The best programmers I know

Not to be snarky, but my definition of best programmers would balance these. I do spend more time than most understanding the depths of tech stacks and the breadth of potentially useful ones. At the same time I strive to make software that gets the job done with only the abstractions that pay off in a short to medium timeframe.

The trap avoid are those business impact folks that demonstrate an unwillingness to get better at actual programming, which ironically would increase their impact.

Edit: an example is fixing a problem without understanding its cause.

> Read the Reference

> Don’t Guess

I find that, when working with a new "thing," I often like to guess for about an hour or so before I really do a deep dive into the reference. Or, I'll read a stackoverflow answer or two, play around with it, and then go to reference.

Why?

Often there's a lot of context in the reference that only makes sense once I've had some hands-on time with whatever the reference is describing.

This is especially the case when learning a new language or API: I'll go through a tutorial / quickstart; "guess" at making a change; and then go back and read the reference with a better understanding of the context.

BTW: This is why I like languages and IDEs that support things like intellisense. It's great to be able to see little bits of documentation show up in my IDE to help me in my "guess" stage of learning.

>> "Don’t go to Stack Overflow, don’t ask the LLM, don’t guess, just go straight to the source. Oftentimes, it’s surprisingly accessible and well-written."

> It's a bit like math books. I dreaded reading formal math during my engineering -- always read accessible text. Got a little better in my master's and could read demse chapters which got to the point quickly. At least now I can appreciate why people write terse references, even Tutte books.

I don't think that's what he means by the advice. I think it's more about systematic knowledge vs. fragmented knowledge. Someone who "learns" through an LLM or Stack Overflow is not going to have the overall knowledge of the tool to be able to reason what's available, so will tend to use it in very stereotyped ways and do things in harder ways because they don't know what's possible. You can still get that systematic knowledge through an accessible text.

same. It is so hard to write good documentation, especially when the author is also the designer. There is so much context that is easy to take for granted and thus leave out.

>> Read the Reference

>> Don’t Guess

> I find that, when working with a new "thing," I often like to guess for about an hour or so before I really do a deep dive into the reference. Or, I'll read a stackoverflow answer or two, play around with it, and then go to reference.

I think that's fair. I've definitely seen "not best" programmers only guess and only read stackoverflow, over and over, forever, and never read the reference. They have no idea what's going on and just spin making a mess in until something sticks. I kinda read that item as a response to people like that.