Google and Microsoft this week traded counter-punches over voice and social networking tools as the two continue to battle each other over emerging technologies.

Google Friday unveiled a voice recognition application for the iPhone that lets users speak search terms into the phone such as “restaurant.” The technology is similar to voice recognition software for the BlackBerry that Microsoft introduced at the Web 2.0 conference in April via its TellMe subsidiary.

On the flip side, Microsoft on Thursday unveiled the Windows Live Wave 3 set of consumer services, a Live Wave API called Project Silkroad, and struck partnerships with social networking sites such as Flickr, PhotoBucket, Twitter and Yelp. The announcement came roughly a week after Google’s OpenSocial celebrated its one-year anniversary.

OpenSocial is a set of APIs that let developers create applications that run across many social networking sites instead of having to create a version of the applications for each individual site.

A team including former Google, Yahoo and Facebook employees is launching a startup that will provide professional support for Hadoop, an open-source computing platform that makes it easier to write programs that process large amounts of data.

Yahoo is among the companies that use Hadoop, which employs the Google-developed MapReduce framework, in their operations.

Now Cloudera’s founders are betting that there is money to be made supporting other Hadoop installations, whether on a customer’s own set of machines or Amazon’s Elastic Compute Cloud (EC2).

“Think Red Hat for Hadoop, but that is just the beginning,” co-founder Amr Awadallah, wrote on his blog this week. Awadallah was previously vice president of engineering on a BI (business intelligence) team at Yahoo charged with improving the user search experience and making more money off the search site.

Former Googler Christophe Bisciglia is also on board. Bisciglia was a main driver behind the Academic Cluster Computing Initiative, a Google-IBM project that has provided Hadoop-powered machine clusters for academic use.

In addition, Jeff Hammerbacher, previously manager of Facebook’s data team, is a co-founder of Cloudera. “The [Facebook data] team made a lot of contributions to Hadoop under his stewardship, most important of which is Hive (a SQL structured data layer on top of Hadoop),” Awadallah wrote.

Other company founders include Mike Olson, former CEO of Sleepycat, a company known for the open-source Berkeley DB database engine. Oracle bought Sleepycat in 2006.

Cloudera won’t immediately feel the expected effects of the struggling economy on venture capital spending, Awadallah said. “We are in the process of wrapping up our funding (we don’t need any more cash at this point), and will soon announce our investors and technical advisors.”

Other details about the company’s plans, such as pricing models, were not available. Cloudera could not immediately be reached Wednesday for a follow-up interview. The company will be based in the San Francisco area, though a specific location was not listed at its Web site.

News of the startup’s plans sparked generally positive reactions from industry observers.

“Given the current economic outlook it’s great to see a new open source start-up rearing its head, and the list of founders indicates that this one has a good chance of survival,” 451 Group analyst Matthew Aslett noted in a blog post Wednesday.

Redmonk analyst Stephen O’Grady weighed in via e-mail. “It’s clear that demand for the services that Hadoop provides is only going up. Most enterprises, even large ones with substantial investments in IT, lack the scale-out skills common to eBay, Google, Yahoo, et al. Commercially available and supported Hadoop would be one attractive solution to that problem.”

Less than a week after the release of Google (NSDQ:GOOG)’s new Web browser Chrome, security researchers detected a buffer overflow vulnerability that could enable remote attackers to completely take control of a user’s computer.

The detected buffer overflow vulnerability, deemed critical by security experts, is the result of a boundary error in the handling of the “Save As” function. If a user saves a Web page serving malicious content, the program could cause a stack-based overflow error, which could open the door for remote hackers to unleash malicious code on a user’s machine.

Remote attackers could then exploit the flaw by constructing a specially crafted Web page infused with malicious code. The attacker could then entice a victim to open and then save the infected page, which would subsequently download malicious code onto the victim’s computer and give the attacker complete access to the affected system.

Chrome’s latest buffer overflow vulnerability is one of about half a dozen errors detected in the newly released beta Web browser, about half of which allow for remote code execution, experts say. Another vulnerability, discovered shortly after the browser’s release Tuesday, included a carpetbombing glitch that stemmed from a fundamental flaw in the underlying user agent Safari 3.1.

However experts say that several Chrome beta version flaws are anticipated and will likely be worked out with the final version as the browser is subsequently tested.

“I think for a new product like Chrome, it doesn’t concern me much that they’re discovering the number of vulnerabilities and the details are getting out there. That’s the point of beta, especially open source beta,” said John Bambenek, handler for the SANS Internet Storm Center. “I think that the people who are really into getting exploits on a number of machines are not interested in messing with Chrome until (Google) gets some distribution out there.”

“If it’s not public information, the hackers don’t have it either,” he added.

And despite some errors that could lead to remote exploitation, experts say that because the browser is still in beta and not yet widely adopted, security threats for most users for the time being remains small.

“I don’t think the consumer impact is very large yet,” said Bambenek, “but that could change very quickly.”

Google Trends was originally released as a tool that let you see visual comparisons between search volume of keywords. That hasn’t changed, but Google now also gives us the keys to more data about actual popularity of websites based on daily unique visitors.

This data is similar to what companies like Alexa already provides, but it’s not ready to be a complete replacement yet. There is not as much data available through Google Trends as there is in other tools. Searches are currently limited to the domain level — so blogs.zdnet.com would translate to zdnet.com before the search happens.

If you have a popular enough website to be included in these results from Google Trends, you may be disappointed to know that there is no way to remove your site if you wanted to. Google doesn’t think that rule should apply to them though, as they have removed all their web properties — searching for things like google.com or youtube.com comes back with nothing.

Facebook CEO Mark Zuckerberg wants to sit down with Google and work out the privacy issues that caused Facebook to block Google’s Friend Connect last week, he said Monday.

“We want to talk to Google about this and see if there’s a way we can make it work,” said Zuckerberg at a news conference in Tokyo. He was in the Japanese capital to launch the a local-language version of the social networking site.

Google Gets Friendly

Google Friend Connect allows Web site operators to add social networking functions to their Web sites. Users visiting the sites will be able to interact with new people or existing friends from social networking sites like Facebook, Orkut and Plaxo. It’s the possibility of data redistribution to third-party sites by Google that caused Facebook to block access, it said last week.

“Part of the issue with Google’s Friend Connect is that when users grant access to Google’s product, Google might share their information with another application, or some part of it, maybe not all of it, without that user knowing. And part of what makes our system work is that people know exactly who they are sharing all their information with,” he said.

Zuckerberg contradicted Google Engineering Director David Glazer, who said last week in a phone interview with IDG News Service that Google had spoken to Facebook about the service prior to its launch.

“They launched that without asking us or talking to us about it first so we had no choice but to follow the rules that we had set forth for any developer on top of our platform and we followed them,” said Zuckerberg. “But Google’s a big player in the space and they make good things and our goal is to work with them to figure this out.”

Zuckerberg also noted that Facebook has had a similar service, Facebook Connect, available since late 2006.

“We think it’s good that other people are picking up on this trend now,” he said.

International Competition

Facebook faces a tough market in Japan. The number one social networking site, Mixi, has the market virtually sewn up with more than 10 million users, and some doubt whether Facebook’s top selling point, that people use real names, will appeal to Japanese users, many of whom only feel free to express themselves when hiding behind a pseudonym.

Zuckerberg hinted that an update to the company’s site for cell phone users, Facebook Mobile, might be coming for Japan and other advanced mobile markets.

“We have a mobile version of the site and that’s just a first version. We realize that in a lot of more advanced, technical cultures that phones are in many ways more important than the Web but this is just a first approach to that,” he said.

The number of people accessing the Internet via cell phones in Japan outnumbers those accessing the network from personal computers, so a strong mobile site is also important.