At the beginning of the year I announced my ambition to learn programming, today I am announcing that I will drop my efforts. The reason is not, that this is getting too complicated for me to comprehend, neither that I don’t see value in it. Actually quite the opposite, the classes so far taught me a lot about the basic principles of web application development, a core knowledge of web technologies, just enough to have intelligent conversations with others. Discontinuing now is not really a matter of giving up, it is the recognition of knowing what I wanted to know, and that the next step to take would take more than I am willing to give.

To get really good at any technology, you got to use it on a regular basis. Learning a programming language, regardless of which, is not really a matter of intelligence, I think anybody can do this. You need to use it day by day to become really good at it. I will not turn into a programmer; I will never have enough practice to become really good at it.

Technology shall help us to make our jobs more efficient. In my current role, I am what others may call a knowledge worker. I work with a lot of data residing in spreadsheets, databases, on the web and in other locations. I must admit, that the way I was working with the tools that are installed on my desktop, the tools that where my home in my day to day activities, especially the MS-Office Suite, was rudimentary at best. I changed that. I decided that I needed to learn more about the tools that I am using every day to make the things I need to get done faster.

Anybody that manages anything manages by numbers. Your ability to manage is largely limited by what you can do with the numbers that are provided to you. There is tremendous power in what Microsoft calls “personal BI” in which knowledge workers can learn pretty fast to mesh up numbers in large volumes, create reports they can analyze, slice and dice and even combine with freely available information from the web. They beauty of it is, that it is to some degree free of charge, you must have the most current Version of MS-Excel to get started, but the plugin is free. It gets you hooked pretty fast, once you see that you can process 10 Mio rows in MS-Excel with blistering speed in memory, you ask yourself how you could live without it. Microsoft learned also from previous mistakes, making this solution not only simple (let’s be honest, MDX and the old Analysis Services were anything but), but also really scalable and sharable. As they drag you in with the sexiness of these reporting tools, they will also charge additional bucks, licenses for additional SQL Server instances you want to run, licenses for Sharepoint Server and other things, so I guess at the end it is not really free. It is sort of the first fix that is free.

To stay on the topic though, the absolute greatest piece of it all is, that the Microsoft people create a tremendous amount of learning materials that you can access online and for free. My prime destinations for learning materials are:

http://channel9.msdn.com             (very technical)

and

http://www.microsoftvirtualacademy.com     (not quite as technical)

Both offer great educational content, especially via video. I download these videos in MP4, store them on my tablet and watch them in airports, hotel rooms and airplanes while travelling. I watched so far more than 200 of these videos across all kinds of topics. Reassessing what to learn was also partially triggered by the way the learning materials are presented to me. While http://www.codecademy.com provided a nice platform and interactive learning and coding, it was not available to me when I had availability – there is rarely internet on planes. In addition, videos are something that I can better learn from, I can replay important areas when I need them and it combines nice visuals with great Audio. My decision to switch was therefore not only based on the subject matter, but also in the way it is being presented to me.

This brings up my last point: Believe me when I say this, but I have not read a 100-page manual in probably 2 years. I mean, don’t even bother sending it to me. My attention span does not last that long to read some poorly written documentation that long. Since I am not reading these either, I have stopped creating these too. This does not mean that I don’t create learning content anymore, it just does mean that the content that I provide will not be in text – it is on video and I have made more than 50 that I share with my immediate community. It is not only that I reassessed what to learn – I reassessed how to teach as well.

When you look at this idiom, you will quickly recognize that the ideal of everything in one comes with significant compromises. Smartphones, as great as they are, have significant limitations based on screen size and input devices (your fingers just are as big as they are), making a smartphone not really ideal to do a lot of reading or writing. The screw driver of the Swiss army knife is probably not as practical as a specific screw driver in the right size for the job; I guess you catch my drift.

Digital convergence that puts more and more features in a single device has basically stopped in the world of software. If you look at current trends, you see that more and more things are done via ‘Apps’ which have comparatively little functionality within themselves, but we use lots of them. Digital convergences in Software today means, that we work with lots of different modules and pieces at the same time, but we want them to seamlessly integrate with each other. In Software, the current buzzword for this is interoperability. More and different software modules work coherently and seamlessly together.

A business in the food industry is managed by numbers which are represented in reports. To manage your business well, you have a range of different requirements that not one single reporting tool can offer. There is no one-size-fits-it-all or ‘egglaying woolmilksow’. Considerations for the choice of reporting tools include:

• Delivery: via File Folder, email, fax, browser, print,…
• Offline:     do you need the reports when disconnected from the internet?
• Real-time: do you need most current data or can the data be pre-aggregated?
• Write-back: do you need the ability to add comments or change data, or do you just read, slice and dice?
• Hardware: do you run this on PC’s or on other form factors, such as tablets, smartphones?
• Internal Data: are the data in a single database or single application, or are they spread out across multiple internal systems?
• External Data: Do you want to incorporate data from external web services like USDA, FDA, Urner Barry, Chisolm and others?
• Organizational: Is the reporting managed by end-users, department leads or IT?
• Data Volume: Are we dealing with a couple hundred thousand line items a year or tens of millions and more?
• ….

Now, if you say ‘all-of-the-above’, you will easily wind up with a number of reporting tools that you need to incorporate and orchestrate to work together. Even BI Solutions from single vendors, whether IBM (Cognos), Oracle (Hyperion), SAP (BusinessObjects), Teradata, Microsoft (my personal favorite!) or SAS are comprised of a range of tool sets that address issues in different ways. Each of them has strength and weaknesses, and it is more common than not where companies decide to use multiple ones in their environment, because they are truly great in one aspect, but too weak in another.

Every reporting tool has its purpose, the build-in reporting tools in your ERP package, the look-up reports and screens that are integrated and even MS-Excel has its purpose. The issue is not how many we use, the issue is how we can keep them peaceful and well living next to each other, like hens, sheep, cows and hogs – sometimes on separate farms.

Most companies turned the corner at some point, replacing their old homegrown system with standard ERP-Systems perhaps complemented with some add-ons that may be still home grown, but it is surprising how many of these legacy applications still exist. This does not come as too much of a surprise to me with really large organizations, but when I encounter a food processor with a couple of locations and perhaps a few hundred to a couple of thousand employees, you got to ask yourself – Why?

I admit, a custom tailored suit fits normally better, and a lot of these legacy applications have certain detailed features that make the move to another application especially hard. I guess the point though should be to rethink the overall solution to the problem and not trying to move the old habits to new software. “Now we got a new system, we can make the wrong things faster.”

Unfortunately a lot of companies do not turn the corner. They may go out of business (Bang!), get acquired by somebody else (Bang!) just because they did not move in time. The underlying reason is twofold:

1. Companies look at software and hardware as capital expenditures and internal personnel as expense. While they like to reduce the latter, they keep justifying that they need these people, because they always had. Since they hardly ever invested in software and hardware, a significant shift in the expense structure is hard to comprehend and execute.
2. People that are dependent on that technology fight with tooth and nails to keep the old stuff. The average COBOL or RPG programmer today is in the 50s. They do not bother learning new programming languages such as C# or JavaScript, they also refuse to learn different ways, such as object oriented programming or Web technologies. Their life depends on staying with the old stuff, because they will hardly get a new job with their current skillset.

This situation is for a lot of companies like driving full speed towards a cliff. Existing management in IT will fight to keep the old stuff until they retire. Their legacy will be that the leave a company with a system that nobody that comes after them can still maintain. Their interest is not aligned with the goal of the organization to become lean and efficient or even long lasting. Their motives are largely selfish. Their strategy though will jeopardize the future of the organization they work for in a dramatic fashion unit the Big Bang occurs – whether that are expiring software licenses, hardware that is no longer supported or spare parts are sparely available or not finding successors that can and want to deal with this, because it jeopardizes their future. The Big Bang will occur at some point, so change or watch it to happen.

PS.:    If you read this, and you are in management of a company that runs legacy software, this article probably does not apply to you. You must be a good and well-intended person.

Traceability has probably been the most dominant topic in industry publications for the past couple of years including this blog. Traceability systems are the ways and means we are using today to perform recalls. The problem that FSMA tasked FDA is resolving the speediness of the recalls, which in the PBA case, the Jalapeno case and the spinach case took weeks until the ultimate source of an outbreak was found, contaminated product removed from the shelves and a lot of innocent businesses harmed. The first step to redefine how traceability systems need to work in the future, FDA and IFT worked together on defining pilot studies to see what is feasible to do. Consider these pilot studies ‘feasibility studies’. The cases that FDA has defined have been published on FDA’s blog and the include Tomatoes, Peanut Butter and Kung Pao Style Chicken. Read the full blog entry real careful!

Kung Pao Style Chicken! If you have followed the history of these pilots, you may have noticed that they were tasked to do just 2 pilots and came up initially with Peanut Butter and Tomatoes. Regulators in FDA and IFT must have thought something like “Let’s look for some serious trouble!” While the first two supply chains are practically entirely contained within the regulatory domain of FDA, Chicken is not!

What do we expect to find out from these studies? I guess the result is clear: Traceability technology that FDA is going to mandate for its foods must be sooner or later expanded to all USDA regulated foods. This means that either the entire effort of making our recall systems faster will stall, because of some pretty obvious large holes, or we will look at another large change in our traceability system requirements for the entire food industry this time around. Stay tuned, the results of the pilots are expected at the end of summer. Enjoy the summer!