How I Taught Myself Python/Django in 8 Weekends

I think that should be qualified. Computers weren't as ubiquitous back in the 70's as they are now, nor was learning material. The TRS-80 with monitor was released in 1977 for ~$600 - about $2200 in today's dollars. I only knew of several kids with home systems.

When I took Computer Math (my high school's programming class), we had to learn programming on a piece of cardboard! It was the CARDIAC, Cardboard Illustrative Aid To Computation - http://www.simnia.com/it/cardiac/cardiac.htm.

After we became adept as manually running programs on the cardboard, we transitioned to dialing into a mainframe. Nothing says fun like dialing a rotary phone over and over until you finally made a connection. And of course, mainframe time was limited, so our BASIC and FORTRAN programs had to be written out by hand before hand. The CLI was an unforgiving beast and disconnections were frequent.

Could people teach themselves to program in the 70's? Sure, but there's really no comparison to the learning opportunities available today, and I don't think it's right to disparage anyone who blogs about their learning experiences.

If it works to get people started then it works. I honestly don't remember how I got started programming - some people rhapsodize about that first moments coding, but for me it was mostly rage at a capricious Java compiler that I felt was always either finding a creative way to do something other than when it should or throwing an error completely unrelated to what was wrong.

After I got past that point everything was gravy. I've always been kind of confused by the learning curve metaphor (If the learning curve is steep, doesn't that mean you learn faster? What's the x and what's the y in this?) but tech now is getting closer to the ideal linear curve, where additional learning and productivity are at a relatively constant proportion no matter what stage you're at rather than being a floor function until you hit a certain level.

Which is awesome.

Apparently down (is it "Hacker News-ed"?). Nyud.net-link:

http://norvig.com.nyud.net/21-days.html

I think I've read this at some point -- I've at least seen the 10 years / 10.000 hours numbers before.

I did a quick calculation with respects to martial arts practice -- and if you want to get in 10.000 hours in 10 years -- you do indeed need to practice quite a lot.

Start with a strong foundation: 2 hours 3 times a week, 48/weeks a year. Add 5 hours of training every other Saturday (or Sunday). Take seminars. Lets assume they amount to a full week of training/year (40 hours). You'd still by below 500 hours/year -- and you'd take 22 years to attain "master" level.

This actually fits quite well with the arts that still do rigorous gradings, like traditional ITF Taewkon-do, or karate and koryu styles that generally have two sets of gradings, first grading up to 4th dan, then to instructor grades -- and 5-6th dan generally being considered "full proficiency".

It also follows a somewhat simple rule of thumb from the traditional grading systems (before splitting the white belt into many colours -- at which point it is no longer white, to paraphrase Gandalf): After 1 year of (intensive) practice grade up to 1st dan, practice two years for 2nd dan, 3 for 3rd and so on -- for 21 years to 6th dan (assuming no breaks or injuries).

I guess all this is neither here nor there, but I found the correlation interesting.