nodebox // metaforix

Ubuntu + Beryl = Computing Hell

My Dell has just survived a trip from Hell – installing Beryl. How doth mine computer screweth? Let me count thy ways.

Installing Beryl the first time was a snap, just added the repositories, installed the beta nVidia drivers (yes, yes, I know I was looking for trouble) then installed Beryl and Emerald. But Beryl refused to start. I figured that it was the X.org configuration. Easy, just a minor change to the X.org configuration file, right?

Wrong! I decided to start from scratch. So uninstalled everything and reinstalled it again. This time it refused to install. So I added and removed a lot of repositories (thankfully backing up my apt-get sources file). Then it worked.

Post-installation, the windows looked cool, but the window borders were nowhere to be seen.

So I decided to remove Beryl. Easy…just uninstall the programs and reinstate the old drivers.

Too bad, the programs uninstalled fine, but the drivers refused to revert back. apt-get recognized the beta driver to be a newer version and refused to let me revert back to the official nvidia-glx.

As such, I had to change back the open source ‘nv’ driver, then install the nvidia-glx driver. But it needed the nvidia-kernel package, which wasn’t present in the repositories. So I had to manually download it.

All in all, reinstatement of my computer took 3 hours. Not the best way to spend my time.

Pageless

Another prime example of AJAX is shown here. Essentially this page advocates the removal of the “page” metaphor, similar to a book. Every website uses this metaphor – when the displayed content comes to an end, the reader is directed to a “Next” link to view more content. The authors of this webpage feel that this approach is the wrong one to take.

Hence they advocate a more revolutionary approach – depending on how far the user has read, the page will display different amounts of content. For example, if a user runs a search that returns 100 results, say, then the first 15 results will be displayed. As the reader advances to the bottom of the pages, the next few results (16-30) are displayed, so that the user will not have to click anything.

While this approach is certainly fresh and interesting, I can think of many ways that it might not be feasible. For one, it takes some getting use to. The finite nature of web pages is a strong draw to some (can you imagine reading 1000+ pages of web results? Breaking it up into chunks makes it more manageable.), so this approach might not work for some. Secondly, it may (I’m not really sure about this) break web compatibility. Does it degrade gracefully on non-CSS compliant browsers? Will it display properly on text-only browsers such as elinks?

Still, I can’t deny that it is an interesting piece of work. If nothing else, it is a useful exercise in AJAX coding.