Migration for the masses
Migration for the masses
While it has long been a mainstay of the server environment, Linux has not had quite the same impact on the desktop environment, at least not until now. Con Zymaris, of the Open Source Industry Australia (OSIA), suggests that with major retailers both here and in the U.S. looking to sell PCs bundled with Linux, and large corporates eyeing it up as a low cost, flexible alternative to Microsoft, Linux' time on the desktop has arrived.
Linux has, for a long time, had the aura of a server platform. When it was initially developed by its original creator however, it was intended to fill his requirements as a desktop operating system. Linus Torvalds needed an implementation of the powerful UNIX system for his 386 PC, but all the versions available were too expensive for him to afford. So, as technologists often do, to the general advantage of the rest of us in the computer industry, he rolled up his sleeves and built his own.
Now, due to the immense effort needed to make an enterprise operating system, he wouldn't have gotten very far alone. This is where he did one singularly smart thing. He released the code under a licence which meant that others could use it and simultaneously help him build it. It was his take on 'If you build it, they will come... and help you'. What had once been a tinker-toy for one, morphed into a viable platform for the multitudes.
Fast forward a few years and we find ourselves in the middle of the Internet boom. Startups needed powerful Internet servers. And they needed them by the millions. Linux (and its thematic cousin FreeBSD) was on hand to fill that need. Sun may have claimed to put the Dot in the Dot Com, but Linux et al provided the low cost building blocks necessary to power the Internet, in turn diverting the platform from its original focus on the desktop into becoming the most popular server on the Net. Linux has entrenched its position in this arena, hitting the #2 server spot and observing 57 percent compound annual growth, according to IDC. Over the next few years it will threaten Microsoft for total dominance of the server space.
By the late 90s, a number of its developers started to pay attention to Linux as a desktop platform as well. Linux had already acquired a reputation for power, stability and security. What it lacked was user focus, polish and wow-factor. With ever-growing effort over the next few years, Linux accrued most of these too. It even picked up the applications that enterprises need to conduct their daily business: groupware, office suites, ERP and CRM.
The reality is that here and now, Linux can be used by most organisations as a totally viable desktop platform. It is at least as user-focused, polished and glitzy as Windows 98, which was used by most as the preferred enterprise desktop platform for many years. Additionally, Linux is at least as powerful, robust and secure as Windows 2000. Many would say far more so. If those platforms are good enough to run enterprise desktops, so is Linux.
"An increasing number of peripheral (printer, video card, RAID, wireless card etc.) manufacturers publish Linux-drivers. Large players like Dell and HP have started selling Linux-based desktops and laptops. In the US, volume resellers like Walmart are pushing sub-$200 Linux PCs into the consumer space. Locally, there have been rumours of similar moves from Harvey Norman and Dick Smith Electronics.
Why then hasn't Linux swept through the desktop space as it did the server market?The short answer is that most enterprises already have established desktop Standard Operating Environments, mostly built around Microsoft Windows and Office. Moving away from these is a lengthy and expensive procedure, fraught with complications and political risk. Migrating from older to newer versions of Microsoft products is also a major, often traumatic leap, but one without the potential for career-limiting damage.
Most enterprises face three roadblocks when considering migration to Linux and OpenOffice. The first is Linux' inability to run most of the important Windows-based applications that enterprises use. The second is the lack of conversion tools which make possible easy and complete transfer of users' address books, email folders, documents and Internet bookmarks. Lastly, there is the issue of users' fear and loathing of switching desktops.
In order to facilitate greater desktop migration uptake, Linux vendors must take into account these concerns raised by enterprises. They must also articulate the benefits they see in TCO reductions, malware security improvements and return on expense outlay. Enterprises will not make the effort to switch to something new unless there is a substantially compelling reason to do so. Nowadays this is mostly centred on reducing costs. If the right story can be told to enterprises, they are far more likely to make the leap.
The story is indeed almost there. Linux desktops, running the latest user interfaces, can comfortably match users' expectations for look and feel. Low cost tools like Crossover Office can run most of the major Windows-based enterprise apps (Microsoft Office, Lotus Notes, Microsoft Project and Visio, Macromedia Dreamweaver MX, Flash MX, Adobe Photoshop, QuickTime and Shockwave). Software vendor Alacos has also released its Linux Migration Agent which will convert users' Outlook folders and address books, Office documents and a dozen other formats from Windows to Linux.
Furthermore, there are a growing number of very large desktop migrations which are being watched very carefully by the industry worldwide. This includes 14,000 desktops in Munich, 7,500 at Allied Irish Banks and optioning the platform on 7,500 desktops in Vienna. If enterprises in Australia see these rollouts reach a successful outcome, the political obstruction factor against Linux will dissipate.
The most interesting move, however, is that an increasing number of organisations are looking to use Linux as a baseball bat with which to beat better pricing out of entrenched vendors like Microsoft. These enterprises know that without viable, strong competition, Microsoft can charge whatever it likes. It can also introduce unpopular money- filching schemes like Licencing 6.0 at will. Where they have suffered severe competition from Linux on the desktop, in places like Thailand, Malaysia and Indonesia, Microsoft has dropped the price of Windows by 90 percent. This should give enterprises in Australia pause for thought.
It also gives them a target to aim for in what will probably become increasingly robust pricing negotiations with the vendor.
Into this fray, Telstra has led the way. It very seriously and very meticulously piloted desktop Linux systems over a period of 18 months. It ascertained that it could in fact migrate to Linux desktops, for no more pain or cost than migrating to Windows XP. It knew and understood this. More importantly, Microsoft knew and understood this too, which is why Microsoft caved in and gave Telstra a whopping discount. The end result is that Telstra acquired the current generation Microsoft software stack for 40,000 seats for a $20 million fee.
To better understand just what Telstra achieved, we need look no further than a similarly-sized volume-licencing deal that Microsoft signed with the Victorian state government which probably had no equivalent leverage. In that 2002 agreement, the government acquired the current generation Microsoft software stack for the same number (40,000) of seats as Telstra. That price? $80 million. By swinging the Linux negotiation bat, Telstra walked away from the deal smiling.
Now is indeed the time for every responsible enterprise to not only investigate Linux, but to actively pursue serious deployment options too. It is only now reaching a point where such a migration can be converted into reality and that the actual benefits of real price negotiations can occur. And if those negotiations don't end in your favour, you merely take that final step onto a new platform. The money spent in determining how your enterprise can migrate desktops to Linux, will be returned in a spades. It's the best return on investment you're likely to get all year.
Related Article: