Are programming languages important?

If 99% of the US population speaks english, its probably not a good idea to start selling books that are written in japanese or latin. If you do, you're severely limiting your audience to a fraction of a percent. This is the same rationale when you develop programs that are written in languages that are not mainstream. Look at all of the people that your book would never reach. Yeah everyone could go back to school and learn japanese or latin but they won't. Besides, we all are very accustomed and comfortable with the language we use the most, which is english. I speak a little spanish but I don't want to learn it in order to read a newspaper. Programming is the exact same way. As the english language changes, we change with it but we never go back to speaking old english.

Also, different programming languages can be viewed as tools. Programmers learn to use the right tool for the job just like we use a philip screw driver instead of flathead for certain screws. When you have to screw 100s or 1000s of screws this becomes more than a minor incovenience.

Imagine that you wanted to build a house and you had to put over 1000 nails in order to make the house hold together but only 250 screws could do the same job. Lets say that you go to Walmart and you notice nails cost $20/1000 and screws cost $50/1000.

Hammers cost $10 and screw drivers cost $1-$6. You notice that Walmart, Target, and Home Depot sell electric screw drivers for around $25 but they don't carry electric nail guns. The only place that carries electric nail guns is 60 miles away and they cost $150 plus the Walmart screws don't work with that nail gun. You have to buy special nails that cost $500/1000 because they are so rare and no uses them anymore (legacy).

What product would you choose, nails or screws?You're thinking, I could go a lot faster and be a lot more efficient with the electric screw driver. Plus its cheaper and I can pick what store I want buy it from.

Programming projects are all about design and construction with programming languages representing the nails, screws and the tools. The stores represent the development communities around the languages.

Lastly, interviewers don't want programmers who have happened to pick a new programming language in the last couple of weeks despite they've be programming in some other language for over ten years. This is why they ask for x number of years with y language. :)

Requirements in job ads don't have much bearing on reality, IMO. Programming is not like writing a novel; I'd argue that within a programming model (procedural, OO, functional, systems) 95% of one's skillset is transferable from a familiar language to an unfamiliar one in a very short period. Experienced programmers will have worked in a number of languages, and should be ready to pick up new ones as the the need arises.

I agree that recruiting will be easier if you look for experience in a popular programming language. This is one of the epiphenomena I was referring to, and can indeed contribute to the cost and feasibility of a project.

Politics and project management -- these are truly the hard problems of software development!

• Yes, I agree with you whole heartedly. by Alfred Smith on Friday November 04, 2005 @ 12:16 AM

  I'm also a programmer and a consultant. I agree with you that good software developers can learn new languages swiftly. I also agree that the focus should be on programmers who understand and apply abstract programming concepts to new technology. The people doing the hiring don't think this way. That's why many programmers find it difficult to get a job outside of their core languages. The reality is that interviewers (the people making the requirements in job ads) are the people who get you hired unless you already have friends who can get you through the door and they want x years in y language no matter how good you are at learning new concepts. It's crazy but it's real. In the real world there is no seperation between what programming languages you use and the politcs and emotions associated with them. And for this reason I believe that you've made a very strong case for why Vista needs to be ported. The human aspect (political, emotional, and perceptions) is what I believe will evetually hurt Vista in the long run. As you said earlier, this epiphenonema can mean the life or death of a project. If this project is ported, it can garner the support of many developers and organizations. Perhaps companies like IBM, Novell, Microsoft, Borland, Oracle and others will create products and services for Vista that will make it more viable and attractive to developers and physicians. A fellow developer, Alfred Smith

  [ Reply to this ]

• Re: Yes, I agree with you whole heartedly. by Ignacio Valdes, MD, MS on Friday November 04, 2005 @ 07:54 AM

  Let's say I agree with your premise that it needs to be ported. Okay, now the question becomes ported to what? Every language has its pluses and minuses as well as its backers and detractors. If you port it to language X, immediately someone will be talking about the tragedy and lost opportunity of going to language X. Come on people! The software is written and it works! IV

  [ Reply to this ]

• Re: Yes, I agree with you whole heartedly. by Malcolm Michael on Thursday November 02, 2006 @ 01:23 PM

  I am a programmer who has worked in C, COBOL, C#, VB, and MUMPS. One BIG drawback of the new languages is the rate of change. Programmers working in MUMPS or COBOL can spend 100% of their time programming. In the "new" world, if you are using Java or one of the .Net languages then 10%-40% of your programming budget is spent keeping up with the latest version, bug fixes, etc. If memory serves, this started with C++. Before C++ compilers pretty much always worked. Occaisionally there was a bug, but those were understood because they were rare. Today people writing the code are not sure if the compiler works or they just don't understand it. Best Always...Malcolm

  [ Reply to this ]