3 ms·
There are a number of problems with Boz's posts. 1) they lack nuance, 2) they often don't reflect the reality of those in the field in his own back garden. I
by KaiserPro 3y ago
There are a number of problems with Boz's posts.
1) they lack nuance,
2) they often don't reflect the reality of those in the field in his own back garden.
In boot camp, I was expressly told that I shouldn't message people for help. Some smug ponytailed twat stood up and said "If I'm messaging you, I can't spend that time improving the system" The reason why he was getting messaged so much is because the code he wrote was "clever" old and entirely uncommented. The wiki was out of date.
Yet, he didn't take it as a signal that his docs/onboarding experience was shite.
So to him, this post basically says: make more code.
To compound the issue, if you're an intern or a junior, you are penalised for asking too many questions. This is at odds with the core message.
- naillo 3y agoI agree with most of what you say. But uncommented code isn't necessarily a bad thing. You can often make it self documenting through long and descriptive variables names and function names.
- KaiserPro 3y agoI mean yes, for maintaining it. But. I just want to use a "message queue"(or what ever), the one I'm trying to use has the semantics I need for a project. I just want to use it, not learn to maintain it. I don't want to have to spend weeks reverse engineering the tests to figure out how to make it work, I just want to build with it. But, how do I know which Queue system to use? does it have the right semantics? what happens when you fill the queue, does it pause the producer or emit an exception? who knows, just look at the code! (this makes selection of systems really hard, unless you have the sacred knowledge) And thats the problem. Meta might have the world's best scaling primitives, but unless you can find the right person to ask, you'll never really be able to figure out _how_ to use it in your project in any reasonable timeframe