Comments for Sliding up the banister

Comment on Intermission: Google Squared by Robin Barooah

Robin Barooah — Thu, 04 Jun 2009 16:38:25 +0000

The results are more entertaining if you search for “programming language” instead of “programming languages”. The ‘image’ and ‘death’ columns contain interesting values.

Comment on Intermission: Google Squared by James Iry

James Iry — Thu, 04 Jun 2009 15:49:15 +0000

I was fascinated to learn the Java’s paradigm is “datagram packet.”

Comment on Jared Richardson at Memphis JUG by Matt Stine

Matt Stine — Sat, 25 Apr 2009 17:01:17 +0000

Joel – thanks for the encouragement on the JUG. I have been very excited to see how much things have developed in the past few months.

Your write up of Jared’s talk is priceless…complete with a nice typo for authenticity.

Comment on Words to avoid: “intuitive” by joelneely

joelneely — Sat, 14 Feb 2009 04:46:22 +0000

@Russ: I recall a sig line that read, “vi is very user-friendly; it’s just picky about who it makes friends with”.

Comment on Words to avoid: “intuitive” by Russ

Russ — Fri, 13 Feb 2009 17:53:08 +0000

There’s a marked difference between something being ‘intuitive’ to a complete newcomer, and being intuitive to someone with some prior experience. Think of it as a property of a system’s coherence. Vim, for instance, is something I found hugely intuitive once I’d got the hang of it. As a n00b it seemed completely impenetrable, but once I’d grasped the concept of motions, text objects, registers etc the whole thing started to come together and previously unknown editing operations became ‘intuitive’ because I knew how to construct complex commands by chaining together basic operations.

I think ‘user-friendly’ is a similarly misused phrase. Everyone assumes ‘user-friendly’ is synonymous with ‘my Aunt Tillie could use it’. But ‘user-friendly’ is a subjective term, and I’m not my Aunt Tillie. When I’m working in linux I use xmonad as my WM because I find myself immensely productive in it. Xmonad is user-friendly to me, whilst AT THE SAME TIME being impossibly opaque to other people.

Comment on Design by proof by joelneely

joelneely — Sat, 13 Sep 2008 11:56:45 +0000

@wrf3: Nice work, Bob. And I’m enjoying the coincidence that 32 years ago is when Dijkstra’s A Discipline of Programming was published. It’s a tad more work to solve a suitably generalized problem, but the return on investment is SO much greater. Unfortunately, that’s a lesson d that even the designers of some of our current development tools have yet to learn!

Comment on Design by proof by wrf3

wrf3 — Sat, 13 Sep 2008 03:45:59 +0000

Some 32 years ago, I worked out the general form for the coefficients for the sum i^k. We both used the same technique. If you’re interested, see http://stablecross.com/files/222ff3ee6fc7f18ca9232db424deb550-0.html and http://stablecross.com/powersum/coefficients.html.

Comment on The burden of FP, part 3 by joelneely

joelneely — Sun, 07 Sep 2008 13:11:18 +0000

@clinton: If I understand your proposed scheme, it has a bug in that duplicates (e.g. when using multiples of 3 and 5, multiples of 15) get accumulated twice.

Comment on The burden of FP, part 3 by clinton

clinton — Sun, 07 Sep 2008 10:19:38 +0000

FIX (sorry, less than symbol didn’t show up, i think cuz no html tags)
what about this for a scheme

A –> [multiples (a, 2a, ...)] –> [less than n] –> [sum] –> Z (warp bcuz cant do angled lines)

B –> [multiples (b, 2b, ...)] —> [less than n] –> [sum] –> Z (warp bcuz cant do angled lines)

Z –> [sum] ——–>

i think it avoids merging and also keeps it functional rather than sequential.

Comment on The burden of FP, part 3 by OJ

OJ — Thu, 24 Jul 2008 03:10:19 +0000

This is similar to the extra answer I gave in my blog post which looks like this:

sum $ [3,6..999] ++ [5,20..999] ++ [10,25..999]

Performance-wise I think it’s a definite win. Though I haven’t verified it by running it through a profiler

Cheers!