January 8, 2004

Prophesies

It's an ancient and not especially honorable journalistic New Years tradition to offer retrospective summaries of the year just past, and prospective predictions for the year just beginning. It's a bit of a gimmick, frankly; but a fun one.

So, as an ancient and not especially honorable journalist with a penchant for gimmickry, I find myself surprised to recall never having produced either a year-in-review or a list of predictions for the year to come. Today I'll correct half of that.

If I'm not mistaken, most of us were here for 2003 and remember what happened. So we'll skip that part and go straight to Pure Prophesy. And if the Alzheimers doesn't set in before 2005, we'll revisit the list and see how we scored.

Okay, so here we go. Ten predictions for 2004.

Prediction #1: Somebody buys Red Hat. My bet is Oracle.

This is based on no inside skinny whatsoever. Just the observation that every company has two markets: one for its goods and services, and one for itself. As the stock market cautiously improves, expect the active market for the sale of "Linux companies" to remain active.

Connect the dots and it's not hard to see where this will go. Novell bought Ximan in August and SuSE in November. EMC just bought VMware in December. Red Hat just bought Sistina. Given the mojo behind Linux, which is growing steadily in just about every respect you can name (enterprises, governments, education, embedded, servers, clients thick and thin), additional acquisitions are a safe bet.

Today, of all the high-profile "Linux Companies" that were around during the boom years of the late '90s, Red Hat is about the only one left that still has a profile. Though I'm no fan of rampant acquisition (a huge percentage of them fail), it's hard to ignore the trends here.

Prediction #2: Quickbooks will come out for Linux. Other mainstream apps will follow.

As Linux continues to establish itself in business, expect familiar accounting apps -- to name just one common category -- to start moving over. For example, as Linux continues to penetrate entertainment and publishing, expect Adobe to try gaining ground on Linux that it lost to Apple on the OS X platform.

Microsoft? Not until they wake up and smell the demand. My guess is that it'll happen eventually. But not this year.

Prediction #3: Outsourcing to low-cost Linux-based services will grow.

Outsourcing is already huge. Folks providing it will naturally rely on the cheapest and most reliable platforms they can mount. This imperative applies whether the customer is a giant retailer like Target (which just signed a 10-year deal for on-demand computing from the Linux giant named IBM), or a small company like the one this reader works for:

Actually I tried something new at this company. Since I end up doing all the support on the systems, we have outsourced everything except the desktops. And I mean everything. Currently our mail server, web server, and ftp server are being hosted by ipowerweb running Linux somewhere in California for $8/month. This gives us 500mb of space and 50 email addresses.

\We have two networks for our people. One network is for development and hooked up to the robot. It is not connected to anything else (so far, never been hacked). The other network is for the office machines and each machine has a firewall and virus protection. No servers yet. Early this year I'm buying a RAID NAS Linux server appliance to hold the backups. I spend no more than 5 minutes a month on maintenance.

From my perspective, the only way to go for small business is outsourced or inhouse appliances managed by a web interface.

That last item leads me to...

Prediction #4: Phrase of the Year: "Mixed Environments."

The all-Microsoft shop is following the all-IBM shop into the growing roster of Things In The Past. But that doesn't mean there will be many no-Microsoft shops. The mature state toward which we are heading is one in which the best tools and materials get used for whatever the job happens to be. Given current cost- and support-savings imperatives alone, this bodes well for Linux.

Prediction #5: New Net-ready embedded consumer electronics devices will get hot.

Why should your digital camera only dump pictures onto your desktop or laptop machine? Why shouldn't it connect by wi-fi or Ethernet directly to a ready server on the Net? For that matter, why should any device that needs to move data not be ready to interact in open and productive ways over the Net with remote storage and all kinds of ready services on the back end? I'm not sure how this one will play out -- whether the trend will be driven by the Sonys and Apples of the world (which will want to restrict you as much as possible to their gear, services and partners), by the ISPs and hosting services looking for value-adds to sell, or what. But I'm sure Linux will make implementation quick and easy in both the embedded device and service providing roles.

By the way, I'll be at CES next weekend, and I'll let you know what I find out there.

Prediction #6: The ROW boat pulls in front, with government help.

In big-picture market data parlance, ROW stands for "Rest Of World." Here in the U.S. we tend to look at everything outside our own borders as ROW. From inside the business sphere we tend to look at government the same way. Well, right now it looks like those are the two categories that will be contribute the most to Linux Leadership this year. One case in both points is a story reported yesterday by the Israeli National News: "Finance Ministry Weaning Israel Off of Microsoft". <http://www.israelnationalnews.com/news.php3?id=55243>

Don't be surprised to see Project Leopard helping that development along. <http://leopard.sourceforge.net/>

Prediction #7: Syndication and notification will go mainstream. And change everything.

Syndication and notification are hot. The leading (or the most "branded") metadata format for syndication and notification is RSS, which stands for Rich Site Summary or Really Simple Syndication. RSS is quickly becoming an infrastructural value-add for the Internet. Thanks to RSS (and its equivalents) when you publish on the Net, a notification goes out to the world, and your content is in a position to be syndicated to everybody who pays attention to the notification, usually through news aggregators. When they write about what you write, services like Technorati (which runs on Linux) <http://technorati.com> let you know what they've said too. More linkage happens, Google takes notice, and juice flows.

RSS is the standard with the highest profile, thanks in large measure to its advocacy by Dave Winer, prime mover behind RSS 2.0 <http://blogs.law.harvard.edu/tech/rss>. It's the first in a series of established and emerging metadata standards, which extend the reach of the category far beyond the online publishing world. All kinds of things can be syndicated. For example, updates to product and parts catalogs. Weather information. Traffic notifications. You name it.

As always, anything that expands the Net as an infrastructure platform inevitably involves plenty of Linux as well.

If you're new to RSS, Amphetadesk <http://www.disobey.com/amphetadesk/> is probably the leading news aggregator for Linux. Hebig.org has a huge directory of other RSS news readers and aggregators. as well. <http://www.hebig.org/blogs/archives/main/000877.php>

Prediction #8: Web services will finally get huge

Loosely Coupled defines Web services as "Automated resources accessed via the Internet. Web services are software-powered resources or functional components whose capabilities can be accessed at an internet URI. Standards-based web services use XML to interact with each other, which allows them to link up on demand using loose coupling." <http://looselycoupled.com/glossary/web%20services>

Most of Loosely Coupled's sources are from 2002. That was the big year for talk about Web services. The following year was a big one for implementation, including all that stuff I just said about RSS, which fits the definition. Now we're reading business magazine cover stories about Amazon's expanding business in Web services. Thanks to Web services, Amazon becomes an e-commerce platform for countless businesses other than its own $4 billion/year retail operation.

And Amazon, of course, runs everything it can on Linux.

Prediction #9: The mundane reigns

The most widely deployed Linux implementations are the least visible. Also, perhaps, the fastest growing. One example is SAMBA, which looks to most enterprises like a Windows file & print server. There are countless millions of SAMBA servers in the world. Another is wireless access points. When I did my study of wi-fi in New York last year, the largest percentage of WAPs I found had a default "linksys" SSID. Linksys, now owned by Cisco, easily has the largest WAP market share. Lots of Linksys WAPs run Linux. Maybe most or all of them. Not sure. In any case, it's an easy supposition to make.

Prediction #10: SCO will lose.

Far too much hot air and ink has already been wasted on this subject. Lawyers withstanding, it will end. Count on it.

Of course, I'd welcome hearing your ideas too. A second guess at this kind of stuff can't be worse than the first one.