But Wait, There's More

Red Hat Profits, But Misses Revenue Targets By A Bit in Q3

Commercial Linux distributor Red Hat announced its financial results for its fiscal 2005 third quarter just before Christmas, and the company reported that profits were up even as it missed Wall Street's projections for sales. Red Hat's stock got a 10 percent haircut immediately, and not just from the missed projections in the third quarter, but because the company said that its fiscal fourth quarter would come in below expectations, too.

In the third fiscal quarter ended November 30, Red Hat posted enterprise Linux subscription sales of $38.1 million, more than double sales in the year-ago quarter, and total Linux subscriptions of $39.2 million, just under double sales from last year. (Red Hat has significantly reduced its presence in the retail and embedded Linux markets.) For the quarter, services revenues (again, dominated by the enterprise market) accounted for $11.7 million in sales, up only a smidgen from the third quarter of fiscal 2004. Overall sales were up 55 percent to $50.9 million, and net income was $10.8 million, up from $4.3 million a year ago. Red Hat said it sold 119,000 enterprise Linux subscriptions in the quarter and 13,000 subscriptions for high-performance computing clusters and hosting services, the latter two areas which Red Hat lumps together as distinct from enterprise customers. Red Hat's HPC and hosting shipments have been dropping like a stone for the past year. (Those subscription figures include both new and renewed subscriptions.).

This all sounds reasonable in terms of raw numbers, but Wall Street was expecting the company to boost sales by $1 million more than it did, and then Red Hat said that sales in the fourth quarter would be between $55.5 million and $56.5 million with net earnings of 6 to 7 cents a share; the Wall Street consensus was $58 million in sales with 7 cents per share of profit. Hence, the share price declines right after Red Hat made these announcements.

Merrill Lynch Says Sun Should Buy Novell or Red Hat

Unix workstation and server maker Sun Microsystems has had an on-again, off-again relationship with Linux since that platform went mainstream about five years ago. Sun bought a Linux server appliance player, Cobalt, for $2 billion at the peak of the dotcom bubble and then just tossed away that investment after a few years of pottering around. With the advent of the new Solaris 10 Unix and Sun's increasing emphasis on Opteron-based servers, Sun is "doubling down," as CEO Scott McNealy likes to say, on Unix. But this, says brokerage house Merrill Lynch, is not going to be a sufficient strategy to counter Linux. In fact, analyst Steve Milunovich says that Sun should just bite the bullet and buy either Novell or Red Hat and just admit, more or less, that Unix is dead.

This is not going to happen.

If Sun wanted to buy a real Linux distributor, it would have done it a long time ago. While Sun can arguably afford to buy Red Hat or Novell, which each have a market capitalization of $2.5 billion as we go to press--and in fact, with a market cap of its own in the range of $18 billion and close to $7 billion in cash, Sun could acquire both companies--Sun is not about to spend that kind of dough on what its current management sees as an inferior style of Linux. Sun believes in Solaris 10, which runs as well or better as Linux on X86 iron (provided you can get drivers, of course) and which has a much larger installed base of enterprise applications. Sun wants to use its cash hoard not to buy into Linux, but to sell Solaris against Linux. It takes a lot less cash for Sun to do the latter than the former, and the profits won't be all that different in the long run if Solaris 10 takes off as many expect among the Unix faithful.

Great Wall to Make xSeries, OpenPower Servers for IBM in China

Only a few hours after IBM announced that it was selling off its entire PC business to Lenovo Group, a Chinese maker of PCs, Big Blue announced that it had gone to rival China Great Wall Computer Shenzhen of Shanghai and given that company a sweetheart deal to make IBM's xSeries and OpenPower servers in China for distribution in the Asia/Pacific region. Great Wall had a partnership with IBM already to make xSeries servers and also had a separate partnership with Big Blue whereby it made IBM-branded PCs for Asian market. Under a separate transaction, IBM bought out Great Wall's stake in the PC operation, which presumably will be now handled by Lenovo. While the great Wall and Lenovo deals were not linked, they are clearly related.

The Great Wall deal marks the first time that another IT supplier has been given the rights to manufacture IBM's RISC-based Power servers. While Bull has for a decade licensed and rebranded IBM's entire RS/6000, pSeries, and p5 Unix servers, the company has never been allowed to manufacture them. Great Wall is apparently only being allowed to manufacture the Linux-based OpenPower variants of the Power5 "Squadron" servers.

Symantec Acquires Veritas for $13.5 Billion

That giant sucking sound you hear is not the Chinese economy eating up all of the natural resources of the planet, but rather the very beginnings of what will likely be a wave of consolidation in the software business. Just before the holiday break, security software specialist Symantec announced a $13.5 billion merger with storage software specialist Veritas. Both companies have been on buying sprees in recent years, and are banking that by creating a software powerhouse, they can get better market penetration, expand into new markets quicker, and wring more profits from their wares.

While any synergies from these very different companies may not be apparent, Symantec and Veritas make products that have an intimate knowledge of operating systems, file systems, and raw data. They are both as platform agnostic as possible--Symantec has concentrated on Windows and dabbles in Linux, while Veritas has concentrated on Unix, has bought its way into Windows (with backup products), and has dabbled in Linux. There are some synergies, and after the merger, there will definitely be some cross-selling capabilities, particularly among enterprise customers, who will represent 75 percent of the company's sales.

Symantec says that the combined company will have sales of about $5 billion in fiscal 2006 and will be the fourth largest software company in the world (behind Microsoft, IBM, and SAP. The company will have 13,000 employees, with about 6,000 in sales and support and 3,500 in development.

Software Market Consolidation Looks to Be a 'Predator's Ball'

Now that Oracle is succeeding in its acquisition of PeopleSoft (and the former J.D. Edwards business too), pundits are declaring that the IT industry is entering a period of consolidation, which is actually something that's been happening for a while. One of the big players you'll likely see more of as this trend accelerates is Oracle. In fact, before settling on a hostile takeover bid for rival PeopleSoft, Oracle's board had seriously considered several other acquisitions, including middleware developer BEA Systems, CRM software developer Siebel Systems, business intelligence software developer Business Objects, and ERP software developer Lawson Software. Oracle may yet succeed in acquiring these vendors. "It's a period of consolidation where the strong are going to get stronger, and the weak are going to get weaker," says Larry Ellison, who, as Oracle's chief executive and one of the world's richest men, is one of the strong ones.

Another strong player to keep your eye on is IBM, which in December announced its latest acquisition, French procurement services company KeyMRO. According to the Wall Street Journal, court records from the PeopleSoft-Oracle fight show that IBM had been targeting more than two dozen potential acquisitions in the business software market, on a list that bore the name "Predator's Ball." Here at IT Jungle, we understand that surviving, adapting, and thriving sometimes requires a hearty meal of red meat, but some of us are not looking forward to a return of 1980s Wall Street values.

LTO Roadmap Extended Two More Generations, but Transfer Speed Increases Scaled Back

The Linear Tape-Open (LTO) format is alive and well, as the LTO Program late last year tacked two new generations onto the end of its LTO Ultrium roadmap. The roadmap calls for the doubling of tape capacity, and a near doubling of data transfer speeds, every generation. In the first three LTO generations, tape capacity and transfer speeds doubled every year, and the LTO-3 tape drives, which are just starting to come to market, offer 400 GB native capacity (800 GB at 2:1 compression) and 80 MBps data transfer speeds (160 MBps compressed). By the sixth generation, LTO will offer 3.2 TB of native capacity and 270 GBps native data transfer speeds. The group says that compression rates will stay at 2:1. The LTO Program also scaled back its anticipated native data transfer speeds for the future of its LTO drives. Instead of an anticipated 160 GBps native data transfer speed, the group said LTO-4 drives will instead offer a 120 GBps native data transfer speed, which represents a 50 percent improvement, as opposed to the 100 percent improvement LTO has made in each generation up to LTO 3. The new LTO roadmap shows that top data transfer speeds will increase by 50 percent for each successive generation.

Congress Allows Another 20,000 H-1B Visas

Congress has approved 20,000 new visas for skilled foreign workers. The cap had been set at 65,000 after years of running at an even higher rate, which had disastrous effects on the IT job market in the United States. IT industry executives had been pushing for 50,000 new visas, while IT workers unions (there are a few) had been lobbying Congress to keep the number of H-1Bs capped at 65,000.

As it turns out, even though the H-1B visas for 2005 were not set until late October, when the omnibus bill passed, vendors were allowed to have applicants fill out visas ahead of time, starting in the spring of 2004. So by the time the bill was passed, the cap had more or less already been filled. And it won't take long for the IT industry to fill the 20,000 new positions, either, even though candidates are required to have graduated from an American university with a master's or a Ph.D.

The question now is what will happen to the unemployment rate among programmers and project managers now that the cap has been lifted. Unemployment rates had been dropping through 2004, and there is a good chance that they will now rise again because of the increase in H-1B visas. The IT industry is giving the same old song and dance about how American universities are not pumping out enough qualified computer scientists, but the truth is that H-1B employees are just "onshored" professionals who are essentially indentured servants for the companies that sponsor them, and can be deported if they lose their jobs. Small wonder they are more qualified and yet will accept lower pay. And whether or not that is legal or ethical is something you don't read about much in the papers.