- The reason is, when I was a newbie in the world of ML, I had a laptop worst than a potato laptop. So, I had to preserve as much resources as I can for actually ML tasks. That forced me to keep things simple & lightweight.
- At that time I tried to use PyCharm, VS Code, etc. but all of them resulted in consuming a lot of resources and very little left for actual model to run.
- Fast forward to today, I now have a decent laptop for coding & a good workstation for my ML related workflows, but I still chose to use only Sublime+Terminal for all my tasks
- To be honest, setting up Conda to work with PyCharm & VS code is sometimes pain in the a*. Some of my students still come to me with problems of conda and other env management tools not working properly with VS code & PyCharm.
- The thing is I like to keep things simple and lightweight. In my experience, Sublime is an excellent code editor with excellent package support. I find running scripts, managing envs, docker, DBs, etc, simpler and quicker directly from Terminal...
- I do not understand why people use VS code or PyCharm if they have to take additional steps to make their tools work with it, let's say Conda (environment management tool). Whereas same task can be done with just Vanilla terminal without any hassle.
So Am I insane for using only Sublime+Terminal in my dev environment while entire world is using various(God knows how many) tools for development. Or is it just a developer inside who got used to keeping everything so optimised at each level, that anything that consumes resources unnecessarily makes my wanna avoid using it...
- I am really curious to know what tools you guys are using in your dev environment? And what's your reason for that setup?
