What essential skills should a software developer learn?

  • Thread starter Thread starter Geranimo
  • Start date Start date
  • Tags Tags
    Summer
AI Thread Summary
The discussion centers on a physics student with a background in C++ seeking additional learning opportunities outside their university curriculum. Suggestions include focusing on programming best practices by reading influential books like "Code Complete" and "Effective C++," which emphasize code quality and maintainability. Learning algorithms and data structures is also recommended, though the student notes this is already covered in their program. The conversation highlights the importance of version control, particularly Git, and suggests exploring scientific programming tools like BLAS and Eigen. While esoteric programming languages are mentioned as a fun diversion, the emphasis remains on practical skills such as shell scripting and database interaction, with a strong recommendation to master version control systems to enhance efficiency in collaborative projects.
Geranimo
Messages
19
Reaction score
2
Hi, I study physics with computer instead of labs, and this summer I will only be taking one course (the first prog course). I will have free time to study other things, and I want to learn things that will not be covered in my uni program. I already know c++ basics from Accelerated c++.

I thought of Qt and opengl or an opensource engine like ogre.
Are these good ideas? Do you have suggestions on what could be interesting for my future studies and personal projects? Thanks
 
Technology news on Phys.org
First, congrats on learning C++ from Accelerated C++. This is one of the most highly recommended books on the subject.

I have kind of a curve-ball suggestion. Instead of learning a new language or two, or maybe in addition to learning them, learn to program well. Get the book _Code Complete_ by Steve McConnell. Get the Effective C++ books by Scott Meyers. Sample the other books about coding style, reducing bug count, and making your code easy to read, easy to test, easy to maintain, etc.

Many is the time I have had to follow behind somebody who learned algorithm and language features, but never the "ethical" way to apply them. It can often be very painful.
 
  • Like
Likes harborsparrow, Geranimo, Greg Bernhardt and 1 other person
Plus one on DEvens suggestion on reading Code Complete.

Some other suggestions:
  • Learn about algorithms. The most important class a non-computer science major can take beyond the basic "can has programming?" class is "algorithms and data structures". Read "Introduction to Algorithms" by Cormen, Leiserson, Rivest, and Stein, or "Algorithms in C++" by Sedgewick.
  • Learn about version control and configuration management. Much as I dislike it (the words "hate" and "loathe" aren't strong enough), I strongly suggest you learn how to use git.
  • Learn the old-style tools of the trade of scientific programming. That would be BLAS (basic linear algebra subprograms) and the various packages such as LAPACK that are wrappers around BLAS.
  • Learn some new-style tools of the trade of scientific programming. For example, Eigen.
  • Learn a very different (but practical) language such as lisp or haskell. Functional programming will give you a new perspective on programming. The newest version of C++ (C++ 2014) introduced a number of functional programming concepts to the language.
  • Learn a very, very different language such as brainf*ck (substitute the asterisk with the appropriate vowel), which is one of many Turing tarpit languages. Here's "Hello, World!" in that language:
    Code:
    ++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++.
  • Learn a very, very, very different language. Or more than one. It's a bit fun. This is LOLCODE:
    Code:
    HAI
    CAN HAS STDIO?
    PLZ OPEN FILE "LOLCATS.TXT"?
        AWSUM THX
            VISIBLE FILE
        O NOES
            INVISIBLE "ERROR!"
    KTHXBYE
 
Last edited:
  • Like
Likes Geranimo and fluidistic
Thanks for your suggestions! My uni program already has algorithm in it so I think i'll focus on something else. Also, these strange languages are just for fun right? Nobody uses them seriously? :nb)
 
h15339A65.jpg
 
  • Like
Likes Physics Beard, Geranimo and DEvens
Geranimo said:
Thanks for your suggestions! My uni program already has algorithm in it so I think i'll focus on something else. Also, these strange languages are just for fun right? Nobody uses them seriously? :nb)

"Esoteric languages" are a bit of fun, but they are turing complete. So library support aside, you can write anything in them that you could do in any other turing complete language such as C++.

I can't really recommend something specific as I don't know your intentions but I will suggest that you learn the tools used by software developers. That means stuff like version control (i.e. github), shell scripting, how to talk to a database, and possibly some kind of dynamic language (Python?) as they're really useful for doing quick jobs.

Why?

I contribute to an advocacy group that teaches developer skills to domain programmers (http://software-carpentry.org/). For example I once met a biologist who was part of a group that shared the results of their software by emailing each other a spreadsheet which they updated and emailed to the next guy. Not only is this inefficient, it also breaks when someone misses an update. We taught them how to write the code that connects to a database, and some basic SQL (the language most databases use for queries), and enough bash (linux shell script) to automate the system, and they were much happier after that.

Anyway my #1 suggestion is that if you don't know how to use a VCS like git (github's backend), learn this NOW.
 
  • Like
Likes D H and Geranimo
Thread 'Star maps using Blender'
Blender just recently dropped a new version, 4.5(with 5.0 on the horizon), and within it was a new feature for which I immediately thought of a use for. The new feature was a .csv importer for Geometry nodes. Geometry nodes are a method of modelling that uses a node tree to create 3D models which offers more flexibility than straight modeling does. The .csv importer node allows you to bring in a .csv file and use the data in it to control aspects of your model. So for example, if you...
I tried a web search "the loss of programming ", and found an article saying that all aspects of writing, developing, and testing software programs will one day all be handled through artificial intelligence. One must wonder then, who is responsible. WHO is responsible for any problems, bugs, deficiencies, or whatever malfunctions which the programs make their users endure? Things may work wrong however the "wrong" happens. AI needs to fix the problems for the users. Any way to...
Back
Top