I've decided that I need a section to collect my own notes on programming. So here they are.
List of Languages
I'll start off with a list of the programming languages that I have used [i.e. I've written at least one line of the language to do something I want to do].
This is a long list, won't be chronological, and I will probably forget a few [taken from this Twitter thread] Yes, I'm aware that some of my stated "variants" are considered additional languages in their own right. Due to their spiritual link in my head, them be the breaks:
BASIC (GW, and variants like QuickBasic)
My dad wrote a menu system in GW-BASIC, and showed me how to change the data section of the code to add in new programs. I tinkered a bit with BASIC code from magazines.
https://en.wikipedia.org/wiki/GW-BASIC
DOS batch files
The BASIC menu system was purely graphical. The backend was a series of DOS batch files that ran when the computer user typed the number on the screen (e.g. 4 + <Enter> loaded Railroad Tycoon)
https://en.wikipedia.org/wiki/Batch_file
Pascal (Including Borland TP7 and FreePascal)
The first programming language I was formally taught was Pascal. The school teacher wasn't great, but the language had amazing help files which found me very quickly writing quite complex code.
https://en.wikipedia.org/wiki/Pascal_(programming_language)
x86 Assembly language
The Pascal help files included some embedded assembly language (e.g. for making a click sound when a key was pressed), so it was a natural supplementary language to Pascal.
https://en.wikipedia.org/wiki/X86_assembly_language
C++
I think my first attempt at a C++ program involved creating 3D images from random dot stereograms. I was later on formally taught it at university, and it's been my go-to language when I need something that runs fast. Really hard to debug, though.
https://en.wikipedia.org/wiki/C%2B%2B
PowerBat
A batch file-like language which I tinkered with when I was trying to improve the menu system. I used it to create a dynamic menu system that would wait for keyboard entry. It was a memory hog, and needed compilation.
[Can't find a web link for it, sorry]
x86 machine code
My first and last attempt at writing raw x86 machine code involved creating a file containing bytes that represented a computer soft reset signal. I wrote it to the first two bytes of the hard disk the system booted from. <facepalm> #DataLoss
https://en.wikipedia.org/wiki/Machine_code
Java
I learnt Java in my first year of university computer science. This was pre-eclipse, so we had to remember the "public static void main(String[] args)" incantation. It used to be my go-to language whenever I wanted a clicky graphical application.
https://en.wikipedia.org/wiki/Java_(programming_language)
Excel
I've tinkered a bit with formulas in Excel. It was extremely useful when I was doing bioinformatics that would fit into a 65,000-line Excel spreadsheet. I wrote overly-complex formulas so that the formula fill would do the right thing.
https://en.wikipedia.org/wiki/Microsoft_Excel#Functions
Visual Basic for Applications (Word, Excel, Access)
When Excel didn't do what I wanted quick enough, I used VBA. My most complex code was a hierarchical 2D bubble sort algorithm that I spent a few months on for an Honours project.
https://en.wikipedia.org/wiki/Visual_Basic_for_Applications
Microsoft Access
Access had a nice form designer for entering data into a binary database. I discovered it when working on genealogy data for Rangitāne, and used it for my first real bioinformatics project, analysing chlamydia transmission data.
https://en.wikipedia.org/wiki/Microsoft_Access
Warbots
It was an interesting exercise writing AI code to join in a last-robot-standing fight. Warbots had radars, lasers, smoke, and robots.
https://archive.org/details/Warbots
Turing machines
I graduated with a maths degree, doing computational logic, with a generous helping of computer science on the side. One exam required us to write Turing machine code for adding two numbers together.
https://en.wikipedia.org/wiki/Turing_machine
Register machines
I learnt Turing machines and register machines at about the same time in both computer science and maths. I think my familiarity with assembly helped for getting the hang of register machine concepts.
https://en.wikipedia.org/wiki/Counter_machine
Prolog
Prolog was a huge breath of fresh air to me, lodging into my head in a very comfortable way. It made sense! I entered "a = 3", and it would reply "FALSE" or "TRUE" depending on what the value of 'a' was.
https://en.wikipedia.org/wiki/Prolog
Haskell
I learnt Haskell at a similar time as Prolog, and got frustrated with the particular interpreter we were forced to use for our course, which wouldn't let me write a show function for a function. Lazy evaluation gave me warm fuzzy feelings.
https://en.wikipedia.org/wiki/Haskell_(programming_language)
R
I started learning #RStats near the end of my Honours year at university. That's when I realised my hundred-line sort code could be simplified into a short one-line command in R. I enjoy using R, especially when I can use '->' and '%>%'.
https://en.wikipedia.org/wiki/R_(programming_language)
Javascript
I'm not a great fan of the javascript flavour of the week / package repository, preferring to stick with local, understandable code when I write my own web pages. I will tolerate it as an end-user though. It keeps Twitter chugging along.
https://en.wikipedia.org/wiki/JavaScript
SQL (including Access, PostgreSQL, MySQL)
Knowing a bit about SQL helped a lot when translating medical codes from one database into another. I don't use it much now; I generally try to avoid using databases for anything that doesn't change much.
https://en.wikipedia.org/wiki/PostgreSQL
LaTeX (including PDFLaTeX and TeX)
My Honours and PhD theses were written in [PDF]LaTeX, and I've used it for a few research papers since then. Many people freak out over formatting hints around text; I'd been yearning for them since WordPerfect 5.1.
https://en.wikipedia.org/wiki/LaTeX
PHP
I've tinkered a tiny bit in PHP, enough to get websites to do what I want, so that I can leave them alone and forget about the language.
https://en.wikipedia.org/wiki/PHP
Perl (might as well include Raku)
My first memorable experience with Perl was hacking a friend's Perl script to load picture messages from a different website to display on a series of monitors. It's my go-to for line-by-line text processing.
https://en.wikipedia.org/wiki/Perl
Dylan
I was challenged to write some Dylan code by a friend of a friend. I didn't find it particularly interesting; it seemed like a mish-mash of Haskell and C++, and the idea of compiling into a different language turned me off.
https://en.wikipedia.org/wiki/Dylan_(programming_language)
OpenSCAD
OpenSCAD is declarative in a similar way to Prolog, but has the interesting idea that everything that is created is a representation of a 3D object. My computer science knowledge has helped in finding workarounds to some of its quirks.
https://en.wikipedia.org/wiki/OpenSCAD
MATLAB (and Octave)
The idea of ultra-fast matrix manipulation, and succinct range slicing, was pleasing to me. Less so was the idea that a programming language could be commercialised.
https://en.wikipedia.org/wiki/MATLAB
Python
Python reminded me a bit of MATLAB in its array notation; I like the simplicity of array slicing. What I don't like is the way that copy-pasted code creates unintended effects due to indentation differences. I like my braces and semicolons!
https://en.wikipedia.org/wiki/Python_(programming_language)
Maple
We learnt Maple as part of our university mathematics course, but I'd learnt it a few years earlier as part of a high-school extension course. I enjoyed watching our lecturer whip up some quick animations, and demonstrate integration solutions.
https://en.wikipedia.org/wiki/Maple_(software)
LilyPond
LilyPond is to musical engraving as LaTeX is to writing. I've written / copied a couple of songs into LilyPond format. It's a bit fiddly, sometimes requiring a bit of web searching to get the right incantation, but it makes nice pictures.
https://en.wikipedia.org/wiki/LilyPond
Logo
I enjoy graphical languages, and was pleasantly surprised to discover in high school (as part of formal teaching) that Logo could also be used for knowledge-based systems (e.g. 20-questions).
https://en.wikipedia.org/wiki/Logo_(programming_language)
Lisp
Most of my interaction with Lisp has been trying to get Emacs to work better. I get easily confused by the parentheses, and generally retreat to R or OpenSCAD, while reminding myself that I should really try to get the hang of Lisp some day.
https://en.wikipedia.org/wiki/Lisp_(programming_language)
C
I've written a couple of C programs, but most of my interaction involves modifying someone else's bioinformatics C code, which is a bit like walking on broken glass. It's mind-numbingly easy to write crashable C code without realising it.
https://en.wikipedia.org/wiki/C_(programming_language)
Brainfuck
Given that brainfuck is an esoteric / parody language, I feel fine stating that my biggest program written in it was created while playing the marble-rolling game Enigma.
https://en.wikipedia.org/wiki/Brainfuck
Enigma
Speaking of which, I might as well mention that as well. I haven't written any Enigma code from scratch, but I have tinkered a tiny bit with levels others have written, mostly to work out how I can solve them. It even has Tetris / EnigTris!
https://en.wikipedia.org/wiki/Enigma_(2007_video_game)
Visual BASIC
It's a bit different from BASIC, and a bit more complex than VBA. I've mostly tinkered with other people's code rather than written my own applications, as I prefer Java and R/Shiny... and getting scratched by a cat, for that matter.
https://en.wikipedia.org/wiki/Visual_Basic_.NET
sh (including Csh, bash and dash)
One-liner shell scripts make up the majority of my bioinformatics coding, but I've made and hacked a few longer script files. Shell scripts are all over the place on Linux systems, so it's helpful to understand them.
https://en.wikipedia.org/wiki/Bourne_shell
HyperCard
I have vague memories of a card-like language that we had to work with in high school, and I think it was probably HyperCard. Despite my enjoyment of visual languages it didn't grab my attention enough for me to remember it.
https://en.wikipedia.org/wiki/HyperCard
QuakeWorld
I played a bit of Quake Team Fortress, and noticed that it's console had some programming functionality which was almost turing complete. I can't remember what I used it for, maybe timed messages.
https://en.wikipedia.org/wiki/QuakeWorld
POV-Ray
Mostly I was peeking over the shoulder of my friend while they were writing POV-Ray scenes. I did a couple of simple things (e.g. a ball on a plane), but I never really felt a pressing need to do 3D modelling until OpenSCAD came along.
https://en.wikipedia.org/wiki/POV-Ray
Awk
Almost the entirety of my Awk experience has been in one-liners to manipulate text fields for bioinformatics. Awk works well for quick manipulation of tab-separated files.
https://en.wikipedia.org/wiki/AWK
MIPS
We learnt a bit of MIPS assembly when studying computer science at university. I liked the idea of a simple structured assembly, and have memories of enjoyment around both learning (as a student) and teaching MIPS (as a tutor).
https://en.wikipedia.org/wiki/MIPS_architecture
Scratch (and BlockSCAD)
I had a play with Scratch for my first sino:bit program, and tinkered with a Scratch-like variant for 3D modeling called BlocksCAD. I liked the idea of a visual programming language with constrained, selectable choices.
https://en.wikipedia.org/wiki/Scratch_(programming_language)
What's Missing?
The most likely reason for me to have not used a programming language is simply that I haven't needed it to solve a problem I had. I don't consider myself to be much of a collector (although I do have trouble throwing things out that others give me), and this philosophy seems to extend into programming languages as well.
Here are some languages that people might expect me to have had experience in, but I don't, yet:
2020-Jun-22
Perl
I use it, love it, and write new code in Perl when R isn't appropriate (e.g. when processing things line-by-line). I will reluctantly use python when I have to (e.g. if there's a hdf5 library that I need to use), or C++ when I have a deep concern about precisely how something is done (usually due to a need for algorithmic speed).
One of the things that people struggle with in Perl is the way that
different types of variables are represented by symbols: single values
are represented by a $ character; arrays of multiple values are
represented by a @ character; maps / dictionaries are represented by
a % character; and file handles are... inconsistently represented by
either nothing, or a * character.
The other big thing that trips people up is the concept of a default
variable. For a lot of functions, if the function argument is not
specified, it is assumed that it operates on the single value default
variable, $_. Calling functions also loads a default array variable,
@_, with the called parameters of that function, so it's fairly
typical for functions to first do stuff with the default array
variable before doing anything else.
On the other hand... I am also currently debugging the Perl code of someone else who just released a paper on a new method for single cell sequencing. It's... an interesting experience, and has helped me realise why so many people hate Perl.
Part of the issue is that the use warnings; use strict; incantation
is not very common, and it's a challenge to get people to understand
the idea that warnings and errors are helpful. The Perl interpreter
does not give much feedback (or helpful feedback) when things aren't
written "correctly", and will happily continue on doing the wrong
thing. The mixture of weird symbols and lack of warnings means that
it can be both difficult to understand what code is meant to do, and
difficult to demonstrate that the Perl interpreter does what the
programmer expects the code to do.