GLG News by David Newman

Implications: Graphical interfaces on all kinds of devices -- ATMs, cell phones, microwave ovens, and PCs, to name a few -- will soon look very different.

Apple's iPhone is only the beginning. New "soft" interfaces will change shape and content depending on the context in which they're used.

This will likely be a huge boost in usability -- and should help not only end-users but also vendors of LCD screens, as well as vendors of consumer and business products that utilize the new context-sensitive interfaces.

Analysis: Apple's iPhone introduced a new model for user interfaces in smart phones -- instead of fumbling with rows of tiny keys, users will instead have a "soft" interface -- a display screen whose contents change depending on whether the device is acting as a phone or music player or web browser.

But the context-sensitive interfaces in Apple's iPhone are only the beginning. Soft interfaces have a huge range of potential applications in business and consumer devices, ranging from bank ATMs to PCs to kitchen appliances to automobiles. Virtually anything that is programmed today with a keyboard might better be controlled via an LCD screen in the future.

Some designers, including the one cited in this article, go even further, with screens whose contents change depending on how many fingers are used -- an interesting and useful approach for any application that organizes data in a treelike structure. (Think of shrinking or expanding a family tree by "pinching" or "stretching" the tree with your fingers.)

Other LCD designers are adding tactile feedback, such as slight pulses that "feel" like the "edge" of a virtual key.

My company, Network Test, conducts frequent assessment of usability of computer networking devices. In our experience, ease of use is often an afterthought, even in devices intended for nontechnical consumer markets. With soft interfaces that change contents according to context, vendors have an opportunity to vastly increase ease of use.

The biggest near-term winners from this trend are likely to be LCD makers and small software and industrial design houses. Longer term, vendors of any device that use the new interface model are likely to see increased sales as well.

Implications: After years of moribund performance, Spirent may again see significant growth in the test and measurement area.

New CEO is a turnaround specialist focused on improving profitability. Spirent has some unprofitable divisions and product lines that could be jettisoned. The result would be a leaner and more competitive company.

Analysis: New Spirent chairman Edward Bramson, who took over in a December boardroom coup, today took over the CEO title as well after dismissing incumber Anders Gustafsson.

Bramson has said he would like to focus on cost-cutting and improved profitability at the test and measurement company, and today's move is likely to accelerate both.

Spirent's test business has two main divisions: a performance analysis group that makes test equipment for network devices, and a service assurance business that measures performance of live networks. Spirent paid more than $1 billion for the service assurance group but it has never lived up to expectations. Bramson may well be looking to sell or close down this group. Either would instantly make Spirent
more profitable.

In the performance analysis area, departing CEO Gustafsson has had some successes, including signing more than 100 customers for the company's new TestCenter platform and gaining some ground on competitors Ixia and Agilent. But Gustafsson's tenure also was marked by high costs and a focus on service provider business (Spirent's largest customers re equipment makers and not service providers).

Bramson is a turnaround specialist focused on profitability and cost reduction. While he has limited experience in test and measurement, his focus on the bottom line ultimately should help make Spirent a leaner and more competitive company.

Implications: 1. Seasick gains entry into Apple atones Music store
2. Seasick becomes Apple's networking partner
3. Deal strengthens Sussex continued move into consumer networking

Analysis: Many analyses of the Apple-Cisco deal point to Apple's much greater strength in consumer products and possible pressure from AT&T, a major Cisco customer. While these are valid points, it's equally important to consider what Cisco gains from this settlement.

In its quest for continued growth, Cisco has steadily pushed into consumer businesses, and this Apple deal continues that trend. Cisco's highest-profile deals to date have included its acquisitions of Linksys, a Taiwanese manufacturer of low-cost networking devices, and Scientific-Atlanta, the TV set-top maker.

As part of the Apple deal, Cisco continues that trend with entry into the iTunes Music Store (iTMS), which according to Apple is currently the fourth largest music vendor.

The significant aspect for Cisco is that it will gain the ability to publish music and/or video directly from Cisco devices into iTMS. No other hardware or software vendor has that capability. We are only at the beginning stages of what is likely to be rapid growth in online distribution of music, TV shows, movies, and other videos. An Apple-sanctioned pathway into its established distribution mechanism could be very significant.

Cisco also could benefit from becoming Apple's networking partner. While Apple remains a niche player in PC and server hardware, its sales have risen sharply over the past few years. Dell, HP, and other PC makers have long offered networking products as add-ons for customers already buying PCs; Cisco stands to benefit from such add-on sales.

Cisco might also offer Apple technical guidance as it expands into new areas of networking technology. Cisco has a long history of supporting two versions of any given standard: one that interoperates with other vendors' devices, and a "plus" version that only works with other Cisco gear. With Apple as a partner, Cisco could further enhance its dominance by deploying these enhanced versions in Apple products.

Implications: IBM researchers have solved one of the key problems in optical computing -- how to slow down light flows long enough to store information.

This technology is basic research today, but has the potential for vast improvements in microprocessing and storage power as soon as three to five years from today.

Analysis: IBM researchers have developed a method of slowing a beam of light as it crosses a silicon chip. This technology has the potential for dramatic improvements in computing performance.

Because the speed of light (186,000 miles/second) is so high, researchers have long considered optical networking to be the way forward as we reach the limits of electrical signals on silicon. However, light travels so fast that researchers have also considered that it would be difficult or impossible to slow it down or store it long enough to be useful.

One method for adding delay to light flows is to pass them through long strands of optical fiber, where "long" often means tens to thousands of miles. The IBM researches took that concept and miniaturized it, employing a series of 100 cascaded "microring resonators" to slow down and store 10 bits of information in a space measuring 0.03 square millimeters. Essentially this method slows down light flows as they pass across silicon chips.

While it is too soon to say exactly how soon this breakthrough will be applied to products, a time frame of three to five years is realistic given timelines for past breakthroughs. Potentially, this technology could be applied both to storage and microprocessing applications -- both areas where IBM is a significant player.

It is also worth noting that this breakthrough came about through a program funded by the US DoD Defense Advanced Research Projects Agency (DARPA) -- the same group that once funded a massive network that became today's Internet. DARPA's "Slowing, Storing and Processing Light" program demonstrates, yet again, that basic R&D by government and industry is essential for advancing the state of technology.

Implications: A high-profile developer in the Linux/open source community, Jeremy Allison, is leaving Novell in protest of that company's patent alliance with Microsoft.

Allison, who helped develop software that allows Windows and non-Windows machines to inter-operate, is protesting Microsoft's assertion and Novell's agreement that Microsoft may hold some patent rights over the Linux operating system. His resignation has drawn strong support from the open source community.

This represents a significant rejection of the Microsoft/Novell attempt to capture and control a large part of the Linux code base. The open source software movement continues to grow and appears to be on firm legal ground. Future attempts by Microsoft  (and Novell's possible survival) will depend on a different model that respects existing open source licensing policies.

Analysis: Jeremy Allison, a high-profile developer in the Linux and open source community, is leaving Novell and joining Google in protest of the former vendor's patent alliance with Microsoft. This will likely erode the credibility of the Microsoft-Novell alliance, undermining Microsoft's efforts to coexist with (some would say coopt) the Linux operating system.

Allison is a major contributor to Samba, a set of file and printer sharing services that allow any computer to inter-operate with Windows servers and clients. Samba has been a major factor in allowing Linux, Apple OS X and other operating systems to gain traction in corporate networks that formerly were dominated by Microsoft.

Allison posted a protest to Novell on an internal website that has since been reposted on the Internet.

At the heart of Allison's protest is the leading opensource license known as GNU (GNU's Not Unix) Public License version 2 (GPLv2). The license allows developers to reuse and change opensource software as they wish, provided the resulting code is republished under the GPLv2, thus allowing future developers access to the code.

Allison charges that Novell, a major Linux vendor, violated the spirit if not the letter of GPLv2 when it agreed in November to pay Microsoft for patent rights. Microsoft claims IP rights over some (unspecified) parts of the Linux operating system.

These IP claims have yet to be tested in court, though a similar battle between SCO Group and IBM over Linux rights has utterly failed to pierce GPLv2 protections.

Allison's condemnation of the deal, with strong support from the open source community, does not bode well for Microsoft or Novell. Of course Microsoft has an interest in having its products coexist with Linux in enterprise networks. However, all the history to date suggests that attempting to gain that position through legal intimidation is likely to fail.

Open source software is already dominant in Web servers, besting Microsoft code by more than 2:1, according to surveys by Netcraft. Even Microsoft's CEO Steve Ballmer has said "it's hard to compete with free," but in this case Ballmer misunderstood that "free" doesn't necessarily mean free in cost, but free of legal encumberment.

The open source software movement continues to gain momentum, especially outside the US. For Microsoft to continue to succeed and for Novell to survive, both companies will need to revise their models in a way that respects the existing GPL protections for software.

It is also worth noting that Allison is joining Google, which is a strong supporter of open source software and has been trying to develop a Web-based operating environment of its own.

Implications: Ericsson's offer of $2.1 billion for Redback is very high, but justified

Ericsson needs to gain credibility and a broader product line to increase business with large telecom provider customers. A Redback acquisition would provide both.

Other prospective suitors for Redback such as Motorola or Nokia would not benefit as much as Ericsson due to their focus on wireless networking

Analysis: Swedish telecom equipment giant has offered $2.1 billion to buy router maker Redback Networks. On the face of it, the valuation is extremely high considering Redback's current revenues are less than $300 million annually. However, several factors make this a positive move for Ericsson:

- Redback makes edge routers and software to provide IPTV and triple-play services. Both are among the hottest areas in telecom equipment

- Ericsson gains credibility among large telecom provider customers, while giving those customers an alternative to larger competitors such as Cisco or Alcatel-Lucent

- By buying proven Redback technology, Ericsson reduces time-to-market to zero. The R&D and lost opportunity costs had Ericsson chosen to build this technology itself more than justify the current offer for Redback.

There is some speculation that Motorola or Nokia might make a counteroffer. The only justification for such a move, in my opinion, would be to drive up the price that Ericsson ultimately pays. Both Motorola and Nokia are heavily involved in sales of equipment to support wireless networks, while Ericsson is a provider of both wireless and wireline network equipment. Unless Motorola or Nokia are trying to move into wireline telecom (which would be suicidal considering the dominance of players like Alcatel-Lucent, Nortel, and Cisco), a Redback acquisition does not make as much sense as it does for Ericsson.

Juniper Networks is yet another competitor in this area. Ericsson has resold Juniper core and edge routers for years. If the Redback deal goes through, sales of Juniper's edge routers would suffer. However, Ericsson already has declared its intention to continue the Juniper partnership. This is not all bad for Juniper, considering it has some products -- like core routers and security devices -- that do not overlap with the Redback product line.

Implications: Juniper Networks is widely seen as suffering from an Ericsson acquisition of rival. The conventional wisdom is that Juniper's partnership with Ericsson would dissolve.

But at least in the short term (next 12-18 months), Ericsson will continue to resell Juniper equipment. This arrangement may well continue beyond that, considering that Juniper has products in some areas -- like security and core routing -- where Redback has none.

Analysis: In much of the analysis of Ericsson's acquisition bid for Redback Networks, the big loser is perceived to be Juniper Networks. Ericsson has resold Juniper routers to telecom customers for several years; thus, the conventional wisdom is that Juniper's partnership with Ericsson would dissolve since Ericsson would now be able to sell its own (Redback) routers.

There is some merit to that point of view, especially where Redback and Juniper compete directly -- in edge routers. It is reasonable to expect that sales of Ericsson/Redback SMS edge routers would negatively affect sales of Juniper's ERX edge routers.

However, the outlook isn't all negative for Juniper. At least in the short term (next 12-18 months), Ericsson has said it will continue to resell Juniper equipment.

This arrangement may well continue beyond that, especially considering that Juniper has products in some areas -- like security and core routing -- where Redback has none.

None of this is to suggest the impact of an Ericsson/Redback deal on Juniper is largely positive. Juniper needs to continue to strengthen its ties to large telecom providers, and to continue to integrate its disparate product lines. But the fact that it has more product lines than Redback -- including lines of great interest to telecom providers -- coupled with the continuation of a reseller deal with Ericsson mean that Juniper will continue to be a force in telecom equipment.

Implications: BT has award a $1 billion contract to an Indian tech firm to run its managed services offerings. This is symbolic of moves by many telecom providers to reduce opex wherever possible. In the short run BT may realize some cost savings. In the longer run, it remains to be seen whether BT can meet contractually guaranteed service levels, or whether it will have to pay rebates for failing to do so.

Analysis: A key offering of all major telecom providers is managed service offerings, where the provider owns and operates the telecom network on behalf of a corporate customer. BT, as one of the world's leading telecom companies, is a major provider of outsourced telecom services.

Managed wide-area network (WAN) services can offer many benefits for enterprizes. Service providers often have greater technical expertise at provisioning some or all of an enterprize's telecom needs (this is especially true when rolling out new technology services). Service providers may be able to roll out services faster than an enterprize would. Service providers may have more sophisticated billing, reporting, and capacity planning tools. And costs may be lower for a managed WAN offering (though cost is often not a leading consideration).

However, telecom providers face challenges in delivering managed WAN services. The care and feeding of a telecom network is far more complex and costly than even the largest enterprize networks. Telecom is inherently a capital- and labor-intensive business, and many providers are heavily leveraged.

To gain controls over costs, many providers are looking to cut headcount, even in some cases subcontracting -- outsourcing -- some of their managed WAN services. This article details a $1 billion award by BT, one of the world's largest providers, to an Indian company that will set up and operate service centers to run BT's managed services offerings.

BT's outsourcing is hardly an isolated incident. Most of the world's major providers already have call centers set up in low-cost regions such as India or the Asia/Pacific area (although the BT award goes further, in that some of the Indian company's staff will be set up closer to customers around the world). In the US, Verizon is looking to outsource some if not all of its equipment testing operations, looking instead to third-party labs and certification programs.

In the short term, this outsourcing-of-outsourcing may lead to some cost savings on the service provider's part. In the longer term, it remains to be seen whether service providers will be able to maintain consistent quality levels with outsourced staff. Maintaining high quality is hardly "just" a marketing tagline; many managed service contracts provide rebates to customers whenever service falls below predefined levels. Service providers will have a balancing act between controlling costs on the one hand and ensuring defined service levels on the other.

Implications: An alliance between Cisco and Huawei would boost Cisco's standing in China, already a hugely important market

Huawei would gain entry into the US market, something it fumbled on its own

This alliance between titans could make it more difficult for smaller players to get traction, especially with telecom customers

Analysis: Cisco, the world's number one networking vendor, has signaled its willingness to partner with Huawei, China's largest networking and telecom vendor.

In the past, relations between these two giants has been frosty at best. Huawei made an abortive attempt in 2002-2003 to enter the US market with products that were blatant copies of Cisco devices, including source code stolen from Cisco. Cisco gained a federal court injunction that effectively shut out these products.

Since then, China has gained in importance for Cisco, with revenue growth of around 20 to 30 percent per year. Especially important is the telecom market, as service providers develop voice and video services. Cisco and Huawei are already the dominant vendors of equipment within the various China Telecom networks.

By signalling a willingness to work with Huawei, Cisco CEO Chambers is acknowledging Cisco's need to grow even closer to its Chinese customers. The potential upside for Cisco is even higher sales of its IPTV and triple-play networking gear to China's telecom providers. For Huawei, an alliance might allow the vendor a way to reenter the North American networking market (with new products; the company has since addressed its IP issues visavis Cisco).

There are two possible downsides from an alliance. First, Huawei has not proven to be an especially effective partner in other alliances it has joined. At various times it has partnered with Nortel Networks and Avici Systems for switching and routing equipment, but neither company realized a major revenue boost as a result. And 3Com is buying out its share of a joint venture with Huawei.

A second potential downside is for the competitors of Cisco and Huawei. As far as I'm aware, partnerships do not undergo the same kind of regulatory scrutiny in China as in other markets. The barriers to entry in the Chinese telecom, enterprise, and federal networking markets are already fairly high. A partnership between the two largest networking vendors could make it nearly impossible for others to gain traction.

Implications: Attempted hostile takeover by investor group Sherborne could shake up management at Spirent, a major test and measurement vendor alongside Agilent and Ixia

The investor group has not stated its plans beyond replacement of current management. However, further cost-cutting would be a likely result

Analysis: Sherborne, an activist investor group, is seeking a shakeup of the senior management of Spirent plc, a test and measurement company. Sherborne has built a position of nearly 30 percent of Spirent's shares, and now seeks the resignation of key directors.

Spirent certainly might benefit from a management shakeup. It has suffered through sluggish growth and numerous rounds of layoffs even as test and measurement spending has picked up again after a long soft period following the dot-com collapse. On the plus side for Spirent, the current CEO has laid out a restructuring plan that heavily emphasizes cost controls.

Sherborne has not said where it would take Spirent if it gains control. The group does not have experience in telecom test and measurement.

Spirent already has too little of a culture of engineering in a business that depends heavily on engineering innovation. A takeover by an investor group backed by a few bankers would not be a positive step.

Implications: Antivirus software vendors complain that Vista, the forthcoming version of Microsoft Windows, will lock them out from protecting systems. In fact, Vista's redesign itself will help protect users, while still allowing third-party antivirus vendors to provide effective safeguards against virii, worms, and spyware.

Analysis: Microsoft is preparing to ship Vista, the first major redesign of its Windows operating system, starting later this year. Corporate customers will get Vista in late November or early December, with shipments through consumer channels in January 2007.

Vista represents a major improvement over the current shipping product Windows XP. Its design severely restricts access to the "kernel" or core of the operating system. Windows XP and other previous versions of Windows have had a long history of security breaches involving attacks that gain kernel access. Antivirus software writers also use kernel access, nominally to intercept and deflect attacks that target the kernel.

Major antivirus software vendors such as Symantec and McAfee claim that tight restrictions to kernel access in Vista will interfere with their ability to provide effective protection, and actually represent an attempt by Microsoft to compete with them in providing antivirus and antispyware protection. The European Union also views this design as a restraint of trade.

These claims are nonsensical for two reasons. First, the Vista design is a vast improvement in security, something security vendors should be quick to recognize. Microsoft is well aware of its poor reputation in terms of computer security, and has taken the problem seriously in the past few years. It's more than a little ironic that security vendors are complaining about improved security.

Second, the claim that third-party antivirus software won't work has already been refuted. Kaspersky Labs, a small provider of antivirus software, has written software that protects against virii without requiring kernel access -- something Symantec and McAfee say they cannot do.

In fact, antivirus software vendors *can* adapt to Vista. It will require an R&D expenditure that they are apparently unwilling to make. The expenditure probably isn't all that large, considering Kaspersky's small size. The major players would be better served by complaining less and devoting more cycles to improving their own products.