Contrary to popular belief, Google Wave is not quite dead. Google plans to expand upon the open source code it has already released to form a more complete, standalone application known as "Wave in a Box." Wave will work (or not work, depending on how you look at it) as it always has, but with the new app, developers can run their own wave servers and host waves from their own machines.

"Since the beginning, it has been our vision that the Google Wave protocols could support a new generation of communication and collaboration tools," wrote Wave software engineer Alex North on the Google Wave Developer Blog. "The response from the developer community to date has been amazing and rewarding. Even more so now, we believe that developers and other projects are a critical part of this story."

North says the project will include an app bundle with both a Wave server and client, a "fast and fully-featured" wave panel in the Web client, persistent wave store and search for the server, and the ability to import wave data from the Web. Additionally, it will have gadget, robot, and data API support, as well as the ability to federate across other Wave in a Box instances. Even more details are available in Google's Wave Protocol Forum.

Google announced earlier this week that the Web version of Wave would remain online through "at least" the end of the year, and that users would be able to export their waves during that time. It was a blip of good news to those few lamenting Wave's short lifespan, but Wave in a Box should be a major step toward helping the project live on.

Read the comments on this post

Amazon has reduced the prices of its EC2 compute nodes by up to 19%, the second price cut this year.

Read More: AWS Blog, EC2 pricing

Read the comments on this post

AMD has announced that it is retiring the ATI brand, and expects to have fully transitioned all of its graphics products to the AMD label by the end of this year. And in an unrelated but no less important bit of news, the company has also confirmed that it is indeed looking at using Bobcat in servers.

The timing of the ATI brand announcement makes sense, given that the company is moving full speed ahead into the Fusion era with the upcoming launch of its Ontario mobile platform, which features an integrated CPU and GPU on the same die. So from a purely technical perspective, it makes little sense to talk about your processor as an "Application Processing Unit" because it combines a CPU and GPU on the same die, and then give your discrete GPU chips a different brand name than the CPU/GPU combo chips. If the CPU and GPU are going to merge in silicon by the end of the year, then the brands should merge, as well.

That said, the move has to give Intel at least a tiny bit of heartburn. Current systems integrators, Apple being the most obvious example, advertise the Intel and ATI brands alongside one another for systems that use Intel CPUs and ATI GPUs, so that will now change to Intel and AMD branding. With as much as Intel has invested over the past decade in the "Intel Inside" program, the company can't be happy at the idea of an AMD logo accompanying the ubiquitous "Intel Inside" badge on laptops, flyers, webpages, etc.

As for AMD's server news, the company's newly hired server chief told IDG today that his unit is currently investigating Bobcat's potential as a server part. This is great news, as it's something that we recommended in our recent look at Bobcat.

"We're definitely in the process of examining this as a design point... It would be foolish not to," AMD's Donald Newell told IDG in an interview.

He later indicated that the company needed to collect a lot more data before it could decide if Bobcat makes sense for a large enough subset of server workloads.

Read the comments on this post

You know the feeling: opening up your e-mail to find hundreds of messages of varying importance. Some are automated reminders from your favorite sites, some are newsletters you have subscribed to, some are actually from real people trying to contact you, and so on. Separating the wheat from the chaff can be overwhelming much of the time, and even the most carefully crafted filters don't keep up with the ever-changing nature of what's important to you.

Google is hoping to address that problem with a new feature in Gmail called Priority Inbox. Aimed at providing users a way to get through their inboxes as efficiently as possible, Priority Inbox tries to learn your e-mail habits in order to decide which messages are important to you, and move them up to the top where you can see them first.

Read the comments on this post

In addition to releasing the most security bulletins ever on this month's Patch Tuesday, Microsoft has released a number of non-security updates, the majority of which are for the latest versions of its client and server operating systems. All the patches are available on Windows Update and the Microsoft Download Center and most will require a restart. With the exception of the last patch, they're all for Windows 7 or Windows Server 2008 R2.

Most of these updates will be rolled into Service Pack 1 for Windows 7 and Windows Server 2008 R2. Testers got the first Windows 7 SP1 beta build two months ago, but just today Microsoft sent out build 7601.17077 to selected PC and Technology Adoption Program partners, according to ZDNet.

The first patch (KB2028560) is vaguely described as one that delivers "new functionality and performance improvements for the graphics platform." 

The second patch (KB2249857) describes an issue that occurs on 2TB+ hard disk drives. If the OS is configured to save dump files to a volume of such an HDD, some of the dump file is offset at a disk offset greater than the 2TB address, and Windows is either put into hibernation or crashes, volumes on the HDD may be corrupted, and data is lost. If the corrupted volumes include the system partition, the computer will no longer boot. 

The third patch (KB982110) fixes a problem when running 32-bit applications on a 64-bit edition of Windows 7 or Windows Server 2008 R2. If the application uses the QueryPathOfRegTypeLib function to retrieve the path of a registered type library, it may return the path of the 64-bit version of the type library instead of the 32-bit one. 

The fourth patch (KB2272691) is for a game, application, or firmware that is either installed incorrectly, causes system instability, or has primary functions that do not work correctly. The update will either prevent incompatible software from running (hard block with third-party manufacturer consent), notify the user that incompatible software is starting to run (soft block), or improve the software's functionality (update). It lists just a single application (Sensible Vision FastAccess) as being affected.

The fifth patch (KB2203330) solves a problem when installing a third-party application for the multiple transport Media Transfer Protocol (MTP) device or for the Windows Portable Device (WPD). Connecting an MTP or WPD device may result in an APC_INDEX_MISMATCH stop error message because of a race condition in the Compositebus.sys driver. 

The last patch (KB979453) is for Windows Home Server and addresses five separate issues that were found since the release of WHS Power Pack 3.

Read the comments on this post

After nearly a month of rumors foreshadowing the deal, it's now official: Intel is buying Infineon's wireless unit for $1.4 billion in an all-cash transaction that will close in the first quarter of next year. To really understand what this purchase will do for Intel, you need a bit of technical background.

Wireless devices like smartphones and laptops historically have two major functional blocks inside them: an application processor (AP) and associated logic, and a baseband processor (BP) and associated circuitry. The application processor is an SoC that corresponds to the CPU northbridge in a PC—it has a general-purpose CPU core that runs the OS and applications, and it typically also has some embedded graphics functionality (often a full-blown GPU) and a few I/O blocks (e.g., USB, camera, GPS, Bluetooth, HDMI, etc.). The application processor is also connected to a pool of main memory (DRAM) and backing storage (flash). All told, the application processor, along with any GPU (on- or off-die), and DRAM (on- or off-package), makes up the part of the mobile device that we typically think of as a "computer," because it runs the OS and application stack.

Read the comments on this post

The LHC isn't simply the most powerful particle accelerator ever created. Handling the huge amounts of data it produces has required the creation of one of the biggest computer grids on the planet. The planning and testing of the compute facilities has been taking place for years, but it's only recently that the grid has had to deal with the output from actual collisions. How did it do? "From the IT perspective, we didn't notice when the beams came on," said CERN's Wolfgang von Rueden, "We had tested it with much higher throughput conditions."

Still, not everything is working quite according to plan. von Rueden said that the initial expectations for the LHC's computing grid had anticipated lower network performance and a reliance on tape; instead, the network has made it easier to shuffle large data sets between compute centers, and the price and performance of hard drives have turned out better than expected. Von Rueden gave us a brief overview of the computing setup at CERN, what they've learned from putting everything in place for the LHC, and how some major companies are relying on CERN's experience to improve their products.

Read the comments on this post

Generally, we assume that making decisions as a group is beneficial because groups can come together to make better choices than their members would alone. Is this true, or is group decision-making a case where a chain is only as strong as its weakest link? A new paper in Science this week has come to a conclusion: it depends.

Each trial of the experiment paired up two of 72 male participants. Over two timed intervals, the researchers used computer screens to show each of the two partners a visual field with six small circles. During one of the two intervals, one of the six circles appeared to be slightly darker than the other five. The participants each had to identify which circle was darker, and during which interval this difference occurred.

In the first section of the experiment, the partners saw identical images on their computers, made their choice, and then were given the chance to communicate before coming to a final, joint decision. By examining each participant's initial answers, the researchers could determine which of the two participants was individually better at the task. During this first set of trials, two heads were indeed better than one: the teams consistently chose the correct circle and interval more often than the more successful partner did on his own.

To determine whether the skill levels of the partners mattered, the researchers pulled a clever trick and secretly introduced "noise" into one of the partner's visual fields, making him perform more poorly at the task. Unlike the first set of trials, the pairs in these trials greatly underperformed when compared to the better participant (who saw no noise). This suggests that when one partner is performing poorly but can't recognize it, the pair's joint decisions will suffer.

In the final set of trials, the partners were not allowed to communicate at all, and only one person had the ability to make a decision based on both his choice and his partner's input. Here, the pairs again performed poorly because the final decision-maker had no information about how certain, or uncertain, his partner was. So, not only should partners be equally adept, they must also be able to communicate freely before making a decision.

The take-home lesson? Work with someone who's as good at their job as you are at yours, and make sure you're both willing to discuss your uncertainties.

Science, 2010. DOI: 10.1126/science.1185718  (About DOIs).

Read the comments on this post

The server market has been AMD's bread and butter for the past few years, and it has been the place that the company has really remained competitive with Intel. AMD's mobile strategy has mainly involved using versions of its server parts to target the low-cost laptop market, so the company has missed out on the netbook revolution of the past few years.

With the launch of products based on its new mobile architecture, codenamed Bobcat, AMD will finally join Intel in having two separate architectural families: one for servers and desktops, and another ultra-low-power design for mobiles. Bobcat, then, is AMD's answer to Intel's Atom, but instead of targeting where Atom is now, the company has chosen to target where Atom is likely to be in the not-so-distant future.

Read the comments on this post

In a presentation at Stanford's Hot Chips conference on Tuesday, ARM added a few more drops to the trickle of information that's coming out which suggests that the UK-based mobile and embedded processor designer is very seriously pursuing the server market. Specifically, ARM's David Brash described a new set of virtualization extensions for the ARM-v7-A architecture, which will be included in the follow-on to Cortex A9. Brash also described an OS-managed address extension that will alleviate some of the I/O and memory pressure that goes with ARM's 4GB memory limit.

This upcoming architecture is codenamed "Eagle," and Brash indicated that it will be unveiled soon. The virtualization and address extensions for Eagle lend credence to what has been heard from other sources, which is that Eagle has some features that are clearly aimed at higher-end, non-mobile or embedded applications and that many ARM vendors are shopping detailed ARM-based server roadmaps to some major cloud customers.

Read the comments on this post

The Microsoft Exchange team has announced that Exchange 2010 Service Pack 1 is available for download now. This comes after a short public beta period; announced in April, and released in June.

Among other things, the Service Pack adds new replication features to improve robustness, a new mailbox online repair mechanism for when things do go wrong, the ability to import historic e-mail directly from a PST into a mailbox, better load balancing, improved mailbox searching, IRM support in OWA when using Safari, the reintroduction of themes in OWA, and a raft of other improvements.

The new version should also usher in the availability of hosted Exchange 2010 solutions. Multi-tenant support, which allows one Exchange installation to provision services for multiple customers, means that service providers can use regular Exchange to provide cloud-hosted Exchange. Previously, this required extra components that only supported Exchange 2007, with the result that hosted service providers have not been able to offer Exchange 2010 (with perhaps one exception). Service Pack 1 should enable such providers to offer Exchange 2010 capabilities.

This fits in neatly with Microsoft's forthcoming Small Business Server "Aurora," which unlike previous Small Business Server versions, eschews a locally-hosted Exchange in favor of cloud services. This obviously becomes more attractive if those cloud services support Exchange 2010, as it means that Aurora will offer the same capabilities as standard Small Business Server, but without the administrative overhead.

Read the comments on this post

Exchange 2010 Service Pack 1 was released today, with improved robustness, PST imports, better manageability, better cross-browser compatibility in OWA, and many other improvements.

Read More: Download, MS Exchange Blog, Past coverage

Read the comments on this post