July 25, 2008

Digg this Post page to del.icio.us

An article entitled Girls = Boys at Math on ScienceNOW reports that a study conducted by University of Wisconsin, Madison researchers found that female students have the same ability to solve complex math problems as male students.

May 20, 2008

Digg this Post page to del.icio.us

I was working on a complicated layout where I was trying to center a div. For some reason, the usual auto margins technique would not work because the div always expanded to the full width of its container even though the width was not set to 100%.

The only way I could get the div to center was to use auto left and right margins in conjunction with setting a width. However, I wanted the layout to scale gracefully if the font size was changed. The solution was to specify the width in em. This makes the div resize with the font. Quick and dirty, but it works well enough until I have time to come back to take a closer look at the problem again.

March 25, 2008

Digg this Post page to del.icio.us

True story that happened to me today: someone asked me to snail mail them a copy of one of my blog entires and gave me their home address.

The first question that comes to mind is how they figured out how to get to my blog and leave a comment on it without figuring out how to save a web page or print it. Furthermore, what made them think I would honour their request to mail them a copy of my blog entry? (I'm not going to.)

Is this type of user representative of an average web user or a less-than-average web user? How many people like this are out there and do they matter?

Being a computer engineer, I really take computer literacy for granted. This type of encounter with lay people gives me an interesting perspective on the type of audience my products may attract when building the Internet.

March 1, 2008

Digg this Post page to del.icio.us

I'm at the MySpace OpenSocial Dev Jam but I'm leaving in an hour even though the event is supposed to run from 9:00 AM - 9:00 PM. All street parking in the area has a 2 hour limit, and all the paid lots in the area are closed.

Aside from the parking situation, who starts a developer event in the Bay Area at 9:00 AM on a Saturday?

February 21, 2008

Digg this Post page to del.icio.us

p3p.gifMy introduction to P3P was purely out of need. I maintain a website where we were using Google Analytics to count our users but we noticed that a disproportionate number of Firefox users were being reported. Further inspection revealed that visits from over half of our Internet Explorer users were not being recorded.

This web site appeared within an iframe on a different domain than the frame parent, and we noticed that the Google Analytics cookies were not being set in the default Internet Explorer environment with a medium privacy setting. A privacy icon also appeared in Internet Explorer's status bar, indicating that site cookies have been blocked.

Why does this cross-domain cookie issue only manifest itself in Internet Explorer 6 and Internet Explorer 7?

It turns out that Firefox and Internet Explorer have different definitions for "third-party" cookies. In IE, a cookie is considered to be "third-party" if the cookie's domain differs from the domain shown in the address bar (even if the domain of the cookie is the same as the domain of the page that is setting the cookie.) In Firefox, a cookie is considered to be "third-party" if the domain of the cookie is different than the domain of the page that is setting the cookie, (regardless of the address that appears in the location bar.)

Here is a summary of what you need to do in order to allow third-party cookies to be set in IE:


  1. Create a human-readable privacy policy for your web site (i.e. in HTML)

  2. Translate the human-language privacy policy into an XML document using IBM's P3P Policy Editor (considered to be the best free tool available.)

  3. IBM's P3P Policy Editor will also generate a P3P compact policy for you.

  4. Emit a compact policy and a reference to the full privacy policy in the HTTP header of your web page

  5. Place your policy reference XML file (p3p.xml) and policy XML file (policy.xml) in the well-known location set forth by the P3P standard (/w3c/p3p.xml and /w3c/policy.xml)

  6. If your privacy policy is satisfactory, your web page will be permitted to set cookies. This MSDN article contains a table of things that will make your privacy policy unsatisfactory.

Although the P3P standard specifies that only the XML policy files in the well-known location and human-readable privacy policy are necessary and sufficient, IE6 doesn't work that way. IE6 only honours the compact policy (and ignores every other implementation method.) IE7 behaves as expected.

Now that we've discussed the solution, we can go on to explain why P3P was created. P3P Toolbox has a lengthy discussion about the need for P3P. The main idea behind P3P is that it is burdensome and unreasonable for the average web user to hunt down the privacy policy for every site that they directly interact with, let alone the sites that they unknowingly interact with (like those displayed in iframes.) Regular people just don't do this. P3P was created to standardize the discovery of privacy policies from site to site.

P3P does not enforce that site owners adhere to their own privacy policies, it merely automates the interpretation of the legalese in which most privacy policies are written. The enforcement of privacy laws is left to the judicial system.

February 14, 2008

Digg this Post page to del.icio.us

I asked a MySpace developer why a few particular links on their site were javascript onclick events applied to a span tag, rather than anchor tags.

He said it was because the product specifications required those links not to be underlined.

*headdesk*

This explains a lot.

February 12, 2008

Digg this Post page to del.icio.us

A couple of weeks ago I attended the Silicon Valley Girl Geek Dinner at Google. I've been curious about Women 2.0 and what they were about since I was supposedly in their target demographic. The event left me feeling disappointed because it felt more like a pep rally than a professional development event. Beyond all the slick marketing and corporate sponsorship, there was little substance.

This conversation I overheard at the conclusion of the discussion panel summed it up best:

Girl: Oh honey, this event was so empowering!

Girl's boyfriend: As long as it was empowering for you, dear...

I agreed with her boyfriend. The event wasn't particularly empowering.

Today, I received an invitation to participate in the second OpenSocial Hackathon.

One particular mandatory field on the registration form caught my attention (mostly because it wouldn't let me proceed without submitting an answer for it.)

Are you a female designer/developer? Your html/css and layout experience can replace server-side requirement.

My first reaction was, "I am a designer and developer who happens to be female, but I don't see what difference that makes, and why this question is mandatory. Are these people sexist?"

This question makes it sound like they're willing to lower the bar to accommodate women, which makes me feel insulted. I understand that they're trying to encourage more females to come out and participate, but the messaging isn't particularly empowering.

To quote Mike:

By trying to compensate for the historical mistreatment of a group you isolate it and reinforce the definition of the differences between that group and the rest of the world

Recent Jobs

Web Portfolio

www.flickr.com