2009-05-31

Google Wave - Game Changer?

Google Wave is their new product aimed squarely at the recent surge in social sites and also at communications. At its best, it could be a new paradigm of communication tools allowing real time communication, collaboration, and sharing. On the other hand, it might be an attack on social media sites and their concentration of power and reach.

While e-mail and even instant messaging are similar in basic operation to plain paper mail, Google Wave could start to branch out into capabilities and functionalities beyond that. The popularity of tools like IM, SMS, and Twitter show that people crave a quick, real-time platform to communicate. But aside from social communications, many practical uses would require documents for collaboration and group effort which are awkward with these traditional methods of communication.

Wave provides an integrated metaphor to handle these traditional methods while adding additional functionality and automation. In addition to having lightweight apps and widgets, Wave allows "robots" that can interact with the wave in predetermined ways. Robots can provide additional content by referencing existing content or pulling from other services for integration.

The other aspect of Wave is that it is going to be open sourced and distributed. This could be a way to build interoperating systems that would vie with the existing social media giants such as FaceBook and Twitter. Being based on Jabber's open XMPP protocol and with Google's promise to open source the project, Wave servers could spring up across the Internet working together and tearing down the walled gardens that tend to concentrate users in a few closed systems.

If Google Wave works as advertised, it could allow users to communicate and collaborate with their own self-determined identities without locking that content and communication on a proprietary system. A successful Wave could end the domination of closed systems that tend to hold the content and relationships hostage. This would be bad for FaceBook and Twitter, not to mention MySpace (who mentions them anymore anyway).







Official Google Blog: Went Walkabout. Brought back Google Wave.

2009-05-16

Apple hires former OLPC security head to harden Mac OS X

Apple hires former OLPC security head to harden Mac OS X - Ars Technica

Most articles I've seen are focussed mostly or entirely on what this could mean for Mac OS X on the Mac platform, but I'm intrigued by the parallels with the iPhone/iPod version.

Currently, each app on the iPhone platform runs only with access to its own limited filesystem area with no access to other apps' data. This makes it difficult for apps to work together, but it dramatically reduces the potential impact of some kind of malware. The malware would be restricted to its own area, which does not allow too much mischief.

This is similar though nowhere near as sophisticated as Ivan Krstic's Bitfrost for OLPC. Under this system, each app is virtualized into a sandbox that only allows interactions that are explicitly allowed by the OS. This prevents an app from accessing hardware, software, or data that has not been explicitly allowed thus preventing any malware misbehavior.

While Mac OS X has always been in a much better position than Windows XP (based on smaller market share and underlying design), this hire signals an effort on Apple's part to really take security seriously. This addition could mean great things with the foundation they already have in place. Just as the geek-set have started to point out some of the glaring chinks in OS X's armor, Apple is moving to fix the problems.

Combining Bitfost-type functionality with the existing application signing infrastructure and Mach kernel features could allow Apple to extend and secure their security lead. And as Apple grows their market share in both computers and handhelds, this will continue to become more important. I'm looking forward to see how they implement a slick UI and make this easy yet still safe for users.



Links:





Hadoop Sorts a Petabyte in 16.25 Hours and a Terabyte in 62 Seconds

Hadoop Sorts a Petabyte in 16.25 Hours and a Terabyte in 62 Seconds (Hadoop and Distributed Computing at Yahoo!)

That is some fast sorting. Damn impressive.