Geeks With Blogs
got net? Kevin Hazzard's Language Interop Blog
| Home |

I'm here at PyCon 2010 in Atlanta. One of the pervasive themes (other than improving Python performance and testing which are very good pursuits IMO) is the recurring, omni-annoying "The Web is the New Operating System" concept that the .com busters (who failed us so miserably in the 90s) just love to espouse. The Web is a lot of things to a lot of people. If you asked 10 people what they thought "The Web" really means, you would probably get several very distinct answers, depending on their experiences and world views. However, when sampling on the street, the likelihood of getting an answer describing the Web as "The New Operating System" is very low. In a statistically larger sample of, say 1M people, you might get some answers like that but the bulk of the responses would coalesce around activity-oriented themes which describe the Web predominantly as:

  1. An e-mail medium
  2. A place for playing games
  3. A marketplace for gambling and other vices
  4. A meeting place for friends or communities of like-minded people
  5. A portal for interacting with product and service companies
  6. et cetera

"The New Operating System" still wouldn't make the top 10 list in a statistically-diverse survey group. Consumers of the web don't think of the Web as an "application platform" but here in the rarified air at PyCon, that's the prototype that we use. We are somewhat like the king's kids. We know how the court really works. We know all the dirty, little secrets lurking about the castle. We understand the lines of history and authority that make the kingdom function. But we aren't too far removed because we rely on that same system of rules for our livelihoods which grounds us a bit.

I sat in on two sessions at PyCon this weekend where the speakers essentially said that if we weren't having fun in software development, first and foremost, then we had lost our way. We needed to be reminded, they said, of the unwritten rule mandating that work be fun first and provide an income second, possibly third. I'd like to believe that but real software development is often not fun at all. It's hard work that requires many late nights, early mornings and lost weekends to stay on top of one's craft to solve world-scale problems. There are fun moments of exploration or epiphany along the way but the thousands of bleary-eyed, sleep-deprived mornings I've suffered over the past 25 years can, in no way, be described as fun.

I've written software that helped my clients to change the world and that, my friends, is what it's all about. Having some fun along the way? You bet. I'm for that. But having fun is not my raison d'être. I think that some of the people breathing this rarified air here believe that the titillation they feel when they code is the primary reason for coding. But that's a bit like saying that the main reason for driving an automobile is to impress your friends. It leads to all sorts of after-market opportunities for automobile parts vendors, for sure, but it sort of ignores the fact that getting from point A to point B on a daily basis has a lot more utility value, especially when points B include places like doctors offices, grocery stores and the spots where you might eek out a living.

You see the phenomena that naturally accompany the titillation-oriented world view all around you at PyCon and similar conferences. There are often dozens of alternative solutions in the bubble of any particular problem space. Diversity is a good thing, right? To a certain extent, that's true. But ask that question of the people who love someone who is severely ADHD and you may get a different opinion. Too many choices can be detrimental if they distract us from solving the problems at hand. At PyCon, shiny objects abound. There's a new technology this year for everything imaginable. Having been immersed in the Microsoft world almost exclusively for so many years, I had often wondered if attention deficit disorders were were common only in that space. Every year, Microsoft seems to drum up an even cooler new way to access your databases than the one they invented the year before, for example. But it's clear to me now, finishing up my first PyCon, that the propensity to create shiny objects and chase after them pervades this world just as deeply as it does Microsoft's. Stabilizing the platform to give customers some comfort isn't a real concern here. Fun has definitely won.

Don't get me wrong. I am not picking on the Python crowd. I love innovation as much as the next developer. The desire to rapidly innovate is strong in all of us. But our customers and clients drift toward technologies that have longer shelf lives and resist change. They have 5 year-long depreciation cycles for most of their capital investments that temper the rate of change for most of their operational expenses, too. And those limiters find their way into our design decisions, like it or not.

So, going back to the premise espoused by some that the Web is the New Operating System, what kind of investment are most companies going to be willing to make in an operating system that can't make up its mind from year to year what it wants to be? You might argue that the Web is largely free so corporations have nothing to depreciate when they decide to adopt Internet technologies. But that's not entirely true. Nothing is free because conventions and procedures of any kind create new dependencies. Using Google Docs instead of Microsoft Office, for example, brings dependence on many networks and tools operated by those whom you don't employ. And those whom you do employ must create Standard Operating Procedures that try to approximate the qualities of the services that they were able to offer when the software and networking were proprietary.

Looking outwardly, consumers may care less that the ground beneath them is constantly shifting. They may even demand that your company implement the best new technologies. Marketing is, after all, an exercise in creating expectations which often drives demand from the bottom, up through your various delivery channels. In the same way that Intel Corporation was so good at self-fulfilling Moore's Law throughout the 1980s and 1990s by creating demand for their ever more capable processors, Google and others are similarly adept at creating demand for their products. And their main product is change.

Caveat emptor as usual.

Posted on Sunday, February 21, 2010 2:27 PM | Back to top

Comments on this post: PyCon's Rarified Air

Comments are closed.
Comments have been closed on this topic.
Copyright © Kevin Hazzard | Powered by: