Font Squirrel | Handpicked free fonts for graphic designers with commercial-use licenses..

HTML5 Cross browser Polyfills – GitHub.

There is a Google Group: Friends of the Unofficial Google Reader API | Google Groups.

And there is a Python library where you can find lots of good API information: GoogleReaderAPI – pyrfeed

There is also a small PHP library: greader-library:

PHP class library for providing access to the Google Reader API. Enables authentication and provides public methods for general Google Reader actions.

I don’t think greader-library supports OAuth as of this writing, but this may change in the future.

Know of any other good resources for the Google Reader API? Let me know!

Here is our scenario:

We are developing code for a client and managing it with some branches on GitHub. The code lives on our local machine in /Users/myUserName/clients/myClient/code. We have Git branches for development and production, and another, local branch where we write experimental code. We are on version 4.5 of this project. So we find ourselves typing the following commands into the command line A LOT:

$ git checkout myClient_4.5_development
$ git checkout myClient_4.5_production
$ git checkout myClient_4.5_local

…but we would much rather simply type…

$ dev
$ prod
$ loc

…to achieve the same thing.

To do this, we need to create some scripts that will contain the git commands, make sure they are executable, and then add the directory where they live to the $PATH variable.

1. Navigate to the directory where we keep everything related to our client.

$ cd ~/clients/myClient

That “~”, just in case you don’t know, is a shortcut to our home directory located at /Users/myUserName

2. Add a directory for the shell scripts and navigate into it.

$ mkdir scripts
$ cd scripts

3. Create the files for our scripts.

$ touch dev
$ touch prod
$ touch loc

4. Make the scripts executable…

$ chmod 700 dev

…and repeat, as in step 3, for prod and loc.

5. Edit the shell script with whatever text editor you are comfortable with. People like vim, pico, nano, emacs… or you can go the long way around and use TextEdit or some other, more GUI-oriented editor. So I might normally use vim, but it can be a little bit difficult for those unfamiliar with it, so let’s use nano instead.

$ nano dev

Whatever text editor you choose, we want the following in our scripts…

#!/bin/sh
git checkout myClient_4.5_development

…and we would change “development” to production or local for the prod and loc scripts.
That first line is the shebang, which tells the command line what program to use to interpret our file. In this case, we want the shell script interpreter (sh), and we typically find that within the root bin file (/bin).

6. Now we simply need to add our scripts directory to the $PATH variable, so the command line will know where to search for dev, prod and loc. To do this, we need to edit our profile file. This might be called .profile on your system, or it might be called .bash_profile, or something similar. You will find it in your home directory.

$ nano ~/.bash_profile

We need to add the following line to our profile.

export PATH=/Users/myUserName/Clients/myClient/scripts:$PATH

We’re done! Now we can open a new terminal window, navigate to our code directory and type our new commands to switch between Git branches. When our version number changes, we simply edit our scripts accordingly.

I’m not the Unix master I would like to be, so if anyone sees a way to improve this, please let me know in the comments and I will update this post. Thanks!

(I feel that I should have this stuff under my belt, but I often have to remind myself of the weird rules of JavaScript. I should probably be studying the writings of Doug Crockford more.)

if(myVariable === void(0){
    //do amazing things
}

So I start to wonder about this sort of thing, and I turn to the awesome JS Bin. I set up a test there, and I can provide you with a link to the test… I love that.

But JS Bin is rather impermanent, and just in case that’s not working out, here is the original test and the results.

code

/* check for an absent argument to a function */

var testIt = function(one, two){
  gnomeSane = '';
  if(two === void(0)){
    gnomeSane += "two is undefined. (two === void(0))<br />";
  }
  if(two === null){
    gnomeSane += "two is null. (two === null)<br />";
  }
  if(two === false){
    gnomeSane += "two is false. (two === false)<br />";
  }
  if(two === ''){
    gnomeSane += "two is an empty string. (two === '')<br />";
  }
  if(two === 0){
    gnomeSane += "two is zero. (two === 0)<br />";
  }

  if(two == void(0)){
    gnomeSane += "two is equivalent to undefined. (two == void(0))<br />";
  }
  if(two == null){
    gnomeSane += "two is equivalent to null. (two == null)<br />";
  }
  if(two == false){
    gnomeSane += "two is equivalent to false. (two == false)<br />";
  }
  if(two == ''){
    gnomeSane += "two is equivalent to an empty string. (two == '')<br />";
  }
  if(two == 0){
    gnomeSane += "two is equivalent to zero. (two == 0)<br />";
  }

  if(typeof two == "undefined"){
    gnomeSane += "type of two is 'undefined'. (typeof two == 'undefined')<br />";
  }

  if(!two){
    gnomeSane += "not two is true. (!two)<br />";
  }

  gnomeSane = gnomeSane == '' ? 'two is '+two : gnomeSane;
  document.write('<p>'+one+'<br />'+gnomeSane+'</p>');
}

testIt('<b>The second argument is not being passed to the function.</b>');
testIt('<b>Now the second argument is null.</b>', null);
testIt('<b>Now the second argument is false.</b>', false);
testIt('<b>Now the second argument is an empty string.</b>', '');
testIt('<b>Now the second argument is zero.</b>', 0);
​

results

The second argument is not being passed to the function.
two is undefined. (two === void(0))
two is equivalent to undefined. (two == void(0))
two is equivalent to null. (two == null)
type of two is 'undefined'. (typeof two == 'undefined')
not two is true. (!two)

Now the second argument is null.
two is null. (two === null)
two is equivalent to undefined. (two == void(0))
two is equivalent to null. (two == null)
not two is true. (!two)

Now the second argument is false.
two is false. (two === false)
two is equivalent to false. (two == false)
two is equivalent to an empty string. (two == '')
two is equivalent to zero. (two == 0)
not two is true. (!two)

Now the second argument is an empty string.
two is an empty string. (two === '')
two is equivalent to false. (two == false)
two is equivalent to an empty string. (two == '')
two is equivalent to zero. (two == 0)
not two is true. (!two)

Now the second argument is zero.
two is zero. (two === 0)
two is equivalent to false. (two == false)
two is equivalent to an empty string. (two == '')
two is equivalent to zero. (two == 0)
not two is true. (!two)

conclusions

If you want to really just test for an undefined argument use:

two === void(0)

or

typeof two == 'undefined'

I prefer the clarity of the latter. But we will need to be careful if we do that, because typeof null is equivalent to ‘object’.

If you want to test for both undefined or null use either:

two == void(0)

or

two == null

Using both might make the code more clear, but it is not strictly necessary.

If you want to test for all “off values” (undefined, null, false, empty string, zero), then use:

!two

This is another test to see how easy I can make it for a client to post to their blog. But maybe my real problem is all the bugs on my computer.

Posted via email from Bill’s posterous

Before now, this was something only possible with tools like Mapnik, Cascadenik and OpenStreetMap. And for white-label maps or extremely customized maps, those tools are still quite relevant.

Resources

• Google Maps API
• Google Styled Maps Documentation
• Google Styled Maps Wizard

Original vCard

BEGIN:VCARD
VERSION:3.0
N:Gump;Forrest
FN:Forrest Gump
ORG:Bubba Gump Shrimp Co.
TITLE:Shrimp Man
TEL;TYPE=WORK,VOICE:(111) 555-1212
TEL;TYPE=HOME,VOICE:(404) 555-1212
ADR;TYPE=WORK:;;100 Waters Edge;Baytown;LA;30314;United States of America
LABEL;TYPE=WORK:100 Waters Edge\nBaytown, LA 30314\nUnited States of America
ADR;TYPE=HOME:;;42 Plantation St.;Baytown;LA;30314;United States of America
LABEL;TYPE=HOME:42 Plantation St.\nBaytown, LA 30314\nUnited States of America
EMAIL;TYPE=PREF,INTERNET:forrestgump@example.com
REV:20080424T195243Z
END:VCARD

vCard Encoded for a URL

BEGIN%3AVCARD%0AVERSION%3A3.0%0AN%3AGump%3BForrest%0AFN%3AForrest%20Gump%0AORG%3ABubba%20Gump%20Shrimp%20Co.%0ATITLE%3AShrimp%20Man%0ATEL%3BTYPE%3DWORK%2CVOICE%3A%28111%29%20555-1212%0ATEL%3BTYPE%3DHOME%2CVOICE%3A%28404%29%20555-1212%0AADR%3BTYPE%3DWORK%3A%3B%3B100%20Waters%20Edge%3BBaytown%3BLA%3B30314%3BUnited%20States%20of%20America%0ALABEL%3BTYPE%3DWORK%3A100%20Waters%20Edge%5CnBaytown%2C%20LA%2030314%5CnUnited%20States%20of%20America%0AADR%3BTYPE%3DHOME%3A%3B%3B42%20Plantation%20St.%3BBaytown%3BLA%3B30314%3BUnited%20States%20of%20America%0ALABEL%3BTYPE%3DHOME%3A42%20Plantation%20St.%5CnBaytown%2C%20LA%2030314%5CnUnited%20States%20of%20America%0AEMAIL%3BTYPE%3DPREF%2CINTERNET%3Aforrestgump%40example.com%0AREV%3A20080424T195243Z%0AEND%3AVCARD

Encoded, but perhaps easier to read

BEGIN%3AVCARD%0A
VERSION%3A3.0%0A
N%3AGump%3BForrest%0A
FN%3AForrest%20Gump%0A
ORG%3ABubba%20Gump%20Shrimp%20Co.%0A
TITLE%3AShrimp%20Man%0A
TEL%3BTYPE%3DWORK%2CVOICE%3A%28111%29%20555-1212%0A
TEL%3BTYPE%3DHOME%2CVOICE%3A%28404%29%20555-1212%0A
ADR%3BTYPE%3DWORK%3A%3B%3B100%20Waters%20Edge%3BBaytown%3BLA%3B30314%3BUnited%20States%20of%20America%0A
LABEL%3BTYPE%3DWORK%3A100%20Waters%20Edge%5CnBaytown%2C%20LA%2030314%5CnUnited%20States%20of%20America%0A
ADR%3BTYPE%3DHOME%3A%3B%3B42%20Plantation%20St.%3BBaytown%3BLA%3B30314%3BUnited%20States%20of%20America%0A
LABEL%3BTYPE%3DHOME%3A42%20Plantation%20St.%5CnBaytown%2C%20LA%2030314%5CnUnited%20States%20of%20America%0A
EMAIL%3BTYPE%3DPREF%2CINTERNET%3Aforrestgump%40example.com%0A
REV%3A20080424T195243Z%0A
END%3AVCARD

Submit to Google

This can now be submitted to Google’s Chart API as a URL:

http://chart.apis.google.com/chart?cht=qr&chs=400x400&chl=BEGIN:VCARD%0AVERSION:3.0%0AN:Gump;Forrest%0AFN:Forrest%20Gump%0AORG:Bubba%20Gump%20Shrimp%20Co.%0ATITLE:Shrimp%20Man%0ATEL;TYPE%3DWORK,VOICE:(111)%20555-1212%0ATEL;TYPE%3DHOME,VOICE:(404)%20555-1212%0AADR;TYPE%3DWORK:;;100%20Waters%20Edge;Baytown;LA;30314;United%20States%20of%20America%0ALABEL;TYPE%3DWORK:100%20Waters%20Edge\nBaytown,%20LA%2030314\nUnited%20States%20of%20America%0AADR;TYPE%3DHOME:;;42%20Plantation%20St.;Baytown;LA;30314;United%20States%20of%20America%0ALABEL;TYPE%3DHOME:42%20Plantation%20St.\nBaytown,%20LA%2030314\nUnited%20States%20of%20America%0AEMAIL;TYPE%3DPREF,INTERNET:forrestgump%40example.com%0AREV:20080424T195243Z%0AEND:VCARD&chld=L

The Result

The image below is coming from the Google Chart API, not an image file on my server.

Resources

• vCard
• URL Encode and Decode Generator Tool
• PHP urlencode
• Google Image Chart API – QR Code Generator

The creepy thing, however, is that I had to give my Twitter login info to Tarpipe to get this rolling. Even if I decided that I trust these people, how can that be secure? They must be submitting this info to Twitter in plain text, right? That means they are storing it in plain text, right? Or maybe I don’t understand their advanced cryptography? My understanding was that once you encrypted something, that’s it. You can’t de-encrypt it. You’re supposed to simply compare two encrypted values… something tells me this is not happening.

I also wonder how this is achieved on Posterous or Tumblr, with their native autoposting to other services.

Anyhoo… it might be incredibly useful at some point, so I’m just checking it out for now. But my Twitter password will probably change very soon.

UPDATE: Looks like Tarpipe is pretty much barely beta, and it’s not as useful as I originally thought… probably will be after they start using PubSubHubbub or cron jobs or some other way of turning regular RSS/Atom feeds into real-time feeds. Apparently, one needs to develop a custom application to use much of what they have built. That’s cool, but not easy for a quick test.