Prince Andrew?Baron Andrew Von Fancypants! We've finally uncovered your secrets!
Some people like to rubber duck debug, except with real people.Timed it today. The two dudes in question ate up 18 minutes combined from our meeting. Retro is in a couple weeks and I'll bring it up. Feels like I'm wasting my time being there.
One dude is very senior, close to retirement.. so it's not like he's a new guy trying to work through code. But he breaks down his problems and gives all these explanations and crazy information we don't need. I literally do not care.
The other guy might be autistic, I'm pretty sure he is. He's a junior and he is always having problems and brings them up. But then there's a lot of awkward pauses and he argues with people who question what he's doing or give him advice. He's more frustrating to listen to than the other guy. He also has this habit of walking around to people's cubes with his laptop and sort of talking through his problem fishing for help, but not asking for help. I usually just offer no advice unless he asks, because it's so hard to figure out what he's actually doing.
Despite all this, the junior is pretty smart and makes a lot of good code changes. But just frustrating.
Yeah, and I understand that. I've certainly debugged problems over email before when I was younger. It's just draining to listen to.Some people like to rubber duck debug, except with real people.
18 minutes, just 2 guys.. that is way too long. Keep collecting more data, if they say "we don't take that long", bam hit them with the exact numbers =).
I'm not going to defend the shortfalls of weak typing, but can you give some examples? I've found some `assert isinstance` will fix typing ambiguity issues. The other one I commonly run into is old style print formatting where you need to know the type or hope the object implemented str() or repr(). That's why I like using .format() instead.I want to rant about my love hate relationship with Python. Now I love Python but I also absolutely hate how my company keeps wanting to use it for everything.
Python pisses me off a lot with its weak typing bullshit that constantly fucks shit up unless you have an intimate understanding of the wonky shit the interpreter does. Derp derp a dict is a dict except for when its a string for... reasons.
format() is old and busted. Fstrings is the new hotness. And by new I mean 5 years oldI'm not going to defend the shortfalls of weak typing, but can you give some examples? I've found some `assert isinstance` will fix typing ambiguity issues. The other one I commonly run into is old style print formatting where you need to know the type or hope the object implemented str() or repr(). That's why I like using .format() instead.
That's not really pythons fault though. Sounds like someone wrote some shit code.This current gripe is from an hour long debugging session I ended up doing for just what I said. There was a method in our codebase that returns a dict. This method is called in another class and then constructs some stuff.
However when the method is called in the other class it converts it to a string. This drove me nuts because the method 100% for sure returned a dict. Which can be corrected sure and you can just declare stuff as dict = {} before you call the method and all but it allows for developers to make mistakes you just can't make in other languages because of strong typing. You can do some REAL spaghetti bullshit in Python and even though its cool to do it causes problems at scale IMO.
In Python 2 my favorite one was that you could use string/byte methods on the opposite of each other because Python would implicitly change the a string to a byte or a byte to a string depending on what you wanted to do to it.
I'm not familiar with fstrings, what do they do that format doesn't? My current employer is stuck in Python 2, so we're a bit behind the times.format() is old and busted. Fstrings is the new hotness. And by new I mean 5 years old
This covers it:I'm not familiar with fstrings, what do they do that format doesn't? My current employer is stuck in Python 2, so we're a bit behind the times.