The Danger of the Magic Bullet

I'm trying to learn whatever lessons I can from various bad situations I've seen over the years. I think one of the biggest problems is management that is looking for a "magic bullet" to solve all their problems. One more reorg, they figure, will make it all better and suddenly unleash the productivity locked in our organization. At one place, someone declared that the "next generation hosting architecture" will be Oracle RAC on Red Hat on Intel blades, never mind the mountains of information indicating the foolishness of its wholesale adoption (RAC works great for heavy read situations and for high availability, but does nothing to solve write-intensive scenarios). If we can just get rid of all the "expensive Sun" hardware, they figure, all their problems will go away. And so on. I see this as a failure by management to face their problems and put in the hard work necessary to resolve the real, immediate issues. I've had a bit of an epiphany in this regard: Technical solutions cannot substitute for missing or broken processes. In the process, they violate the wisdom of ages by throwing out the baby with the bathwater (so to speak).

I can respect the effort to do something to move the company forward (grow or die, as the business adage says), but things are out of hand when you have technically incompetent people making decisions, refusing to work with the technical folks, and even threatening them when they don't agree with their nonsense. They're using the goal of saving money to push their pet technologies, even when they don't fit. In the manufacturing world, they call this type of arrangement a forced fit.

The Value of Standards

As more and more global players come online and start offering IT products (hardware, software, etc), one obvious effect is more competition. But it also means an opportunity for more specialization.

Right now, however, it seems to me that our industry itself is helping to enflame this issue. How many IT departments do you know that refuse to consider anything but the largest one or two players in a particular segment (be it a database, web server, or whatever)? In effect, we restrict our options because of a mindset. You can reasonably argue that you get better support and a more polished product from the #1 or #2 vendors, but you end up changing your approach based on their tools, as opposed to finding a tool that fits your specific needs. You also typically pay a premium for those products.

There is value in "standards", but how far do you go? Is it really necessary for your company to use exclusively a specific version of a single vendor's web server (as a crude example), when there are perhaps a dozen that all comply with larger, RFC-based standards? Don't laugh--I've seen it at larger companies. In that sense, your company's selection of a particular product is not a true "standard", but only a restriction of choice among products that are all standards compliant. You don't want anarchy and chaos in a datacenter, of course, so I'm not suggesting that it should be a free-for-all either. But I do think we need to strike a balance between the extremes that we don't see today.

What I'm trying to get at is that a larger pool of products and technology can be good for everyone. Companies get exactly what they need, and vendors are able to stay in business. One example might be Apple Computer. They've long been considered to have somewhere under 10% of the computer market, yet they're doing just fine. Some would say they represent a "vertical market" segment. Regardless, you'll never see them in the datacenters of many of these corporations solely because they're not considered the #1 or #2 player in the computing market. If you're a Fortune 500 company, you can refuse to do business with anyone but other Fortune 500 companies, and that's your prerogative. But here's where I see opportunity for smaller, more agile competitors to get an edge. I see the recent X Prize winner as an example of that. Boeing and friends don't seem at all interested in a $10M prize--it's too small for them to notice.

I think the key to the future is compliance with open standards (not to be confused with open source) and a change in mindset. If we can do that, I believe market forces will make it possible for players both domestic (meaning US) and foreign to all share in that market.

As a political observation, the Republican party tends to be friendlier to business (both small and large) than the Democratic party. So it seems to me that the results of this year's elections can have a significant impact on the process above.

A Solid Gold Trashcan

One company's story. A nostalgic look back at the heyday of the dot com bubble. Or maybe things we can learn from it. One of my favorite parts:

"One time, a group of us were standing around in the atrium of ArsDigita HQ, and Philip was holding court. (As you may know, Philip is a talented public speaker with an exceptional ability to entertain an audience.) This time, Philip was describing how cool it would be to have a koi pond suspended from the ceiling of the atrium, so you could see the fish swimming from beneath and from all sides. Brian Stein replied, "You know what would be even better? A solid-gold trash can, burning cash 24/7." I don't think we would have laughed so long and hard if Brian had not struck a chord."

http://michael.yoon.org/arsdigita
http://www.waxy.org/random/arsdigita/
http://eveander.com/arsdigita-history