I ask this because so often I see people recommending tools and techniques by saying things like "this dramatically increased my productivity!" or "this sped up my workflow by an order of magnitude!"<p>How can we take such claims seriously? Most things do not improve our productivity THAT much. However, some things truly do.<p>Let me give some examples of things I think truly have "dramatically improved" my productivity.<p><pre><code> - testing, specifically integration tests
- memory safety (garbage collection, borrow checker, etc.)
- version control (git, etc.)
- closures and lambdas
- unix pipes and the idea of composing larger programs out of smaller ones
- Chrome debugger when working with javascript
- I hesitate to say "debuggers" in general because I actually think GDB didn't "dramatically" improve my productivity back when I programmed in C. And I don't even use a debugger at all when I program in python. But the Chrome debugger has numerous features I use all the time.
</code></pre>
Now let me give some examples of things that didn't really live up to the hype, in my opinion:<p><pre><code> - static type systems
- could be a small win but I don't find myself "dramatically more productive"
- TDD
- IDEs
- some IDEs are better than others, but I have yet to experience a dramatic increase in productivity simply by using a new IDE. Text editor + terminal is pretty strong tbh.
- serverless architecture
- I could see the argument for scaling concerns, but at a smaller scale I don't see a productivity boost compared to a monolith Rails/Django app
- classes and class-style OOP, inheritance
- I haven't written a class in a long time, and I haven't used inheritance in a REALLY long time. I just don't think these are huge wins in most cases. Note that I'm not even advocating for FP here. I program in a procedural style most of the time in all honesty.
</code></pre>
I'm not trying to start a flame war, so don't worry about it if you disagree with my particular examples. I was just trying to illustrate what I am getting at. I'm looking for things that TRULY increased your productivity. Not just kinda/sorta. If you are on the fence about something, it wasn't significant enough! I'm looking for things that are obvious. Things that noticeably slow you down if you didn't have it. Things that you can't live without.<p>Thanks! I look forward to seeing the replies.
Maintaining a logbook by date. This enabled me to have some chance of reestablishing state when I knew the date of some action. It also collected all my scribble, notes and references in one place meaning I only had one thing to remember when wandering about. Eventually, two with my phone but they came way later. Yeah, it wasn’t searchable but no electronic alternative worked anywhere near as well nor were as robust when traveling. A good quality engineering spiral notebook..<p>This approach worked for all operating systems and languages that I used. Planning, work notes, summaries and speculations along with the everyday detritus that showed up.
Pycharm basically covers every corners of my workflow. Their git integration is just better. And code highlighting + linting implemented in Pycharm is way better than the others, reducing bug rate by a lot. Also now I can just right click to run doctest inside the code to ensure everything is alright. Testing cannot be more easiler than this. What's great is than everything come out of the box, no extension nonsense (e.g. VScode, you can't even change the background color by yourself).