Comment by gizmo686 | Hacker News Reader

Comment by gizmo686 | Hacker News Reader

Parent Full thread

I'd say there are two different types of "obsession" here. One is the attitude that that you need to use the best tools possible and by doing that you will get a good result. The other type is old fashion geakery. If you look at Adam, he is two things: a maker, and a tool geek. These aspects certainly intersect, but when he is making things, you do not see him reach for the most sophisticated tool in his toolbox. A great example of this is this video [0] where he geeks outs over his new measurement tool: guage blocks (and corresponding guage). The thing is, I don't think I have ever seen him do a build where he uses them, because they are simply beyond what his work demands.

I'm a similar way with programming languages. I'm a PL geek, and enjoy geeking out about the fine details of various programming languages, and pointing out flaws in langauges. But, at the end of the day, when it comes time to pick a language for a project, I typically reach for a combination of Java/C++/python/m4. There is nothing to geek out about those languages (except for how bad they are), but pragmatically speaking, they are typically the best tool available for very boring reasons.

[0] https://www.youtube.com/watch?v=qE7dYhpI_bI

An overlap of the two obsessions: recreate something historical with only the tools available at the time.

Clickspring's ongoing recreation of the ancient Greek Antikythera mechanism, starting by recreating many of the original tools likely used to build it historically (and contributing to academic research on the details of the system as he goes along!), is an incredible example of this (and altogether an amazing channel): https://www.youtube.com/playlist?list=PLZioPDnFPNsHnyxfygxA0...

You might be interested in https://www.guedelon.fr/fr

Went there a decade ago, it was super interesting to discuss with the people over there!

> Java/C++/python/m4

One of these things is not like the other. :) I'd love to learn why you included m4 in that list.

Because I've ended up using it in almost every project I'm involved in. By all rights, m4 should not be a good tool for anything. It has no type system, no syntax checking, no debugger, no namespaces, no real notion of atomic literals. But, it just kind of works for a bunch of situations. Basically any text based format you have can utilize m4 as a pre-processor.

Typically it comes up during integration work, where I am dealing with various technologies from various vendors for various clients in a single project. M4 ends up being a glue language for "compiling" configuration files. Where possible I prefer to use a more proper templating engine like python mako, but those are not always a good fit.

I've also used M4 as a way of implementing DSLs, which is great as long as you are fine with your DSL inheriting all of m4's problems. I've also used it to help write bindings and other highly repetitive files.

M4 seems to fit that niche in between gluing strings together and full blown Python string templating that works in a surprising amount of situations. I use it a lot, especially when templating config files in containers.

It’s the best size of machine screw

Adam Savage is also creating content, and you have to nod to the people most likely to be consuming content at times.

> There is nothing to geek out about those languages (except for how bad they are)...

I would go so far as to say that if someone doesn't at least somewhat hate the tools they are using, they probably don't know much about either the tools or how to use them ;P... regardless, at the end of the day, we still use our broken tools anyway; and so like, to me, if I see someone going on and on about how amazing their tool is, I immediately classify them as an amateur: they will eventually hit the limits and become sufficiently jaded to stop praising anything!

Meanwhile if you look at someone like Ken Parker ( https://kenparkerarchtops.com/archtoppery#new-page-1 ) you see someone who does obsess over the best tools _and_ the best process.

Now, admittedly, Ken Parker is a lightning rod for disagreement, but you can't deny that he makes a well-made product.

For the uninitiated, could you please tell me more about him being a lightning rod for disagreement? I hadn’t heard of him before.

I had not heard of him either.

it's a great point and looking at various YouTube pros I think it's necessary to add a one more type of "obsession" to the list - a commercial one. Thinking about the way YT works a lot of geekery is either related to attracting sponsors or appealing to audience taste to benefit from the ads.

Reply on news.ycombinator.com