Install What You Need with Windows Server 2008
Windows Server 2008 is Microsoft's most full-featured server operating system yet, so it's ironic that one of its most exciting new features is an install option that cuts out most of the other features. Paul Rubens explores why a Server Core installation makes a great deal of sense in many instances.
»
Identify Hardware and Software That Meet Microsoft Standards
The "Certified for Windows. Server 2008" logo identifies hardware and software solutions that meet Microsoft standards for compatibility and best practices with the Windows Server 2008 operating system.
»
Download the Windows Server 2008 Trial
With Windows Server 2008 you can develop, deliver, and manage rich user experiences and applications, provide a secure network infrastructure, and increase technological efficiency and value within your organization.
»
We'll admit to a certain soft spot for end-of-the-year reviews. We can put on a thoughtful face and talk about life, death, rebirth, and renewal. We can get the pain of admitting mistakes from the previous year's review out of the way before blithely sailing into a new year of prognostication and considered opinion. And we can wrap the whole year up in a thematic bow. So with no further adieu ...
Sun: Not Dead, Maybe Even Renewing Itself
We ended last year with a bit of can't-miss prognostication:
"... if Sun pulls together and recovers from its slump, 2004 will go down as the year proprietary Unix held the line against Linux."
Or not.
The point we were trying to make was that the momentum in enterprise Unix belongs to Linux, and that its meaningful opposition among the contenders was down to Sun. Sun's announcement of an open source Solaris was still six months in the future.
So did Sun pull things together in 2004 and present a credible answer to Linux?
We think so, but it did it by building a sensible foundation, not taking over the world. Last year was about the company refocusing on where the market that had disappeared out from under it had resurfaced; this year was about bringing product to that market and drawing attention to a revitalized Solaris, which Sun expects to act as the mortar for its integrated systems.
It was the "drawing attention" part that made us cringe, with Sun's President and COO Jonathan Schwartz using his blog and any passing microphone to assail Linux top dog Red Hat, and the sensibilities of Linux enthusiasts in general:
"There is a fork in the Linux world: Red Hat and the others," averred Schwartz, who eventually managed to pick a cat fight with Red Hat Vice President Michael Tiemann. More anti-Sun snarking from open source enthusiasts over Sun's peace treaty with Microsoft provoked Mr. Schwartz to denounce "bizarro numbskull anti-Sun conspiracy theorists," and most recently he's been hard at work shoring up Sun's case against HP.
When he wasn't pummeling the competition, Mr. Schwartz was out courting, spending some time in New York City in September wooing Wall Street. That encounter, which involved earnest promises to listen better and behave in a more humble manner, was probably the company's PR nadir, fitting in with Sun's broader pattern of emoting and intending so it can stay above the fold when it doesn't have any product to announce.
Enterprise Unix had a sense of dynamism and vigor that's been missing for a few years. Linux has been interesting enough, but its ferment has left the rest of the Unix world in the dust.
Finally, Sun promised to make Solaris 10 available under an open source license, effectively trashing our prediction that it would be the last proprietary Unix holdout. In the immediate aftermath of the announcement, we weren't particularly impressed because we weren't sure what Sun thought the move meant. Over time, it's become more apparent: Very little from a technical standpoint, since Sun will keep firm control over the OS with a license designed to elbow Linux from the table. It will gain a bit more from a marketing standpoint, since "going open source" erodes one of Red Hat's key differentiators.
In the end, although Sun's two most positive actions this year were its push to stress the technical merits of Solaris (with an attendant promise to focus on its neglected Solaris x86), and its partnership with AMD to x86 hardware to market. The former means Unix on the whole will continue to move forward, and the latter represents good timing on Sun's part, as Intel's Itanium founders in the face of AMD's Opteron assault.
SCO: Also Not Dead, Not Exactly Prospering
This time last year we were already tired of the SCO/IBM case and its attendant melodrama:
"... our less measured and less patient perspective is one of exasperation with a company relying on vague threats and misrepresentations of open source software to keep its name above the fold," we sniffed.
In early March, SCO at last filed the Linux-related lawsuits it had promised, revealing it was going after Autozone and DaimlerChrysler. At least, we thought, the company wasn't subsisting on threat alone, even if the DaimlerChrysler suit ended up in the trash can because, hey, DaimlerChrysler hadn't used SCO product in so long it had forgotten all about it.
In April, exasperated with SCO's failure to sell many Linux licenses, investor Bay Star called back $20 million in investments, intimating that SCO needed to buckle down, stop pretending it was selling anything besides protection from litigation, and back off the combative rhetoric. Two months later, we dared to hope SCO had learned something when it made a few product announcements and briefly struck a milder tone.
In May, Linus Torvalds introduced the Developer's Certificate of Origin, a document designed to provide a paper trail for code finding its way into Linux. As the Open Source Development Labs put it at the time, the DCO "goes a long way toward eliminating doubt surrounding the origin of Linux code." In interviews since, Torvalds has said the certificate is enhancing kernel development.
In August, we noted a few other reactions to SCO's campaign, including Red Hat's adoption of software designed to track licenses and copyright infringement risks, and the formation of a company aimed at protecting companies that use open source software from litigation. The SCO suit left an impression on the open source world, even if SCO itself remains one of the most reviled companies in Linuxdom.
Just this week, SCO announced quarterly results: Its losses were quadruple those of the same quarter last year ($6.5 million vs. $1.6 million), and revenue from its Linux licensing program was a paltry $120,000. Along with the poor earnings came news that SCO's majority stakeholder, the Canopy Group, had sent its CEO (and SCO's chairman) packing in a boardroom coup. Not a pleasant way to end the year, as rumors abound that the company faces acquisition.
Expect more from SCO next year: SCO vs. IBM isn't set to go to a jury before November 2005.
Novell: Not Dead Either, Surprisingly Life-Like
We've long had a soft spot for NetWare, but our affection for Novell's operating system and management tools did nothing to protect it from the depredations of Microsoft and Linux. Confronted with Microsoft, which trundles elephantine through any market it pleases, and Linux, which is more of an armed caravan than a marauding juggernaut, Novell opted to join the caravan.
It began the process when it acquired Linux software company Ximian in late 2003. It finished lining up its ducks when it finalized its acquisition of SUSE, a Linux distributor that bore the tagline "big in Europe" like a weight around its Red Hat-dominated neck for years in the United States.
We've seen enough Linux conversion stories that we weren't positive Novell would thrive in its new market. Red Hat has an impressive lead in the United States, and the Linux user base can generate savage and devastatingly negative buzz when its sacred oxes are gored (or even brushed against) by the wrong company.
But Novell moved forward with aplomb, using its annual BrainShare conference to announce NetWare as an operating system was on its way to end-of-life ahead of schedule, the company was releasing tools it acquired with SUSE under the GPL as a sort of dowry to the Linux community, and it planned to merge the two most popular Linux desktop packages into an "integrated" whole.
Later in the year, the company formalized its accelerated migration to Linux and finessed the enthusiast market by both keeping SUSE intact as a brand instead of slapping "Novell Linux" on the product. And it kept GNOME and KDE as separate desktop packages (averting a potential desktop Holy War).
However, this is the first year in a few we found ourselves thinking there's a future for Unix beyond Linux. 2004 was about putting the pieces in place; in 2005 those pieces will need to go into motion.
And the year ended on a sweet note: Novell inked a deal with Dell, which was piqued over Red Hat's high prices . This brings Novell's SUSE Enterprise Server 9 to Dell boxes. SUSE also obtained certification to run Oracle products, indicating that someone besides Red Hat has a shot with the ISVs.
Novell is well on its way to taking a chunk of Red Hat's market and re-establishing itself among enterprise customers. The next year ought to be an interesting one for the company.
Putting On a Bow
This brings us to what the year's three biggest story arcs have in common: Enterprise Unix had a sense of dynamism and vigor that's been missing for a few years. Linux has been interesting enough, but its ferment has left the rest of the Unix world in the dust.
This year, though, with Sun wrapping up the release of an improved Solaris 10, bringing a renewed focus to bear on Linux' core market, and amping up its attacks on Red Hat, along with Novell re-emerging not only as the newest Linux player but also as the most credible competitor Red Hat has faced since Caldera, there's been something to watch. Linux is still the dominant story by dint of its continuing growth within the Unix market, and Sun (despite being the most vigorous Unix holdout) spent the year figuring out how to react to Linux.
However, this is the first year in a few we found ourselves thinking there's a future for Unix beyond Linux. 2004 was about putting the pieces in place; in 2005 those pieces will need to go into motion.
IE 7
Firefox 2.0