Thursday, June 5, 2008

Yahoo Denies 'poison Pill,' Icahn Accusations

Yahoo Chairman Roy Bostock rejected charges by shareholder Carl Icahn that an employee retention package was designed to discourage a Microsoft takeover, in a response to Icahn released Wednesday night.

"To set the record straight, the employee retention program is designed to protect the Company's assets and value during a time of uncertainty. The claim that the plan gives each of Yahoo!'s employees 'the right to quit his or her job and pocket generous termination benefits at any time during the two years following a takeover...' is just plain wrong," Bostock wrote.

Yahoo adopted a new employee severance package after Microsoft announced its US$44.6 billion hostile takeover bid on February 1. The package would have cost Microsoft up to an additional $2 billion, offering employees terminated or who resigned for "good reason" four months' to two years' severance pay depending on their position. Icahn saw the move as a "poison pill" -- usually a corporate maneuver designed to stave off a hostile takeover.

"In fact, the plan was adopted in order to protect the value of Yahoo in anticipation of a possible acquisition by Microsoft which would have resulted in a lengthy regulatory review and a significant period of uncertainty for our employees," Bostock wrote.

Bostock then fired a new salvo in the war of words between Yahoo's board and Icahn, who charges that the board hurt shareholders by not accepting Microsoft's acquisition offer. "Conspicuously absent from your letter is any credible plan for Yahoo other than a repetition of your insistence that the Company should sell itself to Microsoft. Indeed, your stated view that 'the only way to salvage Yahoo in the long if not short run is to merge with Microsoft' demonstrates that you have no other plan and causes one to wonder what exactly would happen to our Company if you and your nominees were to take control of Yahoo," he wrote.

Icahn seeks to oust Bostock, along with CEO Jerry Yang and other board members, and replace them with his own slate of directors,

"I have constantly complained about how far CEOs and boards will go in order to retain their jobs, yet even I am amazed at the length Jerry Yang and the Yahoo board have gone to in order to entrench their positions and keep shareholders from deciding if they wished to sell to Microsoft," Icahn wrote in his letter to Bostock early Wednesday.

Microsoft walked away from its acquisition attempt on May 3, after three months of negotiations to buy the Internet company. Since then, Icahn and other disgruntled shareholders have sought various means of redress, accusing Yang, Bostock, and the board of not protecting their interests.

MSI Plans Next Wind Mini-laptop, Smaller Devices

Micro-Star International (MSI) is already planning a smaller, thinner version of its popular Wind mini-laptop aimed at business people, an official from the company said Thursday.

The Taiwanese company is also planning a new mobile Internet device that's smaller than a mini-laptop.

For its next generation Wind mini-laptop, MSI plans to reuse many of the same components as the original, including Intel's Atom microprocessor, a power-efficient chip designed for the new class of mini-laptop PCs and mobile Internet devices. So the focus this time is on design, and beefing up some functionality so the device appeals to businesses, said Andy Tung, sales director at MSI in the U.S.

"On Intel's roadmap, there won't be any change in Atom until the second half of next year, so the microprocessor and chipset will be the same," said Tung.

The new Wind laptop will likely be out in the first quarter of 2009, he said, and could be at MSI's booth at the International Consumer Electronics Show in Las Vegas, which runs Jan. 8-11 next year.

The company isn't trying to compete with Apple's Macbook Air to be the thinnest laptop around, but MSI will try to slim the Wind down for portability and design reasons.

MSI is also working on a concept for a device smaller than a mini-laptop to attract people who want to surf the Internet wirelessly on a device that's easy to carry around, said Tung. Right now, the company is considering a device with a screen smaller than 7 inches and a keypad that slides out from below for typing.

MSI's Wind was officially launched this week at Computex, part of the growing segment of mini-laptops, which have 7-inch to 10-inch screens, weigh less than 1 kilogram (2.2 pounds) and connect wirelessly to the Internet.

The mini-laptop comes with a 10-inch LCD (liquid crystal display) screen, an 80G-byte hard disk drive and up to 2G bytes of DRAM, and connects wirelessly to the Internet via Wi-Fi 802.11b/g.

Intel declined to comment on an official time frame for an update to its Atom microprocessor.

Tuesday, June 3, 2008

Curl Adds Runtime Support for Mac Environments

Curl, maker of a framework for building rich Internet applications, said Tuesday that it is now supporting Macintosh environments, citing "rapidly increasing penetration" of the platform in enterprises.

Curl's platform, which is aimed squarely at enterprises, consists of an eponymous programming language, an IDE (integrated development environment) for building applications, and runtimes for executing the programs on client machines. The new Mac runtime, available as a free download, is compatible with Power PC and Intel Macintoshes running OS 10.4 and higher, Curl said.

"As we watch Apple whittle away at Microsoft's dominant position, it behooves us to start supporting that platform sooner than later," said Richard Monson-Haefel, vice president of developer relations.

He noted that Curl already had runtime support for Windows and Linux. The move to Apple is "a matter of completing our cross-platform story," he said.

But at least for now, the Apple chapter of that tale may be brief.

A recent Forrester Research survey found that the percentage of enterprise users running Mac operating systems grew threefold in 2007 to 4.2 percent, but concluded that "uptake remains limited to enthusiasts and small workgroups."

The new Mac runtime includes all the features present in Curl's runtimes for Windows and Linux, including support for audio, 2-D and 3-D graphics and animation, and a compiler that transforms applications built with Curl into native code for maximum performance.

However, Curl's IDE isn't yet supported to run on Macintosh, meaning developers who prefer to work on those machines will have to wait. "That's certainly a direction we're heading. The [runtime] development was ahead of the IDE development," Monson-Haefel said. "We're looking at it slowly and at what our customer demand is."

The company's move to support Macintosh has two interesting angles, said Redmonk analyst Michael Coté.

"Curl is going for being the enterprise RIA player, sort of trying to niche itself out of competing with [Microsoft's] Silverlight and Adobe. The thing is that their platform is ostensibly ready and mature, but they don't have the brand power of Adobe and Microsoft.

"Given that they want to be the enterprise RIA people, their support of Macs is interesting because it implies that they see a sizable enough use of Macs in the enterprise to make money off it -- or see growth in the near term therein," he added. "Most people wouldn't consider Macs an 'enterprise' deployment space and, thus, a platform to support. Now, developers of RIAs, sure, that's a whole other story. But users is something slightly new."

Also, Curl's cross-platform emphasis should play well as the RIA market evolves, Coté predicted.

"At this point in the RIA wars, attracting developers is the big battle, and Windows-only toolchains ... are at a disadvantage, as would be OS X or Linux only," he said.

While much smaller than those vendors, Curl has about 300 "large enterprise deployments" and its platform, now in version 6.0, is more mature, Monson-Haefel said.

The vendor is based in Cambridge, Massachusetts. Its platform is based on research conducted at the Massachusetts Institute of Technology. It is owned by Sumisho Computer Systems, a Japanese company that bought it in 2004.

Yet Another Personal Area Network Technology

Ozmo wants to be part of the latest round of Personal Area Network (PAN) technology, which this time leverages Wi-Fi. Ozmo gets a lift by being partly Intel funded, and working closely with Intel on the chip giant's Cliffside development project, which allows a single Wi-Fi radio to serve two separate purposes--in this case making a network connection and serving a PAN at the same time. (Cliffside is part of a larger initiative--"Carry Small, Live Large"--to provide a laptop experience in a small but powerful mobile device.)

PANs are designed for short distance associations with a computer or handheld device; typical uses are for low-bandwidth peripherals like input devices (keyboards, mice, pens, and trackballs), synchronization with handhelds and smartphones, and audio via headsets and headphones, as well as high-bandwidth applications like streaming high-definition video, wireless computer displays, hard drive connections, and so on.

Ozmo is square in the middle of this. The idea with Cliffside in general and Ozmo's solution in particular is that a Wi-Fi radio is wasted by handling just wireless LAN traffic. What if that radio could serve two masters? It could handle a WLAN connection and act like a PAN hub for peripherals. An Ozmo spokesperson told me that while Intel is their first partner on their product--Intel will embed Cliffside technology within a year into their laptop designs--that other Wi-Fi chipmakers could work with them to develop drivers that would allow existing and new laptops to offer WLAN/PAN connections. One of Ozmo's key markets is selling chips to peripheral makers.

This technology tries to give the boot to Bluetooth, which is embedded in well over a billion devices sold to date, and is expected to be built into half a billion cell phones sold in 2008 alone, excluding the tens of millions of computers sold with Bluetooth radios. This seems like a lot of inertia for a technology that has, as its biggest problem, the difficulty of pairing two devices. (That's mostly a legacy issue: Since version 2.1 of the standard, Bluetooth devices can be paired with a simple PIN instead of the former process that took as many as 15 steps on some operating systems.)

Where Bluetooth's current 2.1+EDR (Enhanced Data Rate) standards operates at 3 Mbps, Ozmo is proposing 9 Mbps for its Wi-Fi based flavor, which they claim would provide the overhead for crisper audio among other purposes. Ozmo is also claiming better battery life for peripherals relative to Bluetooth. (Another PAN technology, ultrawideband (UWB), has only barely reached the market. It can offer speeds from 100 to 500 Mbps, depending on distance, but can't yet be built into tiny, battery-powered peripherals. It's really an extension of USB, and some flavors will be sold as Certified Wireless USB.)

Ozmo's argument rests in part on the fact that Bluetooth isn't built into all the devices that have Wi-Fi; nearly all laptops are sold with Wi-Fi radios, but only a third (according to some research) have Bluetooth radios as well. With major Wi-Fi chipmakers all having in-house Bluetooth divisions or partnerships with Bluetooth chipmakers, it seems like a short-term blip that Bluetooth has an undercount compared to Wi-Fi.

Bluetooth has some tricks up its sleeve, too: Bluetooth High Speed. This hybrid mode allows a Bluetooth driver to recognize when a bulk data transfer needs to be made and switch to using a Wi-Fi adapter to handle that part. This uses Wi-Fi at its best advantage, while still preserving the typically lower power usage of Bluetooth.

It's also a bit ironic that Intel, which is pushing future devices that will contain Wi-Fi, WiMax, and possibly 3G cell data radios on a single platform, is carping that much about a second radio that shares quite a lot with Wi-Fi already.

Report: Icahn Will Seek Yang's Removal

Billionaire investor Carl Icahn will try to remove Yahoo cofounder Jerry Yang as the company's CEO if his proxy battle is successful, The Wall Street Journal reported on Tuesday.

In an interview with the newspaper, Icahn accused Yang of being unwilling to consider Microsoft's acquisition offer and costing the company money in his efforts to deter the deal. He also said that Microsoft is hesitant to make Yahoo another offer because it doesn't trust Yahoo's current management, according to the Journal.

Unless the board is ousted and Yang removed as CEO, Microsoft will not make a deal with the Internet company, Icahn told the newspaper.

Icahn said he plans to make a public statement about his concerns with Yang and the current Yahoo board in the next day, according to the Journal.

Icahn's public ire against Yang comes a day after Yahoo lost a battle to keep private some information about the company's operations that was revealed in a shareholder lawsuit against Yahoo, filed in a Delaware court.

The 64-page complaint contains blistering allegations, copies of internal Yahoo documents and e-mails, and blow-by-blow accounts of what plaintiffs characterize as Yahoo bad-faith maneuvers. The complaint had been made available previously but with portions redacted.

Icahn's proxy fight against Yahoo officially began May 15, when he sent a letter to Yahoo's board announcing he was nominating 10 candidates to replace all incumbent directors -- which include Yang -- at the company's shareholders meeting in July. Icahn did not indicate at the time that he was going to replace Yang as top executive of Yahoo, however. Before he sent the letter, Icahn began buying up millions of shares of Yahoo stock.

A few days after Icahn's letter, both Microsoft and Yahoo said publicly that they were open to negotiating another deal, although not one for Microsoft to purchase the entire company but instead to buy only pieces of Yahoo. Comments Icahn made to the Journal on Tuesday indicate that Yang and the current Yahoo board are acting inconsistently with those public comments.

Yahoo's board is scheduled to meet Tuesday, the Journal reported, quoting anonymous sources.

Monday, June 2, 2008

ProcessMaker uses open source inside and out

ProcessMaker is an open source workflow manager that works either on the client side or as a hosted application. Founder Brian Reale began developing ProcessMaker in 2002 after working with the South American Telecommunications Regulatory Institution to create a "paperless office." Once that system was deployed, Reale thought he could create an affordable standalone product that would make it easy for users to eliminate paperwork and create a more efficient workplace. Reale built the new product using open source software, and has licensed it under the GPLv3.

ProcessMaker, the company, uses open source software "for just about everything we do," Reale says. Most of the workstations run Ubuntu and OpenOffice.org, along with Knowledgetree, MediaWiki, Drupal, phpBB, Zimbra, and SugarCRM. "We tend to us open source as much as possible ... to streamline our operations." Reale says telephony server Asterisk is "our most mission-critical application running on open source. We run a PBX in our Bolivian office, and another one hosted in the United States for our Boston office. We have the two connected by an IAX trunk and all of our office extensions connect to each other either locally or across the VoIP IAX trunk. It works great."

Reale says using open source to run the company brings a feeling of security that comes from "knowing there is a community around the product we use. With open source, there will usually be someone somewhere who can carry on giving support, and this gives peace of mind, especially when we are trying cutting-edge software." On the flip side, creating open source software and extending it to the community can be a challenge. "We're pretty new on the open source scene. Most of our challenges have been around building a community and figuring out how to manage it. We want to foster real community, with developers who can own the project. One of the ways we addressed that issue was by offering a full-featured open source version of ProcessMaker.

"For the first few years of operation, we sold turnkey solutions and gave our clients the source code. We never believed that the code should be encrypted or hidden, and we always encouraged them to make their own modifications if they wanted to. But most didn't have the time or inclination to do it all in-house, so we would end up being hired to do most of the implementation work." When Reale discovered that most of his income was derived from these add-on services, he began to look seriously at the idea of open sourcing ProcessMaker. "We believed that if we could gain more input from other develoeprs on how to improve the code, and gain access to new distribution channels, then it would certainly make sense to give away the code."

Reale and his team began cleaning up the code before they invited others to take a look. "Open sourcing felt a little bit like houseguests going through your closet -- you want to be sure everything is in order. We also began attending some of the open source conferences in late 2005 and began speaking with other open source companies." Reale found several potential partnership deals, which gave him the confidence he needed to jump into open source with both feet, and ProcessMaker went fully open in February 2008.

The free Community version is identical to the Enterprise version, but without paid support, training, consulting, and hosting. "Initially there was internal debate around whether to offer additional management features in the Enterprise version," Reale says. "But we decided that it was important that the Community version be full-featured in order to foster our community and create the best experience for all users."

Reale says his existing clients were pleased with the change. "We took great care to explain to our existing clients that with the open source model, not only would they continue to receive software at a lower cost, but they would also benefit from the innovations of the ProcessMaker community."

Reale says the best part about going open source is "being involved with a project where everyone is passionate about what they're doing. Everyone at ProcessMaker loves to see how our project has grown. Every day it changes, with new ideas for the roadmap, new ways to mash up our software with another product. It's a very fast-paced arena, and it is refreshing and encouraging."

For entrepreneurs who may be thinking of producing open source software, Reale has some advice. "Start by using open source. Read as much as you can and educate yourself. This industry is growing and defining itself all the time, so you have to be on top of things. Look at the companies that have done well with the open source model and follow their lead. Put the mechanisms in place that your community will need and start participating in the communities of the software products you already use. You'll learn how to manage a community in part by being part of one."

Facebook Open Sources Parts of Its App Dev Platform

Facebook is releasing as open-source software parts of its application development platform in order to make it easier for programmers to create applications for the social-networking site, the company announced Monday.

Facebook will offer as open source "most" of the code that runs its platform, plus implementations of its most popular methods and tags.

This is another step in Facebook's program for external developers, which it kicked off a little over a year ago when it opened up its platform to them. Since then, about 400,000 developers have created some 24,000 applications for Facebook.

With this move, Facebook is also responding to Google's OpenSocial, an initiative to establish a standard set of common APIs (application programming interfaces) that will let developers create social-networking applications that can run with minor modifications in multiple sites.

OpenSocial is generally considered a challenge to Facebook's platform, because observers believe it could make it easier for social-networking sites to match Facebook's broad catalog of third-party applications.

Among OpenSocial's supporters are Yahoo, AOL and MySpace, Facebook's biggest competitor. In March, Yahoo, Google and MySpace formed a nonprofit foundation to promote the OpenSocial platform as a neutral, community-governed specification.

With the open-source portions of its code, Facebook expects that developers will find it easier to test and tune their applications, and create their own tools, among other things.

The Facebook Open Platform, or fbOpen, as the open-source portion of the platform is called, can be extended so developers can create their own tags and API methods, Facebook said.

The open-source portion of the platform includes the REST API, FBML parser, FQL parser, and FBJS sanitizer and proxy, Facebook said.

Most of the open-source code is being made available via the Common Public Attribution License (CPAL), while the FBML parser is governed by the Mozilla Public License (MPL).

Gartner analyst Ray Valdes calls fbOpen a good move that became necessary for Facebook to make in light of the OpenSocial challenge.

"OpenSocial is gaining enough ground and becoming more real, whereas in the past it was basically a bunch of people talking about the specification. That spec is now being implemented and deployed," Valdes said.

In addition, OpenSocial wasn't significantly open sourced, but that also has evolved to the point where it is now a real part of the open-source portfolio for social-networking applications, he said.

On a broader scale, Facebook is also reacting to the overall trend for greater interoperability, away from "walled garden" sites and embracing the open Web and data portability, Valdes said.

"Facebook needed to respond and they did. Whether what it's doing is sufficient depends on the uptake among developers and partners," Valdes said.

In addition, a possible long-term benefit for Facebook is that this open-source effort may attract more sophisticated developers who are able to create more substantial applications than the typically lightweight widget-like Facebook applications, Valdes said. In particular, Facebook could benefit from developers who create enterprise applications for specific vertical industries, he said.

Ami Vora, senior platform manager at Facebook, said that this move isn't a competitive reaction. Facebook has been tossing around the idea of open sourcing a large subset of its platform since its opening a year ago, Vora said. "This is a release for developers and by developers," Vora said. "The goal is to make developers' lives easier."

Developers have been asking Facebook for more tools and resources. "We thought a good way to do that would be to give them the platform itself so that they can understand how it is put together," Vora said.

By downloading the open-source code, developers get a working subset of the platform on which all Facebook applications run, she said. Now, developers will be able to run and test applications in the Facebook platform but on their own local servers, which should make the process quicker, save on bandwidth and provide more stability in their own controlled environment, she said.

Facebook chose to open source most of the platform code under CPAL because it is based on the MPL, but extends it by stating that contributions and enhancements must be shared even if applications are delivered via a network, like the Web, Vora said.

Thus, Facebook expects that by choosing CPAL, it will foster more sharing of improvements and additions among developers. "We think this is a very appropriate license," she said.

Gartner's Valdes concurred, saying that CPAL should ensure that improvements made to the open-source platform code flow back into the community and aren't kept hostage for competitive purposes by individual players. "It's a very good move on Facebook's part," Valdes said.