I was in much the same position 12 years ago at this company. I am now CIO with 7 people on my team with several business partners to help manage the infrastructure. My advice for what it is worth:

• - Take time every day to assess and analyze the bigger picture before allowing yourself to get drawn into the details.
• - Look at the entire system from a risk mitigation perspective. What areas are most likely to cause "meltdown". Spend the most effort there.
• - What are incremental changes that can be made that improve the overall risk picture? Focus on the biggest bang for the buck.
• - Defer anything that works well enough for the time being.
• - Avoid big bang solutions unless they can be contained and tested well, with the capability of rolling back.
• - Get help where necessary.

...having read the follow-on article suggesting that wiping out mosquitoes might not result in a significant change in the overall environment, I thought of the butterfly effect. It is hubris to expect that we could foresee myriad bad possibilities.

In analyzing the light coming from quasars (active nuclei of distant galaxies), astronomers realized the rays had passed through gas that contained only hydrogen and deuterium, elements that formed minutes after the Big Bang.

The clouds, which are located about 12 billion light-years from Earth...

...a quasar is a compact region in the center of a massive galaxy surrounding its central supermassive black hole.

Wikipedia

Over a long period of time, the slightly denser regions of the nearly uniformly distributed matter gravitationally attracted nearby matter and thus grew even denser, forming gas clouds, stars, galaxies, and the other astronomical structures observable today.

Wikipedia

...so light from "a compact region in the center of a massive galaxy" went through a gas cloud 12 billion years ago. That would mean that if the universe is 13.7 billion years old, the massive galaxy had formed in less than 1.7 billion years... hmmm

We are in the Minneapolis area where a tech recruiter friend of mine emailed me this morning regarding his layoff: "IT unemployment in the Twin Cities is currently at 1.7%, so most of our clients have to use us because they can't come close to finding/recruiting talent on their own." I do not think that my friend will have much trouble in this area.

My friend sent me an email yesterday: "I'm about to go into a meeting where Adobe is laying off my whole team." He had worked on Flash for many years since Macromedia owned the project. After the meeting he said, "Just got out of meeting, I have a job until April 20, paid thru May 15, decent severance, but job will end."

One of the sloppiest, non-informative articles that I have recently read. Coding is the last part of any automation development. What about bad analysis, bad design, poor tool/OS/infrastructure/integration choices? What about the problem of being forced back to the Waterfall development methodology to be able to offshore development? What level of coding are we talking about: embedded systems, business architecture, tools, OS, (etc)? Any security expert would know that code security is the least of the problems regarding overall automation security.

final

adjective/fnl/

Coming at the end of a series

>Why do you have to have the business explained to you? Does it mean that you don't understand the business you're working in? We generally end up explaining the business to the users. The paradigm that we work in is complicated and the users in one small part do not understand the interactions with another part. The problem, then, is more of a business analysis nature than a pure IT solution one. The request might be "Give me a invoice for all of the natural gas transportation costs for all of our customers." The question back is "Give me the costing rules, rates, tariffs for the 15 pipeline pipelines involved." This is not a matter of "understand the business." The simplistic drivel I see in these threads avoids the truly complicated nature of some businesses. Go find a "Cloud" solution that collects and monitors all of the tariffs, rates, schedules, tax rebates, costing rules, inventory rules and costs, daily balancing rules and the like for all natural gas pipelines in the US and can apply them correctly to each customer invoice for all customers across the US on an agency basis with our own costing rules built in.

The article was mostly sweeping generalization with few specifics to back up the argument. As CIO, I find that the problems regarding delivering relevant, efficient, meaningful automation solutions has much more to do with the business units' inability to articulate a problem statement and business case than with our ability to provide a technical solution. The paradigm is generally "give me one of these". The return question is always "What are you trying to solve?" We try to steer our users to third-party hosted solutions whenever there is one (or two or n) that will fit the problem.

Just wait until your grant account is frozen by PayPal...

Taking a step back from the coding errors and even a step back from software development errors, there is a fundamental where failure to adhere to it will produce bad results from start to finish. This idea is not unique to software development but I see large software development projects that do not follow it fail on many levels. Problem statement: Automation is required to solve a specific problem. Each and every part of the project has a problem statement (or should). Gratuitous features are gold-plating. Take Microsoft Word, for instance. A large majority of the problem-solving features (WYSIWYG, spell-checking, grammar checking, etc ) were solved years ago. That means that most of what was left to provide had not much to do with solving the basic problem that the tool is designed for: Editing and printing documents. Yet Microsoft has to create an illusion of need so that consumers will be willing to shell out $400 for the next upgrade. Basically, this is gold plating on a huge scale. So, with no problem to solve, the developers have no fundamental rule to follow. Taking Microsoft Word to illustrate what happens with gold plating: Every single person in our company dislikes Office 2007. Microsoft completely changed the interface forcing the user to re-learn the most basic tasks. The code for the new interface functions perfectly. There are no apparent coding errors. The error was made at the top of the decision ladder. After several years of learning, users had no trouble navigating the tools and options. There was no problem to solve. Microsoft needed to feed its illusion machine so it created eye candy at the expense of the users. Microsoft just happened to be the biggest target, but this issue is apparent throughout the industry.

Quote from the blog: "The key learning over the last year is that when we change the operating system, it takes time to let the ecosystem make sure that the hardware and software that they build works well with Windows Vista. So as we release Windows Vista SP1 to manufacturing, we are going to be thoughtful about when and how it gets distributed." Note that the Microsoft wants to "make sure" that the "ecosystem... hardware and software" must "work well with Windows Vista" and *not* that Windows Vista works at all with the "ecosystem". And as a result of this musing, Microsoft plans to ration SP1. being "thoughtful about when and how it gets distributed". This is yet another declaration by Microsoft that the market marches to Microsoft's beat and that Microsoft can produce any crummy thing that they please and the "ecosystem" has to adjust to Microsoft.