Feeds:
Posts
Comments

How do you select a column of text in BBedit? 1st turn off “Soft Wrap Text,” then option-select columns or rectangles of text.

Rafferty	31
Jones	33
Steinberg	33
Robinson	34
Smith	34

With “Soft Wrap Text” off, hold down the option key while dragging the cursor with mouse or trackpad. With the selection you want highlighted, release option and mouse/trackpad; command-c to copy.

Rafferty
Jones
Steinberg
Robinson
Smith

An alternate if you want to kill the 2nd column of text would be to do a regular expression (grep) find/replace. Find: ^(.*)\t.*$ Replace: \1

SproutCore

SproutCore is an open-source framework for building blazingly fast, innovative user experiences on the web. SproutCore.

Create your Google Sitemap Online – XML Sitemaps Generator.

Sphider – a php spider and search engine. anyone know of an easier way of generating a google site map from a cron job?

This is a simple script I use to consistently embed Google Analytics campaign tracking code within bit.ly short urls. The approach is similar to my Klout user score look-up: it uses Alfred App Powerpack extensions (and Growl)

The GA tracking code lets us know how our overall “direct social” efforts are performing and shows with sources (Twitter, Facebook, LinkedIn) are performing best.

(I’ve stored my bitly.API_KEY and bitty.API_LOGIN in the bitly library in my python path. Get a copy of the library here.)

#!/usr/bin/python

import sys
import bitly

if len(sys.argv) < 2:
	sys.exit(2)
if len(sys.argv) == 3:
	if sys.argv[2] == 'facebook':
		short_url = bitly.shorten(sys.argv[1]+"?utm_source=facebook&utm_medium=social&utm_campaign=sm-direct")
	elif sys.argv[2] == 'linkedin':
		short_url = bitly.shorten(sys.argv[1]+"?utm_source=linkedin&utm_medium=social&utm_campaign=sm-direct")
	elif sys.argv[2] == 'twitter':
		short_url = bitly.shorten(sys.argv[1]+"?utm_source=twitter&utm_medium=social&utm_campaign=sm-direct")
else:
	short_url = bitly.shorten(sys.argv[1]+"?utm_source=twitter&utm_medium=social&utm_campaign=sm-direct")

sys.stdout.write(short_url)

Here’s a little python script that displays a user’s Klout score. It can be run from the command line like this:

[smbrown@smbrown ~]$ klout-score.py smbrown

Or you can add it as an extension to Alfred App if you have the Powerpack.

Invoke Alfred App, type the extension shortcut, twitter username and see the user’s Klout score in less time then it takes to open and load http://klout.com/username.

Marcel Pinheiro Caraciolo has built a nice Python interface to the Klout API, definitely look into using that for anything more complicated then this simple use case.

If the script below is on your path and you can run it from the command like above, then this Alfred App Powerpack extension should do the trick

-- Alfred App Extension:
#Desc: 	Retrieve user's Klout score
#Use:	ks <query>
#Ex:	ks smbrown

URL=$(klout-score.py "{query}")
echo -n $URL

-- Python script follows:
#! /usr/bin/python

import httplib
import simplejson as json
import sys

if len(sys.argv) < 2:
	sys.exit(2)

user = sys.argv[1]

klout_key = "I-am-an-idiot-for-posting-the-version-of-this-snippet-with-my-real-key"
api_url = "/1/klout.json?users=%s&key=%s" % (user, klout_key)
api_domain = "api.klout.com"

conn = httplib.HTTPConnection(api_domain)
conn.request('GET', api_url)
resp = conn.getresponse()
data = resp.read()
data = json.loads(data)

print data['users'][0]['score']

Jawbone Up detailed: tracks activity, food intake and sleep cycles, available November 6 for $100 (video) — Engadget.

At last, the company’s ready to talk specifics. We just got word the Up will be available in the US November 6th for $100, and will continue its world tour on the 17th. With that price, it’s well-matched against the clip-on Fitbit Ultra tracker, and indeed, they have some key features in common — namely, a step counter and a few requisite social networking features. But with a sensor that knows when you’re in deep sleep and a mobile app that can identify photos of food, it has a few unexpected tricks that could give devices like Fitbit a run for their money. We’ll be getting one to test very soon, but in the meantime, join us past the break to learn more.”

Mixpanel | Real-time Web Analytics, Funnel Analysis.

KISSmetrics Web Analytics – Event Tracking, A/B Testing and Conversion Funnel Software.

An example of the challenge in current search vs natural language processing:

What football games are on tomorrow?

I want to know: NFL vs soccer; TV schedule and channels; tomorrow is 10/30/2011; I am in PDT and would like to know TV listings in my local time; sorting by favorite team and/or team nearest to my current location.

Tested on G and bing. Bing result included a Yahoo answers result in top 5 that had the right answer, but: I just wanted the answer, not a list of search results; the ‘answer’ was crowdsourced/turked; the resulting web page will be out of date for the ‘tomorrow’ question in < 12 hrs.

Could Siri do this? Can IBM'd Watson do it.

What if this is possible via a new answer engine? What happens to billions of value in b-models that are search and search-results dependent?

Older Posts »