Disqus - Latest Comments for thakker

Re: Visual Information Theory

thakker — Sun, 25 Oct 2015 11:42:19 -0000

Nice article, but I'm not sure about one point in the optimality proof. When considering the cost/benefit of "investing" espilon in 'a', you say that the costs are now p(a)+epsilon and p(b)-epsilon, but "the benefits are still the same". Shouldn't the benefits be slightly more (for a) and less (for b)?

Re: 120: Rapidfire 32

thakker — Tue, 24 Jun 2014 08:28:14 -0000

Oh, duh, good point... didn't even think about backgrounds.

Re: 120: Rapidfire 32

thakker — Sun, 22 Jun 2014 12:49:01 -0000

Re: the inseparability of box-shadow "sub-properties"

Could it have to do with the fact that you can apply multiple box shadows?

Haven't used it much, but my understanding is that box-shadow is really a *list* of shadows, *each* of which has x- and y- offsets, blur, color, etc. So, you could have something like:

box-shadow: 2px 2px 2px #000, -2px -2px 2px #0cc;

In that case, wouldn't something like "box-shadow-color: purple" be ambiguous?

Re: Inverse trigonometric functions with Sass

thakker — Mon, 02 Jun 2014 09:02:44 -0000

Wow. CSS preprocessing is really not where I'd have expected to find an application of Taylor series!

This is totally delightful. I'm definitely gonna have to dream up some excuse to use trig in sass now.

Re: vitalsigns

thakker — Thu, 29 May 2014 15:15:24 -0000

Currently, this is just a starting point -- something (I hope) on which it might be possible to build a useful app...

Any ideas?

Re: Anand Thakker

thakker — Fri, 14 Aug 2009 01:52:21 -0000

The quote is from a comment on this post: http://www.feministing.com/...

Re: Anand Thakker

thakker — Sat, 06 Jun 2009 00:21:38 -0000

A colleague of mine has a connection with a farm, and has set up a delivery at our school. It's sweet :-).

Re: Anand Thakker

thakker — Mon, 20 Apr 2009 22:09:24 -0000

Lol. That video is hilarious.

Re: We're Bad at Math

thakker — Sun, 29 Mar 2009 19:26:21 -0000

Yeah, I was wrong about predictability being isomorphic to computability. It's conceivable that the person doing the prediction is only capable of it because she can make rational judgments herself, which a computer (probably) can't.

I had another thought about the predictability of rationality, but I'm not sure what to do with it: prediction is itself a rational action, performed by an agent. As such, when the prediction is made about another person, it actually becomes part of an interaction between the two. Doesn't that cause problems for the soundness of the predictions?

Re: We're Bad at Math

thakker — Sat, 28 Mar 2009 16:12:06 -0000

It sounds to me like you're saying that rationality implies predictability, and predictability implies rationality.

Predictability implies rationality:
I'm not sure I see this: couldn't behavioral economics perfectly predict how (irrational?) cockroaches behave with respect to their "economic" resources?

Rationality implies predictability:
To the extent that we could come up with "rules for how people will behave", it seems to me that this set of rules would basically be isomorphic to a computer-like logic.

Am I missing something?

Re: We're Bad at Math

thakker — Sat, 28 Mar 2009 14:56:08 -0000

What do you think would be the results/consequences a) if so, b) if not?

Re: We're Bad at Math

thakker — Sat, 28 Mar 2009 14:40:14 -0000

I completely agree.

In my totally unclearly-stated comment about rationality being "second order", I was really trying to extend your point. You said, "we often pretend that we think in the same way that computers operate. I don't think this is true at all."

I think, going even further, that we often pretend that we're fundamentally rational, and I don't think *that* is true either.

Re: We're Bad at Math

thakker — Sat, 28 Mar 2009 08:59:24 -0000

Well, no, definitely not. That's not what I meant to be doing....

Re: We're Bad at Math

thakker — Sat, 28 Mar 2009 00:50:25 -0000

Lol! Your point, then, is a perfect example of gross misuse of analogy, is it not? :-)

But more seriously: yeah, I agree with your overall point... and I feel like it's a problem more generally, too, with models of the mind that assume that rationality is more fundamental to our processes of thinking than it really is. Why is it so hard to consider rationality a second-order thing?

Re: Anand Thakker

thakker — Tue, 17 Mar 2009 15:21:37 -0000

Your office laughed at using computers during tests? Huh... I wonder if their ancestors laughed when some radical a couple thousand years ago suggested letting students use paper to answer questions.

Re: Why is it increasing?

thakker — Sat, 28 Feb 2009 12:09:52 -0000

Thanks! It was fun to make :-)

Re: Object Orientation is a Joke; Avoid it.

thakker — Fri, 23 Jan 2009 20:15:08 -0000

Okay, I haven't quite fully thought this one out, but:

1... But maybe the web *does* represent a new, fundamental shift beyond merely the dust settling. Here's the before/after.

Before: Object Orientation, because it's all about building flexible, but semi-static systems to manage dynamically flowing information.

After: Functional Programming, because with web now maturing into a system where *interaction* is the primary unit, the distinction between information (changeable) and the processing system (less changeable) is dissolving. Sure, the available fields for a Facebook profile affect what user data is... but more and more, the user data--in particular user *interactions* which have now become data--then feeds back and affects the information management system that is Facebook. The upshot? The very data-transforming processes themselves have to be as dynamic as the data: which means functions need to be first-class data types.

Re: Object Orientation is a Joke; Avoid it.

thakker — Fri, 23 Jan 2009 17:21:35 -0000

Let me preface this: as you well know, I fully agree with you. With that said...

I have three questions for your next post on this topic, relating to each of your three ways that OO is broken:

1. If anyone besides the converted is going to buy this point, it's probably important to give it a little more historical context. Object orientation was a culminating triumph in an effort to address two serious, related problems in software design: the need for data-driven design ("modeling"), and the need for flexible modularity in code. As I see it, these needs coincided with the growing awareness that software was LESS about algorithms than we used to think, and MORE about information management (hence, data modeling) and system-building (hence, modularity). In that context, OO responded well by taking the emphasis away from overly algorithmic / process-based thinking. So my question is this: what is shifting -- or should be shifting -- in our understanding of what software is about?

2. There seems to be an important subtlety buried in your phrasing: "stateful code". It's not the *mechanics* of statefulness you're objecting to, but rather state as a first-order programming concept, i.e., state as a way of thinking about how to make a computer do something. So here's my question: sometimes, thinking in terms of state variables HELPS me think about what I'm doing, because it's a more direct/natural translation of how I think about the real-world problem. Doesn't that make it worthwhile, as long as it's not my only option?

3. Does this imply certain limits on the extent to which modular design can be used as a means to distribute labor in the production of software? (Mostly separate, of course, from the use of modular design *within* an organization to create a rational, understandable architecture.)

Re: Anand Thakker

thakker — Mon, 29 Sep 2008 17:39:52 -0000

Relish the aporia, my friend.

Re: Anand Thakker

thakker — Tue, 15 Jul 2008 08:47:41 -0000

Oh... that was supposed to be sarcastic....

Re: Anand Thakker - Each of us is several

thakker — Sat, 12 Jul 2008 20:57:58 -0000

Thanks :-).

Ever read any John Dewey, or the other Pragmatists?

Re: WashingtRon, City of the Future - And the meetup is....

thakker — Sat, 24 May 2008 16:27:58 -0000

If I'm not down in NC that weekend, I'll be there.

Re: Washingtron, City of the Future

thakker — Sat, 03 May 2008 20:33:19 -0000

I damn near got teared up that something like this exists.

Heart.

Re: Anand Thakker

thakker — Tue, 29 Apr 2008 16:57:57 -0000

Hehe.

Re: Anand Thakker

thakker — Thu, 10 Apr 2008 12:30:09 -0000

Argh. I was being too reactive. Sorry: I fucked up. (heh)

Here's the deal: I'm just tired of the fact that, because of the way *certain* men behave, many people walk around assuming the worst (or, at the very least, having their guard up) about all of us. I'm NOT an asshole/meathead, and it gets old being surrounded by the dominant effect of those who are.

I have no beef with humor, and I think political correctness is absurd.

If the "truth" you found humorous was that *men* fuck up a lot in relationships (moreso than women)-- well, I think that actually *is* a popular view, and most people wouldn't consider it un-P.C.

If the "truth" you found humorous was that "sorry seems to be the hardest word"... well, I completely agree. But your implication that it's harder for men than women was unfair.

Try hearing the following in both a male and female voice: "Why can't s/he ever admit s/he's WRONG!?" I think it plays equally well in both.