if any one is looking for a production designer.

which is clearly a rip off of the classic eye bee m poster by Paul Rand:

The rule to ripping something off is to make it better. In this case, a massive failure. If I ever get old and need grandpa reading glasses, I will not go to Lens Crafters!

Nerding out at WWDC. If you’re here, I’ll see you around. Lets meet at the Thurs Beer Bash!

I finally got around to putting up a video of the Feedabot Viewer.

Every day, the collection of pages rendered by feedabot grows larger. In fact, there are approximately 250,000 pages in the feedabot archive as of a few days ago. Using the current web-based interface to navigate through this collection (one mouse click loads six new images), it would take approximately 35 hours of nonstop clicking to view all of those pages. A choice of layout similar to the Tumblr Mosaic Viewer would certainly help matters, but only slightly. It was only a matter of time before coming to the realization that there must be a better way to navigate (and further explore) such a large collection of web pages.

At this point, we must leave the browser and head back to the desktop. Even with rich media technologies such as Flash, the browser really isn’t the best place to experiment with interfaces for quickly navigating large data sets. The desktop viewer for the feedabot archive, tentatively called the ‘feedabot viewer’, was designed to be clean, fun, and aesthetically pleasing.

Interaction with the viewer is purely mouse-based, containing no popup menus or dialog boxes. A day’s worth of feedabot-rendered pages is loaded with a flick of the horizontal scroll wheel. Navigation through the resulting set occurs through the use of the vertical scroll wheel, and the mouse cursor position is used to reveal the tweet that contained the link of the rendered page. At this point, the user may ’select’ the page with a mouse click to extract the image resources from the page itself. The resulting images are placed on a layer in front of the view, with navigation possible through the use of the horizontal scroll wheel. When the mouse cursor position moves beyond the currently selected page, these images disappear. One can only hope that navigating the web will be this pleasant in the future.

A few more images of the viewer in action can be found on my blog, and in this Flickr set.

Quick Look-like interaction in the browser! Read more here.

Hilarious, but it’s so true. Ask anyone who has a public drawing site what the most commonly drawn “thing” is. Better yet, take a look at what people draw on bathroom stalls next time you’re at a bar.

The Architecture

An overview of the architecture is shown below:

A POST request from the browser is sent to a Rails Metal URL periodically (as opposed to responding to keystroke events). The Rails Metal method verifies sessions, and request parameters and forwards them over to the MudTyper Server. The MudTyper Server then sends a request to MudTyper Renderer to create either a file that is saved to disk or a base64 encoded string depending on the user’s browser. The browser receives the image as a response and includes the rendered image into the page.

Scalability

With this architecture, it is possible to scale for increased load. We can run multiple instances of MudTyper Server+Renderer, and use mod_proxy to get Apache to handle the load balancing.

Font Rendering

The new Renderer adds a lot of new support, including support for OpenType fonts with full kerning support, as well as many OTF features. For instance, we can now render Galaxie Cassiopeia’s contextual alternates, adding smooth transitions between letters.

Over the course of the past few years, I’ve found myself spending a lot less time on social networking sites like Facebook and Twitter. In contrast to popular opinion, I feel like they tend to enhance existing social relationships as opposed to being a replacement for them.

Even though I spend less time with them, I do acknowledge that they are good at keeping me up to date with news and projects that I wouldn’t have otherwise been made aware of. For me, it’s the the links people post that tend to be much more interesting than the comments. This observation, coupled with an interest in saving myself the time of visiting every link that passes through my Twitter and RSS feeds that might be interesting, has resulted in Feedabot.

Feedabot is simple. It monitors feeds (RSS, Twitter, etc.) and looks for links. When it finds a link, it renders and image of it, and places that image on the site. This provides me with additional information about whether or not it’s worth my time to visit any given link that passes through my feeds. Pages are marked when read, and can be placed into groups. Groups can be made public by adding some additional metadata (image, description, and title). I’ve created a few sets in my account here as an example.

While the current list of feeds that Feedabot monitors is fixed, Feedabot users are allowed to submit their own URLs that will be rendered and added to their profile. These can then be added to groups or sets.

Feedabot is a work in progress. It’s growing every day, and could benefit from some additional users. If you’d like to give it a try, send us an email and we’ll create you an account. If you’ve got feeds (Twitter lists, RSS, etc.) that you think would make a great addition to the current set, feel free to suggest one.

We think a lot about the Web. Mostly, we try to imagine new ways to interact and visualize the data we find on it. E15, for example, is a way to allow end users to take a step away from the browser and take it apart, one div at a time.

I’m often surprised at how thinking about the web in this way feels so new and unexplored. A lot of people I know seem to be so comfortable with the browser that they can’t imagine anything that could enhance their web experience in any way. We here at BuzaMoto are still taking steps at trying to change that.

In my most recent attempt at creating a new vantage point from which to view web data, I’ve recruited the Sunflow global illumination renderer, thousands of lines of C and Python, and hundreds of hours of CPU time to create a few visualizations of web data “outside” of the browser. The visualization above, for example, is a collection of data from various web sources (Flickr, Delicious, Google, Facebook, and Twitter), affiliated with myself (or user name), all piled up. The end result is a summary of the data I interact with on the web on a daily basis, put into perspective. What sometimes feels so large and formidible is actually rather small when viewed from the right angle. Maybe if everyone was able to see their own data in this way, they would realize how thin and shallow many of the Web 2.0 social networking sites are.

I’m currently in the process of trying to make this happen. I’d really like to streamline this process and make these visualizations accessible to everyone that wishes to see them.

In the meantime, check out the “Piles of Web Data” Flickr set that I’ve created to host my most recent creations.