Love the subtitle. Not everyone knows I’m a ghost now

Daphne did one too! youtu.be/RaveIxcQuHs

Next step is to get alllllll the tiny islands up one-by-one. I’m still missing Hawaii and Taiwan too!

Poor documentation is a whole ‘nother level of difficulty 🫠

I find reading the source code helpful when available, but that wouldn’t work for a private API like Google Drive

Historically the expectation has been to learn on the job. It’s good to become intimately familiar with all the tools you use but learning a language and 3-10 libraries in depth is a massive barrier to getting started.

Asking questions to senior devs at the same rate as googling is much slower though, as well as discouraging. It’s not bad when the google fail rate is low but it’s been getting higher and higher.

I haven’t seen significantly different results with DDG/Bing but haven’t tried Kagi.

I’ve preordered! It would be nice to see a high level comparison of Bun/Node/Deno security features

I’ll check out Snyk!

Is Bun using permission gates like Deno, or does it have additional security tools?

docs.deno.com/runtime/fund...

Outdated libraries on Stack Overflow/Medium are pretty rough but that’s at least something automation could catch. Linters could suggest newer replacements. (I think @e18e.dev is working on this?)

I’ve had much better luck with native web tech (yay MDN) but libraries are where I run into issues the most, and someone newer won’t necessarily know which is which.

I’ve also seen this be more pronounce with Python, where libraries sometimes inject global-looking functions or hidden proxies.

Medium honestly hasn’t been too bad. What I’ve run into are long blogs that have introductions like a cooking recipe (history of the library, the language, the author’s mood) then a paragraph midway with an incorrect answer.

What’s also helpful (and I need to do this more too) is to pair with a junior dev and see how they’re researching.

I think with experience you learn to dismiss unhelpful resources, even in new programming languages, but that’s hard to build up.

The trouble is it’s difficult to validate/invalidate the results when you’re learning. I’ve seen ChatGPT point someone to the wrong library and then you’re sent down the wrong site for validation.

But then sometimes the results are wrong 😓