August 23, 2007
Links for the day 08-23-2007
How Not to Raise a Genius and Baby Einsteins: Not So Smart After All. My experience, both from our own kids and from seeing others in similar situations, is that many parents feel guilty about using TV as a babysitter. Baby/Little Einsteins acts as a means to assuage that guilt by making the TV-time "educational."
Our kids have learned a few things from Baby and Little Einsteins. Inevitably, though, that is because my wife is sitting there watching the show with them, talking about what is happening, and making the experience active rather than passive. So, of course, in the end, the only way for toddlers to really learn is for parents to interact with them. Any situation where the parent(s) have given up on that and turned over the entire job of parenting to PBS, Cartoon Network, the Disney Channel, and Nickelodeon is a bad situation.
At the same time, the APA pronouncement that children under 2 should never watch TV sounds either like an effort to swing the pendulum in the other direction, or a pronouncement from detached academics and researchers who have never had the pleasure of attempting to put a meal together with an 18-month old screaming bloody murder for no good reason.
Anyway, in other links:
Arthur Miller's Missing Act. Much like "Intellectuals" by Paul Johnson
Google Apps no match for MS Office, report says. Say what you will about Microsoft, there simply are not that many other solutions suitable for enterprise-level needs.
On the other side of the ideological spectrum: Apple Laptops Transforming PC Sales?. The trend away from low prices as the determining mark and towards customer service and reliability also seems to be a good thing.
Posted by maphet at 11:07 AM | Comments (0)
August 07, 2007
Netflix and Software Design
Some interesting notes on Netflix specifically and software design in general at the Netflix Community Blog regarding the use of 1/2 star ratings:
here’s what I learned from months of testing this across the country: when we make the ½ star options possible, we get fewer ratings. Significantly fewer ratings. We have argued these results internally for some time, and our best guess is that the complexity of doubling the number of choices from 5 to 10 deters many people from rating, so they just give up. (“3 stars? No, 3 ½ stars.. no… 3 stars… no… oh forget it…”)
In other words, more options sometimes equals less usability. A lesson that is hard to grasp in software development, with the notable exception of Apple.
Also noteworthy is the remark that Netflix treats 3 stars as a neutral rating in their recommendations. Which means I should probably bump some of my movies up to 4.
Posted by maphet at 09:29 AM | Comments (1)
May 21, 2007
On Geekdom
I have been out of town for most of the last two weeks for various work related activities. Because of the nature of my work, this means I have been immersed in geek-relating activities, particularly at vslive. The fun part of this was watching the types of geeks. While there is a predominant stereotype of what a geek is, the reality is that there are several different types. Over the years, these are some of the different types (with all of the usual fluidity that comes from grouping people).
The gung-ho geek - is surprisingly enthusiastic about his work and technology in general. Will use terms like "b*in'" and "f*in' Awesome!" to describe kernel performance without a trace if irony. If you are OK talking tech for long periods of time, this person can be fun to hang out with, given his/her general enthusiasm.
The snob geek - this person knows that he is much smarter than you and has no time for such trivial pursuits as modesty, humility, or, well, social skills. Enough said.
The hardcore geek - the closest to the hollywood stereotype. Is more at home doing esoteric hardware or software activities that no one else can understand than talking to people.
The cool geek - the geek who somehow successfully rises above all negative stereotypes of geekdom and manages to be the alpha male/female in any immediate social context, tech or not. Usually, these people go on to be project managers or business analysts. More common now than 10 years ago, it seems.
The reluctant geek - has the technical ability to work in the industry and does enjoy geeky stuff. At the same time, has enough social awareness to be slightly insecure about being a geek. Attempts to compensate, but doesn't manage to rise to the level of cool geek.
No word now on where I place myself.
Posted by maphet at 04:56 PM | Comments (1)
April 27, 2007
Searching for Hate
Robert X Cringely has one of the oddest columns on the Virginia Tech massacre I have seen. He comments on his own experiences with disturbed students, and then states that
People fail just like machines do. We break for any number of reasons -- mechanical failure, fatigue, bad programming. But unlike machines, people are not subject to statistical quality control, though maybe they should be. There are probably signs of impending failure we could see if we were looking. Only we don't look, because it never seemed necessary.
Now, he argues, some form of looking may be necessary since the consequences of failure have become drastic. A solution:
There are Internet start-ups scouring the web by the hundreds right now looking for every imaginable form of content or commercial intention, but I'm guessing there isn't a single spider program specifically dragging back signs of hate. Why not? Search the web for hate and vitriol and despair, do some clever parsing and analysis to figure out the where and when, then throw a mapping mashup interface on it all with the simple goal of giving school principals and baseball coaches and worried moms and dads a place to look for trouble brewing in their schools, towns or neighborhoods.
To a degree, I think he has a point. Inasmuch as there are already people searching for sex offenders, it seems like a valid next step to start looking for potential homicidal maniacs.
Three problems stand out, though. First, telling a search engine to "Show Me Everyone Who is Really, Really Angry or Upsent" would generate a massive amount of false positives. After all, this is the age of blogging. Sure, 1 out of a million may be an actual lead, but how do you narrow that down?
Second, the potential for abuse is great. Who defines what is hate?
Third, even if you identify a potential homicidal maniac, then what? Usually, he hasn't committed a crime until he (or she, but usually he) starts shooting. The Minority Report comes to mind.
Posted by maphet at 10:08 AM | Comments (1)