harrisj

--- a/subversion-1.6.5/Makefile	2009-09-23 16:34:05.000000000 -0400
+++ b/subversion-1.6.5/Makefile	2009-09-23 17:53:37.000000000 -0400
@@ -41,8 +41,8 @@
-SVN_APR_LIBS =  -L/usr/lib -R/usr/lib -lapr-1 -lpthread
-SVN_APRUTIL_LIBS =  -L/usr/lib -R/usr/lib -laprutil-1 -lsqlite3 -lexpat -liconv
+SVN_APR_LIBS =  -L#{HOMEBREW_PREFIX}/lib -L/usr/lib -R/usr/lib -lapr-1 -lpthread
+SVN_APRUTIL_LIBS =  -L#{HOMEBREW_PREFIX}/lib -L/usr/lib -R/usr/lib -laprutil-1 -lsqlite3 -lexpat -liconv
 SVN_APR_MEMCACHE_LIBS = 
 SVN_DB_LIBS = 
 SVN_GNOME_KEYRING_LIBS = 

harrisj

require_dependency 'net/ftp'

class FTPMock
  def gettextfile(remote_path, local_path)
    test_file = File.join(RAILS_ROOT, 'test', 'data', 'dbready', remote_path)
    if !File.exists?(test_file)
      raise Net::FTPPermError, test_file + ' not found!'
    end
    
    FileUtils.cp test_file, local_path

harrisj

715944 tweets collected on Wed. night for 80 minutes

Filtering out to clients that had at least 1% of traffic, here is the breakdown
Echofon: 28308 3.95%
Mobile Web: 16266 2.27%
UberSocial for BlackBerry: 47728 6.67%
TweetDeck: 35792 5.00%
Twitter for BlackBerry®: 59479 8.31%
Twitter for iPhone: 136235 19.03%
Twitter for Mac: 12870 1.80%

harrisj

require 'net/http'

# These links were taken from my twitter timeline. No endorsement implied.
tiny_urls = {
  'bitly' => ['http://bit.ly/3S9c5K', 'http://developer.nytimes.com/'],
  'j.mp' => ['http://j.mp/JsML0u', 'http://www.npr.org/2012/04/26/151303230/an-african-trader-and-the-perils-of-business-in-china'],
  'tinyurl' => ['http://tinyurl.com/6rq9vob', 'http://politicsoftheplate.com/?p=1283'],
  'owly' => ['http://ow.ly/axesj', 'http://threatpost.com/en_us/blogs/opera-12-beta-released-new-security-features-including-do-not-track-support-042612'],
  'ht.ly' => ['http://ht.ly/1j7LiX', 'http://rssfeeds.usatoday.com/~r/usatoday-LifeTopStories/~3/8vFV-fFfhAc/1'],
  'is.gd' => ['http://is.gd/gk0FoM', 'http://www.wired.com/dangerroom/2012/04/joe-schmoe-drones/'],

harrisj

# encoding: UTF-8

require 'rubygems'
require 'rest-client'
require 'nokogiri'

class ArticleFetcher
  def self.fetch_text(url)
    case url
    when /video\.nytimes\.com/

harrisj

1.9.3p327 :005 > 100.times { puts "#{Faker::Name.name} is a GitHubber" }; 0
Sid Littel is a GitHubber
Henri Wuckert is a GitHubber
Annette Ledner is a GitHubber
Mercedes Murray MD is a GitHubber
Rocky Stracke is a GitHubber
Miss Noah Reichel is a GitHubber
Miss Ludie Schaefer is a GitHubber
Mr. Isom Hilll is a GitHubber
Elva Homenick DVM is a GitHubber

harrisj

One random idea I have been kicking around with in the past is using Twitter as a crowdsourced moderation interface. This is an interesting approach in the following cases:

1. We receive many more potential items than we want to appear on the official, moderated interface.
2. The moderation twitter account is an interesting behind-the-scenes stream for people who are really interested in the content we are moderating down for the final site.

The final site could be another twitter account that posts only the best or anything else really (blog, tumblr, facebook, pinterest, what have you).

This approach won't really work well if you want to filter out questionable content. It's not an obscenity filter but merely a fun way of separating out the most popular items from the stream. How does it work?

For each item you post to the firehose twitter account, you store the following information in your DB:

harrisj

A Simple Solution for Faked Tweets

Recently, a somewhat large selection of my timeline was shocked by the discovery that it's simple to make a fake-looking tweet on the web. Some feared it would be only a matter of time before some news organization is suckered by a fake tweet that seems to come from a real source.

Luckily, the solution already exists, and it's something you already use constantly: GNU PrivacyGuard signatures Here is an approach for verifying a tweet is authentic and hasn't been tampered with that is so simple even @KimKardashian could figure it out. To get started, we just need to do a little setup first:

1. Of course, you have already installed GnuPG for your own use, generated a keypair and uploaded it to a keyserver so that other people can look it up. Its email address must be publicly listed in your twitter profile.
2. Then, you must collect the public keys of the people you fo

harrisj

So, basically, scanning a sentence becomes like this:

1. Split the sentence into words
2. For each word, create a cleaned version (basically strip off quotes and punctuation) and look up in the dictionary.
3. If not found, try a few fallbacks based on stemmming rules or such
4. Otherwise, add the word to a term_misses table.

These code is really ugly (sorry!), and you might notice I actually am talking about meter in it. This is because the background image for the haikus is generated from the meter. But I don't want to do the meter for new words (sorry, syllable is enough), so for those I just return - as the meter (otherwise it's a combination of 1 or 0)

HAIKU (11000 1010011 11110)

harrisj

akept 2
alighted 3
apexes 3
atemporal 4
azolla 3
benshi 2
bewigged 2
blarney 2
brocaded 3
bunga 2