The Wicked Witch is dead! Now, go back to work.
Ding dong! The Witch is dead. Which old witch? The Wicked Witch! Ding dong! The Wicked Witch is dead.
By the time you read this, you’ll likely know that the greatest threat to Linux and open source in its entire history has been vanquished. The Wicked Witch is dead. The dragon has been slain. The Army of Undead has been vanquished. Godzilla has been pushed back into the sea. Enough monster metaphors? Yeah, I think so.
But while the good guys prevailed in this battle, the war is far from over and we’ve been distracted nonetheless. Through fear, uncertainty, and doubt (FUD) the Wicked Witch and her evil minions set us back about three years and surely slowed enterprise adoption of open source.
In the meantime, instead of standardization, we managed to fracture the landscape even more. There is still too much variance between the popular distributions, and despite the best efforts of projects like Portland (http://portland.freedesktop.org) and the Linux Standard Base (http://www.linuxbase.org), we have no unified API between the GNOME and KDE desktops, and lack a common infrastructure to enable cross-distribution application development, respectively.
Sure, we’ve got universal support for stuff like DBUS for interprocess communication (IPC), but have we seen the tangible benefits of it yet? And in the long run, is it really a good idea to devote energy to two different desktop platforms when the effort could otherwise be amassed into one really good one?
Sure, in the last three years, some great things have been accomplished. Ubuntu has become an incredibly popular distribution, and even Dell has started selling Linux-based systems. Red Hat is making significant progress in displacing legacy Unix systems, Fedora remains strong, and SLES and OpenSUSE have also made significant inroads. Debian released Etch, finally.
But for every major milestone I can think of, I can think of a half-dozen things that still require a lot of attention. Let’s start with one of my biggest pet peeves — the Linux desktop. Ubuntu is wonderful and all that, but getting it (and just about any other major Linux distro) tweaked to the point where a Windows or Mac user gets an equivalent is something different entirely. Oh sure, you can install something like Automatix (http://www.getautomatix.com) but as the community recently discovered, it can have long term effects on stability and can literally break your operating system install.
Ubuntu (and effectively all other Linux distributions) doesn’t officially support packages in third-party repositories, and most of the goodies you need to install to make all that fun multimedia stuff work is illegal to use on Linux anyway, including all the cool browser plugins that you need for a rich Internet experience (see “Plug in Your Browser– http://www.linux-mag.com/id/3131/ ”).
And how about our office suites? Sure, OpenOffice.org is a great open source productivity suite, and its usability has improved considerably in the last several years. But I still am reticent to use it to interoperate with Microsoft Office users. Of course, I can import documents just fine from Office, but export a complex Word or PowerPoint file to its native format and expect it to come out looking like the way it started in Office? Fuhgeddaboudit. For all its great user friendly interface and polish, OpenOffice.org is still not a drop in replacement for Office. StarOffice performs import/export better, but not significantly so. We still have a long way to go before switching off Microsoft Office in an enterprise environment.
On the server, we also have a number of things that still need to be addressed. Open source Active Directory integration, while better than it was three years ago, is still not enterprise-quality.
The commercial solutions, such as Centrify and Vintela VAS, are still superior to the Winbind features built into SAMBA and are far easier to configure and administrate. Xen, for all its promise, still lacks the polish and manageability features of VMware, although hope springs eternal that XenSource’s new corporate masters, Citrix, will mold it into a true enterprise-worthy virtualization solution.
However, OpenVZ and the kvm project looks to further fragment the landscape and make true open source virtualization solutions that much less appealing than the commercial alternatives. I also don’t even want to go into the fact that open source and scalable enterprise messaging solutions seem to be mutually exclusive. For all the promise of stuff like Zimbra, OpenExchange, Scalix, and a half a dozen other messaging and communication gateway players, no organization seems truly interested in displacing its Exchange or Lotus Notes infrastructure for something based on open source, because there’s just no clear cut standout in that space.
So yes, my fellow munchkins in Open Source land, the Wicked Witch is Dead. Now, get the hell back to work!
Jason Perlow is disgruntled because his job in the Emerald City was sent offshore to Treasure Island. You can send employment opportunities to