If you're particularly attuned to tech gossip, you'll know that Google's Chrome OS announcement has prompted a lot of whispers about something called "Gazelle" being cooked up in Microsoft's labs. Part browser, part OS, the word on the street is that Gazelle will be announced soon, and ultimately compete in some way with either Chrome (the browser) or Chrome (the OS). As usual, most of this is just based on hopes and fairytales, but the scholarly folks at Ars Technica have done some digging and come up with a white paper from Microsoft Research that details some of what Gazelle is all about -- and surprise surprise, although it shares some similarities with Chrome, it's actually quite different.

At the most general level, Gazelle is an experiment in building an ultrasecure browser. Like Chrome, it breaks tasks up into different processes, but instead of separating at the page level, Gazelle breaks individual page elements into different processes, allowing content from different servers to be isolated and ultimately providing fine-grained security controls. To manage all these different processes, there's a central "kernel," which is where all the OS talk stems from -- it's all still running on Windows, and the rendering engine is still IE's Trident engine, but Gazelle manages all those separate processes independently, kind of like a virtualized OS. It's certainly interesting stuff, but it's still all just a research project for now -- Chrome OS is still vapor, but it's clear that Google intends to ship something, while Gazelle seems more suited to inspire future versions of IE. Still, it's interesting reading if you're into it, so hit the read link for more.

Read - Ars Technica analysis
Read - Microsoft Gazelle white paper [Warning: PDF]

[Image courtesy of Robert Scoble]

Some researchers at UC San Diego, teamed up with Microsoft Research, have a novel new method in the works to let your computer save on power in sleep mode while still keeping up with IMs and whatever other network activity you pesky folks might be up to. Dubbed "Somniloquy," the USB dongle acts as an external networking card, and takes over most of the communication duties, with its own embedded OS and flash memory (similar to the Killer NIC). When the computer goes to sleep, Somniloquy maintains IM connections, ongoing downloads and torrents, and keeps your computer present on the local network and wireless LAN. If an activity arrives that needs the full computer, Somniloquy wakes up the computer and patches it through -- going so far as to buffer downloads into flash memory, and then dump it periodically on to the hard drive. The wake-on network activity concept is nothing new, but Somniloquy really runs with it, and boasts up to 60 to 80 percent overall energy savings based on the person's usage habits. The device is currently in prototype stage, and in the future all its functionality could be integrated into a PC's network card -- and hopefully won't just be for the pros this time.

Who knew closely investigating expo floor layouts could be so fun and rewarding? Sure enough, the 2008 DemoFest Booth Map hosted up for Microsoft's upcoming (July 27th - 29th) Research Faculty Summit shows allotted space for one "Multi-Touch Spherical Display." Of course, Redmond-based exhibitors may end up dedicating slot 110 to carbonated beverages and unhealthy snacks, but whether those sly cats are ready to admit it or not, we know this thing is real.

[Via ZDNet]

The boys and girls at Microsoft Research are getting set to publish a report detailing a competing (and in their perspective, superior) technology to replace the LCD monitors we've grown to know and love heart with reservations. By using pixels that boast "a pair of mirrors to block or transmit light," displays could theoretically be created which are "faster, brighter, and more power efficient than liquid crystal displays." Dubbed telescopic pixels, the devices would be able to turn off and on in under 1.5-milliseconds, which is quick enough to put "red, green, and blue light-emitting diodes behind each pixel." Furthermore, these pixels are way brighter than those used in LCDs, which means users would see less power draw and be able to view the screen more easily in direct sunlight. Sounds solid from here, but could you not humor us with a release time frame or something?

[Via TG Daily]

We'll go ahead and get this out of the way: the fantastic product you're about to hear more on has "no plans" to go commercial. Now that we've thoroughly killed your buzz, let us introduce to you the LaserTouch. Said device is a prototype that recently emerged from Microsoft Research's labs, which essentially allows people to retrofit any display (monitor, projector, etc.) so that they can use their own hands to control the on-screen action. According to Andy Wilson, who played a vital role in the unit's creation, an infrared camera is used to track how a person touches the screen, while software that he developed handles the majority of the magic. Too bad this could totally undercut Surface sales, right?

If you've been watching CNN's election coverage lately (and we think you have), then you've probably seen anchor John King tripping the light fantastic on the channel's Magic Wall touchscreen. Apparently, Microsoft has come up with its own version of the board -- essentially a Surface flipped on its side and mounted. The device, known as the TouchWall, joins a handful of finger-friendly prototypes that Microsoft Research is working on (including a spherical unit we assume will be used strictly for world domination campaigns). The 4-foot-by-6-foot "wall" uses inexpensive infrared sensors and a rear-mounted camera -- which can be added to a variety of surfaces -- in order to create the hands-on experience. The company appears to be targeting this as a low cost "intelligent whiteboard" solution, so hopefully they'll be turning up in schools, small businesses, and the Engadget offices soon. Check the video after the break to see it all in action.

[Via CNET]

Surely you're not content with just multi-touch and built-in accelerometers, right? Microsoft Research is expecting not, as it's already working towards integrating force-sensing technology that will enable gestures to twist on-screen imagery without crafting UMPCs out of flexible material. Essentially, the gurus behind the idea feel that implementing said tech would "turn an otherwise passive component that just holds the device together into an active input surface." In the future, it's likely that auditory cues would enable users to know when they've applied enough pressure to cause a change, and of course, they expect it to work hand-in-hand with existing human-computer interfaces. Yeah, who needs keypads these days, anyway? [Warning: PDF read link]

[Via BBC, thanks Joe]

Can you imagine controlling your computer with a peripheral that resembles a fluffy bar of soap? Well hear us out, because the Soap pointing device from Microsoft Research offers to combine the accuracy of a traditional optical mouse with the freedom of a wireless, even desk-free peripheral. The device, which consists of a soft, fluffy hull coating a wireless optical mouse core, reads the fabric's movement when the user applies outside pressure, and reports it as a position (think: rolling a bar of soap around in your hand). Through a demonstration in combination with a mobile keyboard in Unreal Tournament (albeit with dumb bots), the creators have shown that Soap offers a degree of accuracy that comes close to a table-bound mouse; when the user squeezes, the Soap can even outclass desktop mouse performance in large wall display setups. If users can cope with the idea of blasting away fiends in UT with a ball of fluff and Soap gets a commercial release, if nothing else you can be sure that an entire cottage industry will spring up around supplying fake eyes and tails to turn the Soap into a line-up of cute, pointable rodentia.

[Via hackaday]