Sweeping Change

Although flying cars are still a long way off,  self driving cars are almost here. Several car manufacturers either have working prototypes or fully functioning production models in selling in other countries.  US models are expected to start selling as early as 2017. This will revolutionize the automotive industry but there are many more consequences,  both positive & negative, that many are not aware of.

With self driving vehicles knowing & obeying all of the traffic laws, law enforcement may have more time to focus on other serious tasks. Valets will have to look for other jobs as your car will drop you off, park itself & returned when summoned. Even personal injury attorneys,  the despised ambulance chasers, will have to look for clients elsewhere as traffic related injuries & death will decline.

ChromeBook - Is it worth it?

This past weekend I purchased a ChromeBook for my wife. It had 4 gigs of ram and an 16 gig SSD for $299. The hardware was solid, light weight and very quiet with a 14 inch screen. Setup was very easy, just sign in with your gmail account. There was no need to install Office as Google Docs were available by default. Boot up was under 20 seconds, very impressive.

I browsed the Chrome Web Store and installed several apps like Facebook, Twitter, Blogger, Drop Box, EverNote & RDP. The available apps both performed well and never crashed. My surprise was that the Chrome Store has a limited number of apps compared to the Android Play store. I understand these are 2 different operating systems but being both from Google, I expected more. Apps like Skype, Opera and Dolphin Browser were not available. In face, since the ChromeBook runs it apps inside the Chrome browser, no other browsers could be installed.

After 5 days, we returned the ChromeBook. The product would be good for the my mother who only sends email and browses the internet, but for a seasoned user, the functionality is lacking. If all of the apps you use are available in the Chrome Store, great buy a ChromeBook. Otherwise, save yourself the frustration and go with Windows or OSX.

MongoDB 2.1.1 / Raspberry Pi Install

Setting up & configuring your Raspberry Pi + Installing MongoDB 2.1.1.

You must first install the OS to your SD card.
Download the OS from http://www.raspberrypi.org/downloads
I use Raspbian Wheezy but there are other choices.
You'll need an installer as well to format the SD Card and install the OS.
Check out: http://www.raspberrypi.org/wp-content/uploads/2012/04/quick-start-guide-v2.pdf

Once you Pi is running, the follow the steps below.

1. From the config menu, select "Enable SSH"

2. Change your password

3. Expand root to fill SD Card & reboot

4. Test available space: df -k

5. To re-enter the config screen:
sudo raspi-config

6. Set timezone

7. Find the IP address from the router and SSH to the Pi from another computer.
ssh pi@192.168.1.XX

8. Set static IP address.
sudo vi /etc/network/interfaces

Change: iface eth0 inet dhcp
To:

iface eth0 inet static
address 192.168.1.XX
netmask 255.255.255.0
broadcast 192.168.1.255
gateway 192.168.1.1
network 192.168.1.0

Once the file is saved, you must reboot for the changes to take effect.
sudo reboot

9. Install VNC Server:
sudo apt-get install tightvncserver

This must be run after every Pi restart:
vncserver :1 -geometry 800x600 -depth 24

10. Connect using RealVNC Client

11. Shutdown and take a backup the SD Card.
sudo shutdown -h now


--- MongoDB Install 2.1.1 --------------------------------------------
(Thank you RickP for this install)

1. Reboot the Pi and login through ssh

2. Installing the supporting softwares. (This will take a few minutes)

sudo apt-get install git-core build-essential scons libpcre++-dev xulrunner-dev libboost-dev libboost-program-options-dev libboost-thread-dev libboost-filesystem-dev

3. Checkout this repo:
git clone git://github.com/RickP/mongopi.git

4. Build it (this will take very long! Hours!):

    cd mongopi
    scons

5. Install it:

    sudo scons --prefix=/opt/mongo install

6. Install mongo in /opt/mongo, to get other programs to see it, you can add this dir to your $PATH:

    PATH=$PATH:/opt/mongo/bin/
    export PATH

7. Create a directory for your MongoDB data files.

   cd
   mkdir mongoData
   cd mongoData
   pwd

8. Start MongoDB:
   ./mongod --dbpath /home/pi/mongoData

SQL Server Backups on Amazon EC2 Instances

If you have an EC2 Windows instance with SQL Server, you've probably noticed the SQL Server Agent missing. So how do we perform backups?

Option 1: Select each database in Management Studio, right click and run a backup, one database at a time. (Really? Does anyone have time for this)

Option 2: Use Transact-SQL. You can run the statement below, one command per database. This is quicker and easier than option 1 but it's still initiated manually.

backup database DATABASENAME to disk='c:\backup\FILENAME.BAK'

Option 3: Run a script to generate the Transact-SQL and run the commands with a scheduler. You can do this with any server side language with a SQL Server connection. The example below is written in Coldfusion.

<cfsetting requesttimeout="180">

<!--- Current Datetime --->
<cfset t = dateformat(now(),"yyyymmdd") & timeformat(now(),"hhmm")>

<!--- Delete backups older than this date --->
<cfset deleteDate = dateadd("d",-4,now())>

<cfset request.datasource = "CFDATASOURCE">

<cfquery datasource="#request.datasource#" name="backup">
   backup database DATABASENAME to disk='c:\backup\FILENAME_#t#.BAK'
</cfquery>

<!--- Delete old backups --->
<cfdirectory action="list" directory="c:/backup" filter="*.BAK" name="dirQuery">

<cfloop query="dirQuery">
   <cfif dateCompare(deleteDate,dirQuery.dateLastModified) eq 1>
      <cffile action="delete" file="c:/backup/#dirQuery.name#">

   </cfif>
</cfloop>

Connecting Python to SQL Server on OSx

After several hours and many frustrating errors, I was able to query SQL Server with Python. Previously, I went through several searches suggesting to install FreeTDS or find ODBC drivers. The steps were usually overly complicated and the instructions produced errors that send me down other tangents.

So here is it simply:
1.  Install pyodbc. (https://code.google.com/p/pyodbc/)
     Make sure the Python version and bit count (32 or 64) match your system.
2. Buy the driver from http://www.actualtech.com/
     I know, buy a driver? But it's $35 and I had it installed and running my test scripts in 5 minutes.
     Given my billable rate the client was happy to pay for it.
     If you're skeptical, you can install the driver for free with a few restrictions.

Here's a sample of my code:

import pyodbc

cnxn = pyodbc.connect('DRIVER={Actual SQL Server};SERVER=127.0.0.1;DATABASE=somedatabase;UID=username;PWD=password')
cursor = cnxn.cursor()

q = cursor.execute("select firstname, lastname from users")
row = cursor.fetchone()
if row:
    print(row)

Temp File Cleanup with Python

In the short few months I've been using Python, I've been impressed by it's versitility, performance and small footprint. Here is a quick of example of how to used Python to loop through a directory and delete old files based on the last modified date.

On Windows:

import os, time, sys
from datetime import datetime

path = r"c:\tempFiles" # Target Directory
logfile = r"c:\log\tempFileCleanup.txt"  # Log File
days = 10  # Delete files older than X days

now = time.time()

for f in os.listdir(path):
 full = path + '\\'
 full = full + f
 #print(full)
 #print(os.stat(full).st_mtime)
 if os.stat(full).st_mtime < now - days * 86400:
  if os.path.isfile(full):
   d = os.stat(full).st_mtime
   print(full)
   print(datetime.utcfromtimestamp(d))
   print(" ")
   os.remove(os.path.join(path, f))

# Create the log file if it does not exist
try:
   with open(logfile): pass
except IOError:
 target = open (logfile, 'w')
 target.write("\n")
 target.close()

# Write to the log file
with open(logfile, "r+") as f:
 old = f.read() 
 f.seek(0)
 f.write("Job Complete - " + str(datetime.utcfromtimestamp(now)) + "\n" + old)
   
print("All Done")

GPS with jQuery Mobile

Having never developed on a mobile platform before, I was very curious about using the GPS feature within a web page. Having seen a "request to access my GPS" when I viewed certain sites, I knew it was possible. After a little research I found it was much easier than I through thanks to jQuery mobile.

You must download jQuery and jQuery mobile. I also used the jQuery mobile css.

Click here for demo

<!DOCTYPE html>
<html>
<head>
<meta name=viewport content="user-scalable=no,width=device-width" />
<link rel=stylesheet href="jquery.mobile-1.3.0/jquery.mobile-1.3.0.min.css" />
<script src="jquery-1.9.1.min.js"></script>
<script src="jquery.mobile-1.3.0/jquery.mobile-1.3.0.js"></script>
</head>
<body>

<div data-role=page id=home>
  <div data-role=header><h1>Where Am I?</h1></div>

  <div data-role=content>
  <span> Status : </span> <span id=stat></span> <br />
    <span> Time : </span> <span id=time></span> <br />
<span> Latitude : </span> <span id=lat></span> <br />
    <span> Longitude : </span> <span id=lng></span> <br />
  </div>

  <input type="button" name="ss" value="" id="ss">

  <div id="out1" style="width:96%; height:200px; overflow:scroll; border:1px solid #000; margin-left:2%;"></div>
</div>

</body>
</html>

<script>
(function($) {

$.fn.changeButtonText = function(newText) {
return this.each(function() {
$this = $(this);
if( $this.is('a') ) {
$('span.ui-btn-text',$this).text(newText);
return;
}
if( $this.is('input') ) {
$this.val(newText);
// go up the tree
var ctx = $this.closest('.ui-btn');
$('span.ui-btn-text',ctx).text(newText);
return;
}
});
};

var runGPS = 0;
$("#ss").changeButtonText('Start');
$("#stat").text('Paused');

$('#ss').click(function() {
if (runGPS == 1){
$("#stat").text('Paused');
$("#ss").changeButtonText('Start');
runGPS = 0;
} else {
$("#stat").text('Running');
$('#ss').changeButtonText("Stop");
runGPS = 1;
}
});

var timer = setInterval(function(){
if (runGPS == 1){
navigator.geolocation.getCurrentPosition (function (pos)
{
date = new Date();
$("#time").text (date);

var lat = pos.coords.latitude;
var lng = pos.coords.longitude;
$("#lat").text (lat);
$("#lng").text (lng);

var t = date + " lat:" + lat + "lng:" + lng + "<br>";

$("#out1").append(t);

});
}
},2000);

})(jQuery);
</script>