Udderings

(For those who don't want to read my silly rambling and just want to look at the simple example, here are "Linked" Errors.  It's simple enough that you should understand within 30 seconds of looking at the HTML.  If you don't, then you have to read my rambling.)

Error messages are something that have to happen in almost every application.  Even if you follow the Robustness Principle, and allow liberal input (why can't people's usernames contain spaces, or semicolons?  I don't see a good reason why they shouldn't...), sometimes you are forced to present errors.  I love many javascript implementations that let you know if your two password fields are agreeing or if you've entered a valid email address and use color coding to let you know of that interactively.  However, ultimately you have to validate all that input server side and let the user know if there were still errors ("Someone else already has the username 'I love ;s', please choose another").

There are many pages on which these errors are put next to the field they apply to.  I sometimes have trouble finding these error messages, so I decided a long time, that I should put my error messages at the top of the page.  This usually works well, but as forms become more complicated, associating the error message with a specific field can become difficult.

As exhibit A, I present the 2006 SLADM Registration page.  This was a monster with way too many fields and may have presented the largest design challenge in my life.  I was never completely satisfied with how I packed all those fields in and it was a huge balancing act between having it work in Safari, Firefox, and Internet Explorer.  A total of 46 inputs in the form, 22 for each person + 2 that applied to the pair of people.  Some would say the number of fields should be reduced, but they were all needed for either legal or sponsorship reasons, so that isn't a possibility.  I thought about moving error messages closer to what they applied to, but the large number of fields makes the problem of searching a page for the error message even more difficult and the layout probably would have fallen apart if you tried to force an error message in somewhere.

My solution at the time was to work on my copywriting skills.  Well written error messages could direct (smart Washington University) students to correct their errors.  However, I knew from experience watching people that still didn't work.

Eighteen months after the fact, my brain came up with the solution.  It is a ridiculously simple design pattern I've never seen on the internet, which I'm naming "Linked" Errors.  Yes, the word linked is in quotation marks.  Because it's not really a link.  It's a label.

<div id="errors">Your <label for="email">E-mail Address</label> must belong to an educational institution or the US Government</div>

For those of you not in the know, clicking on the text inside the label, will cause the user to be taken to the input with the id in the for attribute.  In this case, clicking on E-mail Address will take you to the input that has an id of email.  You then use CSS to make it look like a link.  For example:

#errors label {
   color:blue;
   text-decoration:underline;
   cursor:pointer;

}

After some basic testing, browsers seem to implement the :hover pseudo-class on labels, but don't implement :active.  They also don't have any idea of this having a tabindex so that one can force tabbing to the "link" or an accesskey.  I'm sure you could code these concepts in Javascript, but that's outside the scope of this simple example.  Now it's time for to see the end product:

See "Linked" Errors in action

So, now you might have looked at the HTML and you might have a major criticism.  How can an input have two labels?  Will this cause problems?  I don't know.  In practice, good copywriting would tell me to make the text in both the labels the same, and that might alleviate certain problems I can imagine in my head.  But really, I want to emphasize that I don't know.  If you have examples where this would cause problems, I'd love to hear about them because this is something I plan to incorporate into everything I build in the future.

Since last April, a number of my friends from Virginia Tech have commented on guns and their beliefs about flaws in the gun laws.  I don't really disagree with them about closing things like the gun show loophole and so on.  One thing has struck me though.  It seems the majority of people I know have never touched a gun.  This came back on my mind when Tammy said one of her students presented her with the first bullet she had ever seen.

Recently, I've seen lots of chatter on the internet about people moving back into cities from suburbs, probably mostly driven by an article in The Atlantic.  I disagree with what many are saying about people wanting to live in cities.  I believe people are looking for a new type of suburb, which I will call the "town".  I can't really think of good examples of towns in Saint Louis, except maybe contrasting the Loop and downtown Saint Louis.  For Silicon Valley, I would say downtown Mountain View vs. San Francisco, but the problem is that SF and the Valley are very two-way commuter while most suburbs are classified as one-way commuter.  But, I can think of dozens of examples in the DC Metropolitan area.   In fact, the article gets into this, describing what I call a town as lifestyle centers.  And the "granddaddy of all lifestyle centers" is in Northern VA, specifically Reston Town Center (opened in 1990).  Since its opening, dozens of communities of this sort have opened, and I had dozens who lived in places like Kingstowne.  Notice they all describe themselves as towns.  People want to live in towns.  In Virginia, I would say a lot of this urge came from the desire to live in the Town of Vienna or Old Town Alexandria, both of which predate any of this stuff.  As a corollary, people looking for towns go where they can find what is most similar.  Many cities are trying to add the features of towns to draw in population.

So, what describes a town?

• walkable, but not so walkable it prevents driving
• there are apartments, but it's easy to find nearby single family homes
  
• single family homes have yards on all 4 sides of the house, there likely are no townhouses to fill a gap in between apartment and home, despite them having town in their name.  Townhouses are crowded suburbia, not what anyone desires.
  
• has a downtown shopping district
• has a small bus system, but probably no trains
• has either a miniature golf course or an ice rink
• probably has a man-made lake with trails
• has parades
  
• has features designed not for economic stimulation, but to get people to live there (this will be key in talking about Falls Church later)
  

Now, much of that may sound like suburbia, but it really isn't the way suburbia turned out.  It instead turned into urban sprawl where people commute to Winchester into DC and are miserable or they live in houses all crammed next to each other and it takes 15 minutes to get to the grocery store because traffic is awful.  A town takes the city, puts it in your backyard, but lets you still have a single family house with a nice lawn.

So, towns are a solution to this problem of people moving out of your little suburb.  What do you do?  You build a city center of course.  As described in the Washington Post article, Falls Church Turns to the Future, my hometown of Falls Church has been trying to do that for the last 8 years.  Lots of things in the article make me laugh.  Like the quotes by the planning commission's feeling that the project "felt rushed" when developers have been modifying plans for 8 years.  The Falls Church Planning Commission has a long-time reputation of being anti-building, so much so that to find an example, I went to the Falls Church News Press and just digged up the oldest issue in the archive.  Like all issues, there's an editorial criticizing the commission for being anti-development.

In examining this project, it is important to see how it differs from what the article in the Atlantic describes and it is much closer to my description of a town.  Despite what the city council and developers claim about economic stimulation, this is not prime real estate for building businesses that would bring in money.  Those locations would be the East and West Falls Church Metro stations, which are both outside of walking distance.  This is quite different from a town center like Rosslyn that is based on spanning two closely located Metro stops (so much so that they describe themselves as the "Rosslyn-Ballston Metro corridor".  So, if it isn't about economic stimulus, what is it about?  It's about making people want to live in Falls Church.  Families want to live in cities where they can walk to do their shopping, but they also want to live in the suburbs where their kids have a big enough lawn that they can let the dog run around outside without a leash.  Cities have always tried to provide this with public parks, and in more recent years with dog parks that have leash-free zones and adding similar amenities of suburbs. (Big example in Boston is gardens where city residents can plant vegetables.  The green space in front of my building is about the same size as my apartment, nothings going to grow there).

So, what I hope this shows is that both suburbs and cities are both trying to head toward this concept I call "town" that is a happy medium of the two.  When people fled the cities, they didn't realize how much they were going to miss some of the benefits.  However, they are too addicted to the suburbs to give up certain amenities.  The return to cities has not been driven by a dislike for suburbs, but by cities providing many of these amenities.  In addition to those I mentioned, a huge part of this has been cleaning up pollution in cities.  You no longer need to be in a suburb to have clean air.  Newly built suburbs are designing themselves to be towns, or what The Atlantic calls "Lifestyle Centers".  Meanwhile, established suburbs, like Falls Church, are trying to build themselves City Centers that will provide them with the same feeling.

One thing I've wondered for years is how changes in what we know can filter down into elementary schools.  And in some cases how that will shift culture and make certain parts of culture unfamiliar.

That sentence sounds quite awkward, so let's give an example.  You probably think there are 5 senses:

1. Sight
2. Sound
3. Smell
4. Taste
5. Touch

But based on what is taught in most physiology courses, you would be wrong.  We have at least four more:

6. Pain
7. Position
8. Temperature
9. Balance

Generally, we don't teach our kids these senses, even though some of them are among the coolest in the human body.  Who would have ever guessed little hairs in our ears play such a significant role in us not falling down all the time?  And while the first five are fairly orthogonal to each other, except for smell and taste, when we add the next 4, a lot of the interactions between the senses and how we perceive the world begin to be built up.  Like, why does pain go away when we rub the area that hurts?  (of course, the traditional 5 sense do mix in some people, in what is referred to as synesthesia)

But I'm getting away from my original point, which was how does us having more than five senses impact culture?  Well, obviously we can't talk about a "sixth sense" any more if we teach our kids that we have 9 senses.  Is there a way we can transition to talking about a "tenth sense"?  I doubt it.  This is an example from science because I'm familiar with that, but the same things are very true of many things about history.  Many of the stories about the early Americas and the beginning of Thanksgiving are just plain wrong.  Yet, we continue to teach fables before crushing them later on in life with the truth.  Or at least that's what I'm told, it's been a long time since I was in elementary school.

More things in science to talk about.  Do they still teach the five kingdoms of life?  Or are we doing the 3 domains everywhere now?  That doesn't seem like it would be too hard to correct, but I learned 5 kingdoms until high school, which was a full 9 years after the change was proposed by Carl Woese, and even then it was taught as a controversial change.  It was only controversial in that some very vocal people didn't like it, and as usual the vocal disagreeement had to be heard for a balanced presentation.  It didn't matter that the biology of cellular membranes and flagella fully backed up the 16srRNA data it was based on.  Then there's the fact that Pluto is no longer a planet.  When will that change in curricula (or has it already)?  And why did Pluto losing planet status make so much news, but noone was told about these 4 extra senses?  I think it has to do with the fact you can't point to something physical like the nose or eye or tongue or ear and say, "That is responsible for your sense of position."  Without something physical, it's much harder to explain.  It's also harder to explain how the "skin", which was previously only responsible for touch, is now simultaneously taking care of pain and part of temperature.  And why those aren't actually part of touch.  You "touch" something to know it's cold, right?  So isn't that touch? (no, nice try...)

Basically, I'm just curious whether the amount our culture depends on a scientific misconception affects it's propagation.  Will the loss of Pluto affect the Greek God and Mickey Mouse's dog?

One of my favorite "blogs" is Indexed.  Apparently, there is now a book.  Me want now.  You should get as excited as me.  And subscribed to the RSS feed for indexed.  You'll love it, trust me.

This summer, I got used to using Yahoo! search.  While many people would disagree with me, I would argue it's better than Google unless the thing you're looking for was added to the internet in the last 2 days.  Part of the reason for my difference in experience may be how rarely I go to search engines.  Most of what I look for any day is in Wikipedia, and I don't even use the browser to do Wikipedia (I use Pathway).

The problem is, my browser of choice is Camino, which unlike Firefox, does not have search plugins.  And like all Mozilla products, it comes with Google by default.  So, I wanted my Yahoo! search.  So, after months, I got around to fixing my problem.

So, this is what it looks like.  Beautiful.  I should add Flickr and del.icio.us to it as well. (They're added now)  I added all the services I could find from search.yahoo.com, so if there's more you think should be added, let me know.  It's fairly easy to delete choices once the file is downloaded, so I'm aiming to add as many as possible with the idea that it's easier to customize via deletion than adding.

So, how do you install this?

1. Download the .plist file from http://static.cowsandmilk.net/yamino/SearchURLList.plist
2. If you want to remove some of the options, open the file with the Property List Editor (should open with this by default) and delete the ones you don't want.  You can also change the PreferredSearchEngine if you like.
   
3. Move SearchURLList.plist to ~/Application Support/Camino
4. Restart Camino
5. You're done

So, I have Yahoo! Search in my life and you can too.

Sometimes you need questions answered about which of two pieces of code are faster.  If you're a computer science major, someone probably taught you how to do this at some point in your life.  I never really was taught this and instead relied on using microtime at breakpoints to profile code.  Lately, thanks to DNA & Protein Sequence Analysis, I've been playing with Perl.  On this visit to the language, I started looking at the Benchmark module, which I think makes pretty output for benchmarking perl code.

I've got this way in perl to make output I like, how do I use it to answer questions I have about programs not written in perl?  Enter comparecommands , a simple script that compares shell commands using Benchmark.

#!/usr/bin/perl
use strict;
use warnings;
use Benchmark qw(:all);

#first argument is how many times the functions are run
my $count = shift @ARGV;

#$test will contain dynamically generated code for comparing the commands
#open up cmpthese
my $test = "cmpthese($count, { ";
#add the anonymous functions to run the command in the shell
$test .= "'$_' => sub { system '$_' }, " for @ARGV;
#close cmpthese
$test .= '});';

#run the dynamically generated code
eval $test;

The first argument is the number of times to run the 2 commands, and then the rest of the arguments are a list of commands to compare.  For example, comparecommands 1000 "php foo.php" "python bar.py" .  Hey, that's amazing, we get pretty output saying which was faster. The obvious question in my mind after writing this was whether I could have done this in PHP, say using PEAR's Benchmark module.  Well, yes, I could have, but it would not have been nearly as simple.  Perl's anonymous functions just make dynamically increasing the number of compared commands easy.

If you have two pieces of code you want to compare that are both in PHP, it probably is better to just throw together a benchmark using the PEAR module, but when you want to compare code in two different languages, I think this code is useful.  If someone wants to clue me in on more useful ways of doing benchmarking cross-language, hopefully you know where to find me.

For those who don't know me, I haven't spent any time in the public education system.  In fact, my education was:

• a couple years in a private preschool
• 11 years at The Congressional Schools of Virginia
• 4 years at Saint Stephen's and Saint Agnes School
• 4 years at Washington University in Saint Louis
• and now I'm at Boston University

I just put that out as a disclaimer that I think is actually irrelevant to what is going in this post, but I thought someone might criticize me if I didn't come out with that up front.

So, moving on to the important bits.

In Ron Paul's talk at Google, he says one of the first departments he would eliminate is the Department of Education.  One Googler asks if eliminating the Department of Education would hurt the education of poor children and not provide a fair chance to all children.  He also presents a scenario where a rich state/locale decides the private schools are adequate and eliminates the public schools.  The Googler makes several assumptions that I refuse to accept:

1. Local governments in poor areas are unable to fund their schools and/or manage public schools
2. The solution to all underperforming school districts is the same as it can be mandated from the national level
3. Private schools are the mainstay of the wealthy and driven by accepting wealthy in order to have lots of money

While I disagree with much of libertarianism, I must agree with Ron Paul on the Department of Education as I have only observed bad consequences as it has become more powerful.  In fact, I would argue that even certain state level education laws in Virginia have hurt public schools at the local level, but less severely.  I also would like to argue against point three by examining a level of education where private schools are much more prominent, that of the universities and how they have performed in regards to the poor along with some ways that private institutions are driving forward the education of the poor at a young age.

Exhibit 1: Fairfax County Public Schools vs. the Department of Education
Most of my knowledge of this has come from the Washington Post's coverage, so my feelings may be influenced by theirs, but the fact that they set up a special page for this indicates how much this is a big deal.  Fairfax County has over 100 languages spoken exclusively in the homes of children (where exclusive means the parents speak no English).  For this reason, their website is in seven different languages in addition to English.  Due to this unique situation, one of Fairfax County's high schools, Jeb Stuart, was featured in the September 2001 issue of National Geographic.

Recently, Fairfax County has been fighting how No Child Left Behind measures literacy for non-native speakers of English.  Before No Child Left Behind, literacy progress for students was measured as improvement over the year before.  Under No Child Left Behind, literacy is measured as a definite point at different grade levels.  If a student moves to the US at age 14, never having heard English, and not even literate in his native language, no longer is it sufficient progress to get them to grade 1 level by the end of freshman year, grade 4 by sophomore year, grade 7 by junior year, grade 10 by senior year.  You would have failed this student under No Child Left Behind, but far exceeded any reasonable person's expectations for your educational system.

In April, this came to a head when Fairfax County bluffed saying that they would not perform the tests required under No Child Left Behind.  They gave in because of the $17 million bounty they would lose if they didn't do the new tests.  And noone pays attention to the results.  All the parents of children at Fairfax County schools know they are excellent and understand why their school is viewed as "failing" in the eyes of the Federal Government.  So, is No Child Left Behind at all useful for Fairfax County?  No.  Some people would argue that Fairfax County is the exception to the rule, but I'm sure there's plenty of other school districts where the Federal Government defining their goals impacts how the students are being educated in a negative way.  In Fairfax, they luckily are smart enough to know the federal government's goals are silly and are happy to define their educational goals as they see fits the local situation.  Now imagine a school district that is actually underserving its population.  If there was no federal involvement, they would examine their system and create priorities and self-examine to find the problems in their system.  But since there's a federal government, they may just be willing to accept the federal government's criteria for success and follow their recommendations rather than doing self-examination.  This runs counter to my own experience of solving problems.  The first step in solving a problem is clearly defining the problem from your local perspective and finding the root of that problem.  Often this first step is where the creativity happens and ideas for solutions emerge.  Without self-management and assessment, that first step never occurs and the creative solutions to problems never emerge.

Exhibit 2: Ivy Leagues vs. Public Education on costs
Your family makes $35,000/year and you just got into a state school and Yale, which is cheaper to attend?  Believe it or not, the private school, Yale, would probably be cheaper.  So would Harvard.  And many other private schools.  If your family income is under $45,000 a year, your expected parental contribution at Yale is $0.  So, what happened to private schools being for the rich?  I believe that assumption came from the fact that these schools tend to let in children of alumni and alumni tend to go on to be successful and rich.  And because rich people who have smart children give them money so they can fund the children who aren't as rich, but are also smart, to come and get an education as well.  But these schools are need-blind, so that means your financial situation is not considered in admission.  Not that going to a school in a rich neighborhood with a good reputation wouldn't be considered as a plus over attending one that isn't so stellar, this discrimination correlated to need likely still occurs.  But the key is, a private education isn't reserved for the rich and wealthy and need has been almost completely negated from attending a great university.  Of course, you still have to get the education before college that can get you in, which is where this next exhibit comes into play.

Exhibit 3: Bring Me a Book
When I was working at Yahoo, for Yahoo! for Good day, I volunteered at the Bring Me A Book Foundation.  It was quite an interesting experience because they have a warehouse in Mountain View, which is not cheap.  In fact, nothing about their operation was cheap.  But that isn't surprising because their goal was to bring "High Quality Children's Books" to children in need of reading programs.  High quality in both how the book was written and illustrated, but also in the set of books sent into a situation and in the physical quality and durability of the book.  These books aren't meant to affect one or two kids, but whole communities for years to come.

So, what impact can this have?  A huge factor in success in life is the ability to read when entering kindergarten.  Most low-income areas don't have age-appropriate books for children.  The goal of Bring Me a Book is to use the wealth of Silicon Valley to place age and culture-appropriate books into low-income areas along with encouraging reading aloud.  They even have programs for teaching people how to read aloud and encouraging non-native english speakers to use the opportunity to improve their English skills along with reading aloud to their children (or audio tapes are included with the books to help in this learning).  The ability to read isn't the only thing impacted by reading aloud, creativity is cultivated in the young mind by stimulating stories, so even making up stories to go along with picture books is encouraged among these parents.

I'm not going to pretend like Bring Me A Book is the definitive solution to education problems in poor areas, but its proof that the goal of the wealthy isn't to hold down poor students, instead they believe in improving education for all children and Bring Me A Book is a great example of this.

Today I was on the T and picked up a copy of the BostonNOW off the floor as I tend to do.  Since I like scandal, I immediately turned to the lead article on teachers and sexual harassment.  It provided the scandal factor I needed for the day, I was quite satisfied.

But there was something that bothered me tremendously.  In an inset to the article (present at end of article online), it says, "Nine out of 10 of those abusive educators were male."  Now, this is nice to know, but the article has three examples of sexual harassment by teachers and they were all female.  What was the deal?  Well, the byline says Associated Press, so it's time go visit their copy.  The AP's story doesn't have the nine out of 10 fact, so it really makes it seem like a bunch of female teachers are taking advantage of our students...