Paw Prints: Writings of the maddog

Lately I have been seeing blog posts and hearing statements where (if I had to boil them down to an appropriate theme) seem to be along the lines of “all companies are evil”.

The authors of these statements are not quite that strong in their words, but the tone of their articles loudly questions various things that "all" companies do, and usually in a very negative way.

I must admit that I sometimes offer the opinion that one or two particular companies are “evil”, but I believe I am very selective in my choices, and for the most part I believe that companies “do good”. Companies produce things for customers, employ people, and for the most part are generally “good”. I give most companies the benefit of the doubt, and part of the reason for this is having spent a lot of time inside large companies helping to make some of these (sometimes difficult) decisions.

Recently I saw a blog on Red Hat and its decision to pay a company for the use of that company's royalty bearing patented software. The author of the column lambasted Red Hat because (in the author's opinion) Red Hat had (on one hand) stated a policy against software patents, and (on the other hand) seemed to be “encouraging patent trolls”.

I submit that Red Hat could do both legitimately.

Red Hat, as a company headquartered in the United States, has to obey U.S. law. That includes things like obeying laws about copyright and patents. Whether you believe in those laws or not, if you violate them you put yourself, your company, your employees and your customers in jeopardy. Red Hat was using software that was judged to use patented techniques. Red Hat could try to remove that software. If Red Hat can not remove the patented software, as long as the patent laws remain the same and that patent is deemed to be legitimate, Red Hat has to pay the royalty fees if they are shipping the software. Red Hat also was right to negotiate with the company to try and come up with the lowest possible royalty payment.

Red Hat is also building a patent pool. Some people could criticize them for “supporting the concept of software patents”, but Red Hat has continuously testified in many venues against the concept of software patents. As a defensive measure, and with a promise never to use the patents against Free Software, Red Hat should not be criticized for “fighting fire with fire.” The day that software patents are dissolved, I believe Red Hat will be as happy as anyone. Until that time, they have to mount a defensive measure.

Another “old saw” of being evil is about the concept of computer companies “locking in customers” by creating operating systems that made it difficult to move programs from vendor to vendor. I am not going to say that companies never went through this thought process, but I do not remember any of the companies that I worked for saying to the engineers “create some interface that locks our customers to us.” The conversation was more about “how can we make our systems more efficient and useful for the customer”.

Remember that in the early days of computers the memories were small (measured in Kilobytes, not Megabytes and Gigabytes), the CPUs were slow (cycle times in milliseconds, not GigaHertz), disks were small (my first hard disk was one-half megabyte in size) and all of this was expensive (I remember paying 128,000 dollars for 64,000 bytes of core memory in 1976). If the system vendor could make the computer ten per-cent easier or more efficient to use, you could literally save the customer tens of thousands of dollars.

If “locking a customer into our computer through the operating system” had been Digital's mantra when I was there, we would have needed only one operating system on the PDP-11 instead of the seven that we supported:

• DOS-11
• RT-11 for real-time
• RSX-11 for “multi-tasking”
• RSTS/E for time-sharing/education
• IAS – real-time and time-sharing
• DSM-11 – Digital Standard MUMPS
• V7M-11 (a Unix variant)

which all came from Digital as well as

• Unix (various flavors).

These operating systems had different interfaces trying to deliver the capabilities that customers needed for these (by today's standards) very slow and small machines. Often we addressed different market segments with the different interfaces.

I do not remember feeling “evil” at the time, nor do I remember anyone using the phrase “locking the customers in”.

I recently ran across an article where a pundit was complaining about how some companies were making customers pay more money for systems without advertising and without “bloatware”. While I do not like “advertising bloatware” either, I pointed out that the advertising was probably used to offset some of the cost of development and deployment of the system, and the customer had the choice of getting the system at a higher price, but without the “bloatware” or paying a lower price and spending the time to remove the "bloatware" later. Of course I do not suffer through this as much as other people, since I normally remove ALL of the “bloatware” at one time when I remove the operating system that came with the computer, and having some companies pay the system vendor money to put their “bloatware” onto the box simply allows me a less-expensive computer, which is fine with me.

But is this advertising bloatware “evil”? No, it is a business practice that may allow for an over-all less expensive solution for the end-user customer and an acceptable profit margin for the vendor.

Another business practice often criticized is “bundling”. This is where a software company will strike a deal with a system vendor to put a copy of their software into every single system. Often the software company (who may charge hundreds of dollars a copy for their software when sold “off the shelf”) will charge only a few dollars (or perhaps nothing at all) to get their software (or a portion of their software) delivered to every customer. If the software is useful to the customer, this is often a very appreciated practice. It reduces (or almost eliminates) cost-of-sales, and gets the software into the hands of every customer at a low cost.

When I was at Digital I formulated a business deal where we bundled a database engine from a well-known commercial database into our operating system. The bundling deal allowed our Unix customers to count on that particular database engine to always be available to them, which allowed them to develop sophisticated software that could rely on having that database engine.

Our bundling deal did not prevent any other database engine from also operating on the platform, nor did we “brand” the database to create a preference, although admittedly our choice did create a preference for our customers. However at that point in time the use of a relational database on Unix systems was about four percent, spread across five major vendors and with a relational database engine often costing more than 100 thousand dollars per system. Our action made it “easy” to justify the use of a relational database in a project, whereas before you had to allocate hundreds of thousands of dollars for that decision.

Bundling is not necessarily a bad thing in the industry, but when it eliminates choice by approaching monopoly proportions, that is when it becomes bad.

Not all companies are “evil”, and not all of any company is necessarily “evil”. What some of us perceive as “evil decisions” may be based on rational thought that took the company down a particular path. If we knew more about the decision process, we might understand more about the company, how it works and why it made a seemingly “evil” decision.

I once had a customer at Digital that was upset about a decision DEC had made as a company. The customer was so upset that they swore they would never buy anything from Digital again, and were saying this very loudly on a mailing list. I called the customer and explained our decision. The customer saw the rationale behind our decision, agreed with it, but still felt “tricked and trapped” by his purchase. So I arranged for him to return two (now used) computers to us and instead receive two used computers of equal value and power from a competitor that he favored.

The customer was so impressed that I took this step to make him satisfied, that he returned to being a customer of Digital's, and although he described himself as a “small shop”, I found out later that he was the research arm of Boeing Aircraft. What he bought, Boeing tended to buy.

Carpe Diem!

Comments