"Don't quit, even when you know your stuff sucks" - Ira Glass on Storytelling | Hacker News Reader

"Don't quit, even when you know your stuff sucks" - Ira Glass on Storytelling

youtube.com

68 pointsiamwil19 comments

I normally wouldn't post stuff on here from youtube, but I was watching stuff on "This American Life", and I ran across this clip. TAL is a weekly radio show. Regardless of what you might think about the show, here, the host of the show talks about getting better at what you do, and how many people quit when they can tell their stuff sucks. I thought it applicable to builders that frequent this forum as well, as well as being in line with PG's essay about a taste for makers.

I've not watched to video, but I think of a woodworking page I read a long time ago following the paths of a power tool guy and a hand tool guy learning their craft. Through enough practice, they eventually converge on a generally common set of tools, mix of hand and power. Through the progression of skill development were milestones of

make stuff that sucks that no one wants, throw it out.

make stuff that sucks that you can use, keep it for a while.

make stuff that sucks less, keep it for longer.

make stuff and hey, someone actually wants you to make stuff for them

make stuff better

and so on...

To me, its about making mistakes and learning from them, trying not to repeat the same mistake twice. Then you are bound to get better. There was some quote I heard that if you make enough mistakes and don't repeat them, you will run out of mistakes. It was a bit satirical, but I like the general line of thinking.

I took a furniture making class and Dan, the experienced woodworker, who taught the class told me something important when I was struggling with my table. He said "Do you think master woodworkers don't make mistakes? They do, they shim* stuff up. You can't tell when it's done". Another woodworker also told me 'there's no such thing as the right way. there's only smarter, safer, faster" e.g. if you run into a problem, you work around it with a creative solution.

A shim is a small piece of wood, chipped to fit into a joint and glued in place so that the joint is snug.

yep, and mistakes that do not negatively impact functionality or form are collectively called "character."

> There was some quote I heard that if you make enough mistakes and don't repeat them, you will run out of mistakes. It was a bit satirical, but I like the general line of thinking.

There's a quote by some physicist to effect of, "An expert is someone who has made all the mistakes one can make in a chosen field."

Really interesting interview -- thanks for posting it! Just the sort of content I love to see on news.yc

A more common problem is lack of said taste - when people don't know their stuff sucks.

A sign of this is not loving other people's work. People who love to write, but not to read, produce dreck.

There are lots of people who love to program. But how many people love to read really good source code, or learn from books that teach difficult concepts in programming?

I've been thinking a lot about this lately myself. Where does one go to read "really good" source code?

Great question, I'd like to know too. Sometimes I poke around popular open source projects, like the rails source. Here's how they implemented the syntatic sugar for things like "1.month.from_now":

http://github.com/rails/rails/tree/master/activesupport/lib/...

I recently found http://refactormycode.com/ and it looks interesting for learning, though not comprehensive. Sometimes it's nice to see the "before and after".

Programming Pearls, by Jon Bentley:

http://www.amazon.com/Programming-Pearls-2nd-ACM-Press/dp/02...

Actually, I'm not sure if this is true. I recently watched a lecture (that I can't find right now) on something like "How to write a good math paper". In it, the speaker made what I thought was a wonderful observation:

Often, he sees new graduate students reviewing paper submissions to conferences or journals. These students are usually very critical, and their criticism is usually dead on-- the things they complain about are genuine defects of the papers. However, when those same students write papers they do not live up to the standards that they have set!

So, the problem may be more of what Ira Glass hints at-- people who have good taste, but are unable to apply that taste fairly to their own work.

"Knowing the path is different from walking the path" etc.

They may well not be happy with their work either, cf Ira Glass video.

His story totally reminds me of Jerry Seinfeld's "Don't Break the chain" tip.

see: http://www.dontbreakthechain.com http://lifehacker.com/software/motivation/jerry-seinfelds-pr...

His series on the Chicago radio is great. It's called "This American Life". I caught an episode of it on Showtime, the TV series is equally brilliant. It's good to see good TV like this at the time when cable's dominated by nonsense like The Hills and all the dumb reality shows.

I normally don't watch online videos, but when I saw this one was exactly "5:19" long, I knew he was talking to me :-)

01:35 "the most important possible thing you can do is...do a lot of work"

Reminds me of the time I asked Tony Robbins what made him such a great public speaker and he said, "10,000 speeches".

David Heinemeier Hansson gave me a similar response after his memorable Startup School talk.

I've written 10,000 programs and I'm still not close to being satisfied. I'll just keep plugging away.

10,000?

10,000 is a large number. For instance, 10,000 days is about 27 years, 5 months.

When you look at it that way, it seems impossible, no?

I have 2,100 original programs in one client's library right now. I have worked for them part time for 5 years. That works out to more than one program per day. I suppose that's right. There have been many days where I've written 5 to 10 new programs.

In all fairness, I count everything. Largest program --> maybe 3,000 LOC (shame on me). Smallest --> 5 LOC? Lots of utilities, functions, subroutines, and one time runs, too, I imagine.

In an interview once, I was asked how many programs I had written. I was never asked that question before (or since). I did a little math and answered 8,000. The interviewer didn't believe me. I showed him the math. I eventually got the job and it worked out well for both of us. I reminded my boss many times about that question (I'm up to 8,049.)

I first got paid for a writing a program in 1979.

I once went to a company that had in its library a program I had written for a software vendor 5 years earlier. I was so embarassed by it, I actually changed the author's name :-).

Tony Robbins giving 10,000 speeches. Many days he gives 5 to 10. I believe it.

Neat. I've had those clips bookmarked for a long time now... I rewatch every month or two. <3 Ira Glass... religious TAL listener.

I also watch this interview with Nick Cave on the creative process every once in a while: http://www.youtube.com/watch?v=ZgHatnzQ9ZM Semi-related, theme-wise.

Reply on news.ycombinator.com