Information overload
My friends look at me like I'm crazy when I tell them that I'm trying to find an RSS/Atom feed aggregator that won't choke when I try to subscribe to over 350 feeds. I wonder what they will think when they find out that I'm also trying to keep track of:
- 466 Twitter feeds,
- 188 Facebook friends,
- 30 podcasts,
- 10 IRC channels,
- 6 Meetup groups,
- 3 instant messaging networks,
- 2 voicemail boxes, and
- 1 email address that sorts 2000 messages per day through 231 procmail rules and a spam filter.
Does that sound like information overload?
Let's put it into perspective: The ATLAS detector (part of the Large Hadron Collider) generates 25 megabytes of output every time a packet of protons crosses it, which happens 40 million times per second. That's a whopping 1,000,000 GB every second!
Today's technology can't store that much data for very long. Even assuming the availability of $100 1TB hard disk drives that could store the information fast enough (they can't) and that were 100% reliable (they aren't), ATLAS would need 1.4 million of these drives (and a budget of $144 million) per day just for this initial data storage.
That is information overload!
The scientists and engineers working on ATLAS are well aware of these physical limitations, and they deal with them by using a clever multi-stage process that ultimately figures out what's important and discards the remaining 99.99997% of the data, resulting in a data rate of 'only' 320 MB per second (see page 5 of the linked PDF).
Think about those numbers. That is the state of the art.
Yet, I struggle with a handful of RSS feeds and a couple of social networking sites! I've thought about just pruning my subscriptions and changing my email address every few years like everyone else does, but I'm too embarrassed to admit to myself that I am so far behind the state of the art.
Darnit, I'm a programmer! If the ATLAS scientists can do what they did, I should easily be able to put together something that can sift through a measly fraction of the data while throwing away much less of it, especially when I can let it work for seconds or even minutes instead of nanoseconds. It's a simple, routine automation task.
Right?
“Feeling blue? Look to hugs before religion” – G&M
A friend and fellow atheist on Facebook posted a link to this article in the "Health & Fitness" section of The Globe and Mail:
Feeling blue? Look to hugs before religion
A new study finds Canadians who get hugs 'all the time' are more likely to report better mental health, while regularly attending church did not lift respondents' spirits.
My first reaction was to think, "Oh cool! They must have taken 300 depressed people, sent 100 to church, 100 for hugs, and left the other 100 alone as a control group, while making objective measurements of the subjects' mental health..."
Except they didn't. At least, that's not what was reported. Instead, they just correlated the answers to some very subjective questions in a Statistics Canada survey.
I've taken what I think are the relevant questions and quoted them below. Try to guess which one is different from the others:
(GEN_Q02A) How satisfied are you with your life in general?
[Very satisfied/Satisfied/Neither satisfied nor dissatisfied/Dissatisfied/Very dissatisfied/Don't know/Refuse to answer](GEN_Q02B) In general, would you say your mental health is:
[excellent/very good/good/fair/poor/don't know/refuse to answer]?(SSA_Q10) How often is each of the following kinds of support available to you if you need it:
… someone who hugs you?
[None of the time/A little of the time/Some of the time/Most of the time/All of the time/Don't know/Refuse to answer](SSA_Q22B) When you needed it, how often did you receive this kind of support (in the past 12
months)?
[Almost always/Frequently/Half the time/Rarely/Never/Don't know/Refuse to answer](SPR_Q6) Not counting events such as weddings or funerals, during the past 12 months, how often did you participate in religious activities or attend religious services or meetings? Do not include special events such as weddings, funerals, baptisms, bar mitzvahs, etc.
[Once a week or more/Once a month/3 or 4 times a year/Once a year/Not at all/Don't know/Refuse to answer]
If you guessed SPR_Q6 (the last one) you're right! It's the only one that wasn't completely subjective. It only contributed to a null result:
it didn't matter whether people went to church every week or not at all – how they described their mental health didn't change.
*Yawn*
It gets better though:
But don't rule out the health benefits of being spiritual, researchers say – especially in times of stress. In the past few years, a growing collection of studies have explored rates of mental illness among the religious, many of them finding that the incidence of depression and anxiety is lower among this segment of the population. A new, two-year Australian study that tracked people hospitalized with depression found that those who expressed core religious beliefs recovered faster – and that faith had a greater influence than either medication or community support.
OK, so we have a reporter citing unnamed researchers and unnamed studies that allegedly point to "health benefits of being spiritual", but of course without a citation, effectively no verifiable evidence is given to support this.
Well then, if they can publish random assertions that conveniently support their preconceptions while providing little to no supporting evidence, so can I:
- Maybe the unnamed study's uncredited authors made methodological errors.
- Maybe the unnamed study was misquoted.
- Maybe the unnamed study was withdrawn.
- Maybe people who express core religious beliefs are so busy reading their holy books (which they've read dozens of times before) that they don't notice all the crap that passes for health science journalism.
- Maybe people who express core religious beliefs have a negative impact on the non-religious people around them, as evidenced by all those angry atheists in the world.
Without any citation, who knows? We just have to trust some reporter who is almost certainly no medical researcher. As one commenter put it:
"The headline's claim is not supported by the reported work."
PyCrypto 2.1.0 released
PyCrypto 2.1.0 has been released. This is the first stable release since I took over as maintainer of PyCrypto. It is compatible with Python 2.1 through 2.6. (Python 3.x is not yet supported.)
See this post for more information.
