No, not this one...

Some dude has an posted an archive of ‘high-res’ stills from Kubrick films posted at Stanley Kubrick, deserving of worship, a statement with which I wholeheartedly agree.

While I came across the link to the publicity stills from 2001: A Space Odyssey on boingboing, after looking around, I think the most intriguing of these are these behind-the-scenes shots from Dr. Strangelove… and not just because I have designs on writing a stage musical called FailSafe. The process of seeing the artist at work is fascinating enough (if you’re a Kubrick-o-phile like me and you haven’t seen Stanley Kubrick: A Life in Pictures, it’s worth checking out, even if it is narrated by Tom Cruise).

Strange technicolor

But what blows my mind the most? Seeing shots from Dr. Strangelove in color. I’ve become so accustomed to the story being tied to the static monochrome of black-and-white that it’s part of my very being, and the shot in color below of the Russian Ambassador is a blow to my worldview. And it made me wonder – how much of Dr. Strangelove’s fantasticness comes from the fact that it was shot in black+white, and how would it have been different were it shot in color?

Although the film was shot in black and white, using varying aspect ratios which lends a gritty realism to the interplay between narrative and newsreel footage, seeing Kubrick’s perfectionism shine through just as clearly in color, from  the maps on the walls, to the green baize on the table in the war room to give the actors a sense of that they were playing a game of poker to determine the fate of the world.  And this makes me wonder: how would the impact of the film would have changed had it been shot in color?

Kubrick in the doctor's chair

I’d love to know the answer, but then again… maybe I wouldn’t.

Mein Führer...! I can walk!!

Mandrake pontificating (I don't remember this shot...)

No related posts.

Welcome to summertime in the city.

SF Timelapse from 1 on Vimeo.

Views from and of the Downtown Skyline, Corona Heights, Golden Gate Bridge + Dolores Park.

The video’s a little buggy, but once you get it to work, it’s worthwhile – so many stories in every frame.

No related posts.

So that I can reclaim this blog for some of my other interests (read: topics other than software development and user experience), be more like my normal irreverent self, and intersperse random facts everywhere, I’m starting to migrate my work-based content to a new blog. I’ll still geek out here, but more likely on language, food, and random youtube videos.

You might also want to check a couple of other blogs I’ve been working on lately:

A Room Full of Idiots – watching the decline of western civilization through the lens of Yahoo Answers

Stuff that tastes like other stuff – on one hand, a journey into the human condition, examining how indulging in stuff that tastes like other stuff is a safe way to conform while perceiving oneself as nonconformist, to experiment within accepted social channels, and to give in to curiosity without the risk of being ostracized. On the other hand, a collection of a bunch of food that makes you go WTF?

No related posts.

Over the past few weeks, I’ve discovered that most web forms that contain a field for a url have a major usability failure when it comes to validating the content of the field… the validation fails if you neglect to enter “http://” preceeding the variable portion of the url.

Sure, sure, I get it. Perhaps the form is anticipating that a user may be linking to a site using https://, and thusly requires the user to specify the protocol… but for users linking to blogs, corporate websites, and the other most common use cases for these types of fields, the likelihood of linking to a secure server is (or should be) inversely proportional to the number of users on the site… or in other words, as the number of users increases, the probability of a user’s url requiring https:// approaches 1/∞ (remember algebra? it gets close to zero).

So, a note to all my fellow developers out there… chances are, if someone is linking to something on your site, it’s far more likely that it will require http:// rather than https://… Browsers don’t require that you type http:// before your URL anymore… neither should our forms.   Adhering to the same conventions users are used to elsewhere in their travels throughout the internet will create  a much more satisfying user experience.  Here are a couple of recommendations:

1. Write up your validations accordingly.  Don’t require a prefix, but allow users to add one, on the off chance that they *are* referring to a secure server.
2. If, for some reason, you deem it necessary to require users to specify the appropriate prefix, gosh darn it, specify the format for a valid url below the field, or at least display the required format as a tooltip/popup when the user is focused on the field, and draw attention to the required prefix.  Novice – and even seasoned users – may otherwise gloss over it – and be forced to endure the frustration of a failed attempt at submission.
3. When the field loses focus, alert the user to the fact that their url is not formatted properly.  In big Red letters, and all caps if necessary… preferably somewhere near the field (if it’s still visible in the user’s frame of vision) or near the submit button to remind them to fix their error before continuing

Sure, chances are if they submitted the field once, they’ll fix their error and submit it again… but they’ll be mumbling and grumbling under their breath at you, and I know that some of you out there wouldn’t want to hear the names I’ve called you after a failed form submission. So be good to your users, and build your forms and validations with an eye towards the most common use cases, and make those who subvert convention be the ones who do the work.

For the less geeky, in case you’re wondering why http vs. https matters at all, take a look after the cut for the random facts.

Here’s the gist: HTTP is an unsecured connection, while HTTPS creates a ‘secure’ connection over an unsecured network (e.g., the internet). The connection is deemed ‘secure’ if a “trusted authority” has deemed the server to be trustworthy. You’ve seen those popups before:

From SSLShopper

Think about it this way – you go to the pharmacy to pick up a prescription, and they ask you for your address. If the address you give them doesn’t match what they have on record, it raises a red flag. Essentially, the same concept works here – your browser alerts you if what’s on record doesn’t match.

Also, http:// connects to a server using port 80 by default, but  HTTPS URLs (aka https://) connects on port 443… and that does makes a difference, like if you went to visit a friend and walked into apartment 443 when they lived in apartment 80… kinda.  Really, that bit isn’t such a huge issue, and in most cases it could be solved if the site directed you to the appropriate url if you used the wrong prefix… but that’s another story for another day.

No related posts.

A quick checklist for creating a new app with geolocation on heroku’s 1.8.7 stack… so you can use the count method to your heart’s content. I’m sure that there’s something I’ve missed, but it’s a start…

1. sudo gem install geokit (if you haven’t already)
2. script/plugin install git://github.com/andre/geokit-rails.git
3. for new app, create heroku app (
   • on 1.8.7 stack – heroku create APPNAME –stack bamboo-ree-1.8.7 –remote REMOTENAME
   • add the following to app’s .gems file:

     1 2 3

     rails -v 2.3.5 pg geokit --version '=1.5.0'

4. get new developer keys for your heroku app
   • http://developer.yahoo.com/maps/simple/
   • http://code.google.com/apis/maps/signup.html
5. add config.gem “geokit” to config/environment.rb
6. edit your application controller file:

   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

   include Geokit include Geokit::Geocoders include Geokit::IpGeocodeLookup class ApplicationController < ActionController::Base   helper :all   protect_from_forgery   geocode_ip_address     before_filter :geokit   def geokit     store_ip_location     @current_location = session[:geo_location]  # @current_location is a GeoLoc instance.   end   def store_location     session[:return_to] = request.request_uri   end   end

From this point forward, the geocoding methods of the Geokit gem are available. There’s more that needs to be done if you want to treat one of your models as geolocatable… but I’ll pick that up later :)

No related posts.

If you’re a Mac user forced back into the world of .NET, you might come across the following error message when creating a new Visual Studio project over Parallels:

Invalid URI: the hostname could not be parsed.

Chances are, you are trying to create the project in a directory that is inaccessible to Windows. In the New Project window, check out the Location field. If it contains psf://, you’re trying to save to the Parallels shared folder – a location Visual Studio apparently can’t handle. Change the location to somewhere on your C: drive, and you should be good to go.

Related posts:

1. Display Line Numbers in Visual Studio

Why Visual Studio 2008 doesn’t display line numbers by default? I don’t know. Why there isn’t a way to toggle line numbers on and off from the interface? I don’t know. Why it…. well, you get the point. Here’s how to fix it so you don’t have to waste time figuring it out:

From the Tools menu, select Options > TextEditor > All Languages > General…… then check the box beside Line Numbers.

You can also turn on line numbers for specific languages too – just select the target language instead of “All Languages.”

If you’re on the right screen and you still don’t see it, make sure the “Show all settings” option in the bottom left corner of the Options windows is checked.

Related posts:

1. Creating Visual Studio projects from Parallels

I just witnessed the first sign of my impending geekpocalypse… rather than go through the ‘hassle’ of working with a gui-based calculator, I launched irb to do a few quick calculations.

Oy vey.

No related posts.

I’ve a great adoration for the music of Blondie, and in more than one way, I think that Debbie Harry is a great inspiration… but I often can’t handle some of those high-energy tunes on Monday morning. Personally, I’ve always thought that Call Me would be fantastic downtempo in an abandoned cathedral at 3am.

So to celebrate the release of my first iPhone app (MuniMe, in the app store now :), I decided to take a little bit of my spare time away from coding and play around with music again, so I spent an afternoon a few weeks ago tinkering around with my keyboard and Garage Band, and this is the result.

The production value might not be exquisite, but since I’ve been a bit of a purist in the past, I wanted a chance to muck around with different effects and explore beat frequency by overlaying an auto-tuned track with the original (e.g., the last chorus).

And no, this song is not a secret message to anyone. There’s no subtext here. None at all. (You’ll know if this message is directed at you. Don’t call me. Really.)

No related posts.

What is true evil: evil for the sake of being evil, or evil for the sake of being good?

No related posts.