One of the most striking details of the merger is the extent to which Compaq, supposedly the junior partner in the deal, seems to be coming out on top as the deal is executed. Headed for the boneyard is HP's NetServer line. The "New HP" will adopt Compaq's Proliant servers, running the client's choice of Windows, Linux, or bare bones. HP will also dump its Journada line of PDAs in favor of the iPAQ. Capellas said aloud what many observers thought when he announced, in advance of the official HP announcement, "You are what you eat. The New HP in many ways will resemble the old Compaq."

Not only is the hardware mostly Compaq. So is much of the mid-level management team. For example, HP's channel chief Kevin Gilroy has been moved over to concentrate on reselling consumer HP boxes at Kmart and Staples. Compaq's Dan Vertrees will run the high-profile job of handling the enterprise integration and VAR markets. Outside the United States, Compaq executives are heading up many of the country manager positions. In India, for example, every division except printing will be run by an ex-Compaq exec.

Fiorina may well prefer ex-Compaq managers to her own managers, many of whom have been downright hostile to their CEO. During the course of the merger, the tech press was flooded by letters from retirees and employees suggesting that Fiorina is little more than an East Coast "suit" who doesn't appreciate HP's "engineering leadership." Polls conducted by merger opponent Walter Hewlett found that HP employees were opposed to the deal. Fiorina's stint at Lucent Technologies, where she positioned a tired communications company saddled with tons of old-technology big iron into one of the biggest telecom IPOs of the dot-com boom, has not helped her reputation with the "old bulls" in the HP engineering and R&D staff. It has not done much for her reputation among many institutional investors, who bought lots of bad paper, and who have long memories.

That negative sentiment nearly sank the merger. Once investors began lining up against the merger, Fiorina only won the deal by declaring war against the Hewlett family. Most observers believe the deal went through because the company had stalled so long that many institutional investors feared a complete implosion if the merger failed. All the same, the deal nearly didn't get done. Fiorina only got a majority after one German investment bank changed its vote. The final tally was 838,401,376 shares in favor, 793,094,105 shares against. In an era when most mergers are typically approved by 9:1 ratios, this is no vote of confidence.

While Fiorina may have gotten her merger, the investment community is still throwing rocks at both the new company and its leader. Several investment analysts greeted news of the merger by re-iterating their "neutral" rating, which, in consultant-speak means, "run for the hills." Bear Stearns' Andrew Neff and Naveen Bobba expressed their doubts about Fiorina's ability to execute the merger plan, scathingly suggesting that she needed to move from PowerPoint slides to reality. Some of HP's "friends" were even more negative. The one analyst who gave the company a "buy" rating asked why the company chose HPQ as its new corporate ticker symbol when both "YUK" and "PU" were available.

Credit market types are even less impressed. George King at Alliance Capital Management said that the only debate among his peers was "just how negative it will be." Moody's Investor Service recently put the company's unsecured debt on review for a possible downgrade. Fitch Ratings Service assigned an F1 rating to the newly merged company's commercial paper program and gave HP a rating outlook of negative. While unpromising, this actually represents a slight upgrade from an even more negative opinion during the last few months.

There is a good reason for much of this negative sentiment. According to Dataquest, HP suffered a 35% drop in demand for its workstations in the last year. By comparison, Compaq only lost 1% of its business. In any other year that would not be much to brag about, but this was a year when most players not named Dell or IBM suffered double-digit losses. The combined HP and Compaq dominate PC sales in stores and through resellers, but their market share has been falling off -- rapidly in some cases. HP's share of the PC market has dropped 6% in just the last month. Compaq's decline has been less abrupt, although its store sales have fallen by more than a third over the last year and it has lost share among commercial customers as well.

The situation places Capellas in an interesting position, not much different from the one that emerged when he took over at Compaq originally. Capellas, who built a reputation as a skilled corporate infighter, has been happily driving up expectations in a way that further exposes his new boss. Most analysts have suggested that the company will take anywhere from 18 to 36 months to recover, and that only deep cost cutting will justify the merger. However, Capellas stated that the company had only a "six month window" in which to justify the merger, and that "the only reason for making a big strategic move like this is to gain leadership." He add that the market should judge HP (and presumably, his boss) by the degree to which the merger "stimulates innovation" by HP's unhappy engineers, not by how much it saves in costs.

Capellas has also made a number of other interesting statements deviating from the official HP catechism. While the company's official position is that it will continue to support both its PC lines and storage solutions, Capellas has pointed out the merged company will start with an unwieldy line card of nearly 10,000 products. The company announced Tuesday that it will support its HP and Alpha Unix servers, and will rely on a triad of Linux, Windows, and proprietary platforms. However, the week before, Capellas told reporters that products like HP's HP-UX were all but dead, declaring, "You're going to see Linux and Windows absolutely eviscerate the midrange Unix market." He also managed to dismiss HP's investment in proprietary components, stating that HP would focus on "Microsoft Windows, Linux and Intel chips."

While contradicting the boss is generally a poor career move, Capellas' situation is somewhat different. The investment community was already spooked by the large sell-off at Sun Microsystems when Ed Zander "retired." Several of the institutional investors who finally did vote for the merger have warned that they would look "very negatively" at any exodus of "senior managers" resulting from the merger. If the merger is to succeed, it will have to do so by growing the computer side of the business, and the most prominent computer guy in the combined entity is Capellas.

A lot that future growth may depend on how just how well the company manages Open Source. As Grant Gross noted in a column last week, HP's solutions team recently won a number of high-profile deals with workstations running Linux. While the company's efforts to win design and imaging deals with companies such as the DreamWorks movie studio captured most of the headlines, the Linux sale at the U.S. Department of Energy was more significant. One of the few places HP has maintained market dominance is in engineering environments, especially computer-assisted instrumentation. Linux is a natural winner in this environment.

If HP is to rebuild its workstation and computer integrated solutions business, it will need to do so using superior, Compaq-based hardware, and Linux code. Now that Capellas has emerged as the Open Source advocate with the highest pay grade at HP, he will certainly want to make sure HP continues to feature Linux solutions and that those deals are high-profile money makers for the company. Success at HP will be important for the future of Open Source, and for the future of Capellas.