Disqus - Latest Comments for llimllib

Re: Scott's Blarney :: 091024a.html

llimllib — Sun, 25 Oct 2009 10:56:34 -0000

I am 100% anti-vaccine fear mongering, and that is not what this is, so let's start by dispensing with the name-calling.

To make my biases plain, I am not the author, I am the opposite of scared of needles, and I am related to several traditional health professionals. Traditional vaccination fear-mongering (mercury, etc.) makes me extremely angry. Fucking pertussis is coming back because of those assholes.

What the author is doing here is positing that the 86% number you're giving is not worthwhile due to two factors. First, the overall incidence of the flu does not decrease even when people are successfully immunized from one particular strain of virus. Second, that cohort bias makes that number untrustworthy.

I have not read any of the studies he points to, so I'll only update my probabilities a bit before I do, but it's worth considering whether this particular line of reasoning is valid without lumping it into general anti-vaccination fear mongering.

Re: Brian Sloane — asciirun: PhotoSketch: “We present a system that...

llimllib — Tue, 06 Oct 2009 10:24:50 -0000

I was inspired by this to go download the original paper about the GrabCut algorithm; I may try to implement it for fun.

Re: SIGUSR2 > Python Worlds

llimllib — Thu, 01 Oct 2009 20:49:17 -0000

I'm pretty sure that all this dodgy with-hacking indicates a real need for a new language construct.

That's not an insult, by the way, super neat hack.

Re: rodeo dog - Today I learned a new vocab word, guilloché, and...

llimllib — Fri, 25 Sep 2009 10:45:57 -0000

There's a fun guilloche toy here: http://www.subblue.com/projects/guilloche

Re: Canvas Tutorial - The Mouse

llimllib — Mon, 10 Aug 2009 09:58:44 -0000

I think I'd rather keep all the logic in the draw() function. In fact, if I went back and did it again, I'd probably just save the mouse's x position in the onMouseMove function and move the paddle in the draw() function, instead of in the onMouseMove function like I've done.

It's mainly an aesthetic decision.

Re: Canvas Tutorial - Library: an Interlude

llimllib — Mon, 03 Aug 2009 13:51:43 -0000

Yeah... I don't know how I missed that? I looked at my code today when you first commented and moving those into the init function was immediately obvious. Will fix.

Re: Canvas Tutorial - Bounce

llimllib — Thu, 09 Jul 2009 17:33:47 -0000

Check the code in Finishing Touches :)

Re: Kamikaze Marketing

llimllib — Sat, 04 Jul 2009 21:44:22 -0000

nit: replace "principal" with "principle".

Re: Canvas Tutorial - Library: an Interlude

llimllib — Thu, 28 May 2009 20:16:04 -0000

On what browser? You mean the $("#canvas").width() and $("#canvas").height() calls return undefined?

Re: Facebook's latest raise to cash out employees is disgusting

llimllib — Sun, 17 May 2009 01:55:50 -0000

I agree with your general thesis - it's silly for Facebook to raise cash to hand out to these developers (if indeed that's what they're doing), but I think your article makes its claim in histrionic fashion with irrelevant statements.

> Tell that to some family who ran their local Chevy dealer for the last 20 years who just got news this week that they're getting the ax.

This is a totally irrelevant statement designed to appeal to emotions instead of reason. My father-in-law and his father have run a Chevy dealership for the last 70 years, and were closed this week; if you want to discuss that situation and why it's different, that's fine, but I believe that if you think about it honestly you'll see why for yourself.

> I'm sorry, but this is exactly what's wrong with this country and why we got here in this economic mess--people looking for a handout before actually creating any real value.

Well, no, not in any proximal sense at least. This is another statement designed to elicit a "yeah, you tell 'em!" that doesn't actually make sense.

What I'd like you to do to prove me wrong is to give me a theory of how "people looking for a handout before actually creating any real value" are the (exact!) root cause of the US's economic troubles. I don't think you can do it without stretching the phrase to meaninglessness.

My opinion? Facebook sold shares of its company to investors who bought them of their own free will, and it wants to reward some of its employees. It makes *me* not want to invest in a company that will do so without making some cash first, but I don't see a reason to get all hysterical about it and try to blame the economic collapse on this type of behavior.

It's this type of shoddy, emotional thinking that makes the *actual* problems of our economy much harder to discuss.

Re: There are many like it But this one is mine. - When they finally arrived at Aqaba, Lawrence’s...

llimllib — Tue, 05 May 2009 13:34:36 -0000

fact.

Re: There are many like it But this one is mine. - When they finally arrived at Aqaba, Lawrence’s...

llimllib — Tue, 05 May 2009 12:12:04 -0000

Although I agree with you about his overall point, I think he definitely pushes the basketball thing too far. His primary example is of *12-year old girls*. OF COURSE 12-year old girls can't beat a super-hard press. And his opponents absolutely are correct when they say that the game should be (at least somewhat) about learning skills at that age - not about winning at all costs.

What *really* gets my goat about the article, though, is that he doesn't talk about the risk/reward tradeoff that occurs when you try a radical David strategy, or the idea that equilibria exists.

First off, playing as a David is an extremely risky proposition - basically you gain a shot at winning in exchange for a large chance of getting killed. Gladwell doesn't mention this at all.

Second, not everything is a war or a zero-sum game! Basketball is, but business isn't. Playing as a David in the business world often means you'll give up opportunities to establish a positive equilibria. Lowe's co-exists with Home Depot - they didn't need to kill the giant, they established a competitive but mutually beneficial arrangement. When you try a David strategy, you explicitly attack your enemy and run the risk of killing *both* your businesses.

I might blog about this, I definitely have things to say about this article.

Re: There are many like it But this one is mine. - At many blogs (Sullivan, Yglesias, DeLong, among...

llimllib — Tue, 28 Apr 2009 10:05:03 -0000

It's just hard for me to swallow the idea of living in a country that's not governed by the rule of law, and feels like we have a chance to tray and make a stand on it right now. If we don't prosecute people who knowingly ordered torture, what hope is there for ever regaining the rule of law? Should I just concede that we live in a country whose politicians are above the law?

Re: Canvas Tutorial - Introduction

llimllib — Mon, 27 Apr 2009 11:18:20 -0000

Others have suggested it, and I remain unopposed to using explorercanvas. However, I do not have Windows, and so cannot test it, so I don't want to post it up here.

If you'd like to modify these pages and test them with explorercanvas for me, I'd welcome the work - you can check it all out at http://github.com/llimllib/refresh-canvas/tree/... .

Re: Canvas Tutorial - Draw a Circle

llimllib — Mon, 20 Apr 2009 11:35:02 -0000

What browser are you using? The error means that $('#canvas') is returning null, so that the index [0] doesn't exist.

Re: Pycon: Unladen-Swallow

llimllib — Thu, 26 Mar 2009 14:56:57 -0000

They say on that page that the times are calculated as (new-old)/new, so (.409 - 1.023) / .409 = -1.5 -> 150%

I agree that this is silly; .409 is 2.5 times faster than 1.023, not 1.5 times.

Re: Shooting with the oppposite eye

llimllib — Tue, 24 Feb 2009 10:29:25 -0000

My left eye is only half as good as my right one, so my nice zuiko lens would look like a holga all of a sudden... I'll have to try it.

Re: Shooting with the oppposite eye

llimllib — Tue, 24 Feb 2009 10:29:25 -0000

My left eye is only half as good as my right one, so my nice zuiko lens would look like a holga all of a sudden... I'll have to try it.

Re: Canvas Tutorial - Introduction

llimllib — Mon, 23 Feb 2009 23:09:17 -0000

I alerted the disqus guys to the fact that canvastutorial/ and canvastutorial/index.html display different comments, and didn't receive any reply.

I'll look at the contents-off-the-screen issue, that is annoying too.

thanks for the reports.

Re: My Git Workflow

llimllib — Tue, 20 Jan 2009 11:04:57 -0000

I actually didn't even notice that you used that too :)

Re: My Git Workflow

llimllib — Tue, 20 Jan 2009 10:56:13 -0000

My workflow has one difference: I hate stash. It feels too "magic" for me; I'll use it if I expect to have a change stashed for no longer than 5 minutes, but not otherwise, because it hides the change too well and I've forgotten them several times, with painful results.

Instead, even when I'm working on my solo github projects, I'll work in branches (why work on master? It's so easy to branch). Then, if I need to work on a bug, I'll just git commit -a -m "stash" on the topic branch I'm working on, git co master && git co -b <bug-fix-branch>.

When I want to merge the topic branch back into master, I use "git merge --squash <topic-branch>" so that I can write a good commit message and nobody sees my "stash"es.

Re: My Git Workflow

llimllib — Tue, 20 Jan 2009 10:43:18 -0000

> that SSP workflow looks daunting

Really, all the stuff after "run the tests" is usually really quick. In my head it's only 3 conceptual steps:

1) make sure master is up to date: git co master && git pull
2) make sure our changes don't conflict with what happened in master since I checked out: git co <topic-branch> && git rebase master
3) merge our changes in: git co master && git merge <topic-branch>

The only way it becomes a problem is if you let your topic branch go too long without rebasing - then step 2 can be painful as you have to decide whether or not to merge each change, or use a merge strategy, etc etc.

(edit: I use git aliases so that git co == git checkout; see http://git.or.cz/gitwiki/Aliases )

Re: between scylla and charybdis - Anyone else up for a walk through Greenmount West...

llimllib — Tue, 20 Jan 2009 01:52:57 -0000

I am. This weekend? (I'm llimllib on flickr too or bill.mill@gmail.com)

Re: Canvas Tutorial - Coda

llimllib — Wed, 07 Jan 2009 09:57:33 -0000

rahul, I would change the condition of the onMouseMove function from:

if (evt.pageX > canvasMinX && evt.pageX < canvasMaxX) {

to:

if (evt.pageX > canvasMinX && evt.pageX < canvasMaxX &&
evt.pageY > canvasTop && evt.pageY < canvasTop + HEIGHT) {

where canvasTop was initialized to 0 in the variables list at the top of the library, and set to its final value by a line like this in init():

canvasTop = $("#canvas").offset().top;

Re: The question is wrong

llimllib — Sat, 03 Jan 2009 18:12:24 -0000

Paul,

That was my comment on the post where he *asked* the question, not where he answered it:

http://www.codinghorror.com/blog/archives/00120...

unfortunately, there's no permalinks on the comments there - grep for "bill mill".