If you wish to look at a specific exit status for a piped command you can use the $PIPESTATUS array which is the exit codes for each command in your pipe indexed by the position of the command in your pipe.

http://colorschemedesigner.com/

The above is great for picking coordinating colors.   You can adjust the overall brightness or saturation of the theme, the degree of contrast in the light or dark areas, whether you want monochromatic or coordinating colors etc.

It is a little heavy or sluggish but I still am a big fan of this implementation.

The total transaction volume we are anticipating for this is about 60 million transactions a year with an average response time, including all application overhead of 0.6 seconds.

We have built that. Now its on to phase two. We need to add redundancy. We are scoping out a second identical environment at a physically separate facility. To do that I will need to focus on replication of postgres. The large proprietary read only store can be manually refreshed and kept in synch easily. Postgres is the challenge. Not because of limitations in postgres but just because that is the db where all our changes will be made. If we go live on both sites simultaneously with load balancing I’ll have to configure and support bidirectional replication in some cases.

Can we do it ? Yes we can. But it will be fun to figure out how it all fits together.

lucias@blackwater:~/src/rails/testauth$ rake db:migrate
(in /home/lucias/src/rails/testauth)
rake aborted!
uninitialized constant User::Authentication

(See full trace by running task with --trace)
lucias@blackwater:~/src/rails/testauth$

What did all the above mean ?   Well the long and the short is that “include Authentication” in a module named …. was failing.   It needed to have the plugin installed and, for reasons of simple human error, I’d skipped that step.   To install the plugin I cc’d to the vendor/plugins/ directory and ran :

lucias@blackwater:~/src/rails/testauth/vendor/plugins$ git clone git://github.com/technoweenie/restful-authentication.git restful_authentication
Initialized empty Git repository in /home/lucias/src/rails/testauth/vendor/plugins/restful_authentication/.git/
remote: Counting objects: 1773, done.
remote: Compressing objects: 100% (617/617), done.
remote: Total 1773 (delta 1030), reused 1636 (delta 957)
Receiving objects: 100% (1773/1773), 738.50 KiB | 176 KiB/s, done.
Resolving deltas: 100% (1030/1030), done.
lucias@blackwater:~/src/rails/testauth/vendor/plugins$

Which installed the plugin, note I used an underscore and not a hyphen in the name I used locally.   That is important as on occasion the hyphen in the name can cause problems.

Anyway now all it works as follows.

lucias@blackwater:~/src/rails/testauth$ rake db:migrate
(in /home/lucias/src/rails/testauth)
==  CreateUsers: migrating ====================================================
-- create_table(:users, {:force=>true})
   -> 0.0191s
-- add_index(:users, :login, {:unique=>true})
   -> 0.0060s
==  CreateUsers: migrated (0.0253s) ===========================================

lucias@blackwater:~/src/rails/testauth$

All is well.

Opening this package or using the print cartridge inside confirms your acceptance of the following license/agreement.  The patented print cartridge and its contents inside this package are licensed for a single use only.  After this single use the license to use the print cartridge terminates, and the used cartridge must be returned only to Lexmark for remanufacturing, refilling or recycling.  Visit www.lexmark.com/recycle for return information.  IF YOU DO NOT ACCEPT THESE TERMS, RETURN THE UNOPENED PACKAGE TO YOUR POINT OF PURCHASE.  A replacement cartridge sold without these terms is available through www.lexmark.com

This is right up there with the licensed plants I saw at the nursery a while back which prohibited you from propagating them.

I’m speechless so I’ll not say anything other than “WOW”.

I decided to install subversion on a new linux box I recently migrated to at home.  As with anything in Ubuntu its a simple install.

sudo apt-get subversion

And follow the bouncing ball which installs the following ( and other supporting ) package :

subversion                                 1.6.6dfsg-2ubuntu1                              Advanced version control system

In my case I decided to setup a simple repository as follows
svnadmin create /home/lucias/lib/SVN

That created the repository in the SVN directory.

Then I created a simple test to validate and imported a rails application tree I have in the source directory

svn import isbn_google file:///home/lucias/lib/SVN -m “Initial import of code”

That worked fine. To validate I renamed the directory I had just checked in and then checked it back out from subversion.

svn checkout file:///home/lucias/lib/SVN isbn_google

Which laid the files down correctly for my working copy.   Then I modified a file and validated that svn was aware my working copy was revised

svn status

Next I checked to see if anyone else had revised the file while I had it checked out with

svn update

Since I am the only user no one had made revisions   So with that taken care of I then

svn commit -m “Fallback to text_field to create working code”

Some quick validations showed the logs of changes were there

svn log

This was run on various files and directories.

In short thats the core of Subversion, naturally much more to it but that was enough for my quick validation the install was working fine.

First install the pg gem.   I installed pg 0.9.0.

gem install pg

You may need to install some other gems as prerequisites or and an extra package or two at the Linux OS level.   Once done you then simply edit your config/database.yml file in your rails application directory.   Here is how I setup my development environment to use Postgres :

development:
  adapter: postgresql
  database: railsdb
  user: railstest
  password: *****
  host: localhost
  port: 5432

Obviously **** isn’t my password but you get the idea.
Now when you run a rake db:migrate the changes are made in Postgres.

There are many ways to do this.   In this method I use the DB2 drivers and depend on the fact that we have already cataloged the location of the database name, thus mapping to a specific IP/Port/Database name.  

First you need to add the DB2 drivers to your referenced libraries.   Right click on the “Referenced Libraries” under your project then select “Build path->Configure Build Path”.     Select Add External Jars on the Libraries tab of the dialog.   My file was in C:\Program Files\IBM\Sqllib\java\db2java.zip.

Then I ran the following java program ( formatting didn’t paste correctly due to tab issues ):

import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.PreparedStatement;
import java.sql.SQLException;

public class DB2ConnectTest {

	public static void main(String[] args) {

		Connection con1;
		String theSql;
		String dbName;
		String userName;
		String userPassword;
		PreparedStatement schemaQuery1;
		PromptUser prompt = new PromptUser();

		// Get the input
		dbName = prompt.GetInput("Database Name");
		userName = prompt.GetInput("User Name");
		userPassword = prompt.GetInput("User Password");

		String url1="jdbc:db2:"+dbName;     	     

		try{

			Class.forName("COM.ibm.db2.jdbc.app.DB2Driver");

	       	con1 = DriverManager.getConnection ( url1, userName, userPassword ) ;
	       	theSql = "select distinct tabschema from syscat.tables order by tabschema";
	       	schemaQuery1 = con1.prepareStatement(theSql) ;
	       	System.out.print("It worked");

	       	// We don't do anything with prepared statement
	       	// We just validate we can connect and prepare ok.

	    } catch (SQLException e) {

	       	e.printStackTrace();

	    }catch ( java.lang.ClassNotFoundException e ){

	      	System.err.println("Exception "+e.getMessage() );

	    }
	}
}

I am not really able to blog about the administration and coding I perform at work due to restrictions.   But I am able to blog about what I do in my spare time relative to development, so that I will.

IT isn’t the only thing I do in life, but it is the main thing I will blog about here.

Technical books are not the sort of thing one keeps around.  They are obsolete pretty quickly.    You don’t tend to reread them either.  Once you have learned what they have to say I find myself using google to find specific answers.   But while one learns a topic I find a well presented book the best way to learn.

I am very partial to Oreilly books.   In fact if there is an Oreilly book on the topic I almost never buy a different one.   They are far more than a giant tome of all the documenation.   They present the matter in a way that I understand quickly.   No I’m not on comission for them.