Live data from Hacker News

The worst tool for the job

johndcook.com

21–30 of 196 posts

Re: The worst tool for the job

#21
post #2

What a great principle. One of those things I've felt for a long time but never express so well as John did. This idea extends to many other things, like... sports equipment. Beginner runners/bikers/skateboarders/whatever will not really feel the difference between equipment that cost $100 and $1,000, so get the cheap one and use it to (its) death. If you're still into the sport by then, go ahead and get the nicer on…

I found a pretty big difference between a $20 tennis racket and a $60 racket. For one thing, any tennis racket is going to get loose if you play with it long enough. It cost more than $20 to redo the strings on a $20 tennis racket and $20 racket is going to send balls in totally the wrong direction in the next season.

I guess the point is that if you find redoing the strings of your $20 racket more than once or twice, it means you're really using it and that warrants buying the expensive one.

The idea is to avoid buying the expensive racket before you even know if you'll like playing tennis at all.

Re: The worst tool for the job

#22
post #2

What a great principle. One of those things I've felt for a long time but never express so well as John did. This idea extends to many other things, like... sports equipment. Beginner runners/bikers/skateboarders/whatever will not really feel the difference between equipment that cost $100 and $1,000, so get the cheap one and use it to (its) death. If you're still into the sport by then, go ahead and get the nicer on…

Sometimes you don't know what the minimum viable is because you are a beginner and don't know yourself well in that field.

For example, I started running with shoes that were somewhat fine for beginners when I would double-check online, but happened to have limitations that affected people who practice a lot, or people who don't practice properly, or in my case undiagnosed issues (hypermobility of the knee). In my case getting those shoes with too much shock absorption was actually dangerous because it prevented the feedback loop of auto-correction from being built. This led to some habits that I didn't notice were bad and later on led to a pain in my knee that has, despite my best efforts and much rest over the last 8 years, entirely prevented me from the freedom of running.

Another example I've encountered is with the guitar: the first one I picked up was this entry-entry-level thing that was viable for my budget but hindered my interest in practice because I didn't just encounter the physical pain you encounter when learning the guitar, but also the pain you get from badly adjusted guitars with high action and uneven frets and fairly crappy wood. I gave up on it a few times. But I eventually tried again and made the hypothesis that the hindrance was the guitar, and after testing that hypothesis I committed to buying a better one which enabled me to actually learn things that flat out seemed impossible on a crappy piece of wood. Years later, I'm far away from all my good guitars and I buy this entry-level thing because I miss playing.. and it's been hanging there mostly untouched because of how bad it just feels.

I think it's much smarter to spend some time testing out things with other people's help and to borrow stuff for a while whenever possible, but on average to commit to something above entry-level if you're trying to be serious about anything. You'll save yourself a decent amount of time, you can give yourself a progress curve that looks much smoother, and when comes the time to truly upgrade you'll have something solid to use as a base to look at higher quality tools.

edit: notwithstanding the fact that in many cases you have zero resell-value for physical tools that are entry-level, while something average will at least go for a reasonable percentage of the original price

Re: The worst tool for the job

#23
I play bass, and my advice for people who are new to an instrument is to invest a bit of money. (For bass guitar, this means spending $250-$350 for your first guitar).

You can buy a bass for under $100, but many of these guitars cannot be set up properly. The strings sit really high, so you need to apply a lot of pressure to play (the alternative being that they buzz against the frets).

Learning on an instrument like that adds a lot of frustration. What's worse, it can teach you a lot of bad habits, like over-gripping.

Re: The worst tool for the job

#24
post #18
post #16

I'm not happy with the premise that 'best' means 'most expensive' as i'm pretty sure anyone involved with software will undoubtably know - Oracle anyone? :)

There's a definite link between aggressive sales and how terrible a product is.

The shittier the product the more agressive sales people are required to sell it?

Re: The worst tool for the job

#25
Interestingly, no one asked me what software I used when I created diagrams in the vector graphics app sketch[0], but once I started using ascii diagrams using monodraw[1], everyone wanted to know what I was using.

I find the aesthetic extremely "precious;" that's why I do it. Affectations are memorable.

0: https://www.sketch.com/ 1: https://monodraw.helftone.com/

Re: The worst tool for the job

#27
As a data scientist I interpret this as : if you can do it in SQL, do it in SQL. Once you really can't improve your, say, recommendation engine with SQL move to ML/DL.

And sometimes it is better not to know anything about deep learning because the tech debt is not worth it it the medium term.

Re: The worst tool for the job

#29
post #2

What a great principle. One of those things I've felt for a long time but never express so well as John did. This idea extends to many other things, like... sports equipment. Beginner runners/bikers/skateboarders/whatever will not really feel the difference between equipment that cost $100 and $1,000, so get the cheap one and use it to (its) death. If you're still into the sport by then, go ahead and get the nicer on…

It's also related to the 'rule of least power': Choose the least powerful computer language suitable for a given purpose.

https://en.wikipedia.org/wiki/Rule_of_least_power

Re: The worst tool for the job

#30
I agree that an ASCII diagram helps concentrate on the content by making the attention to the irrelevant impossible.

But it's way clumsy and hard to edit when you need to change the text and move the boxes because of that, making it... not the best tool for the job.

I would take Graphviz [1], or went to https://plantuml.com/ and drew a nicer diagram while using fewer characters. Also, a much more easy to edit one.

So no, don't take "the worst tool for the job" unless you're masochistic.

Instead, realize what the job is really about, what is important and what is not, and then take the best tool for that important part, which would free you from ever having to deal with the irrelevant details.

[1]: https://graphviz.org/Gallery/directed/hello.html

Post reply on HN