Tim O'Brien

Home Care Payroll Management

2014-07-02T12:07:00.000-07:00

As the population gets older, more and more people are requiring home care services. Government Health & Social Service departments in Canada often cover all or some of the cost associated with services provided.

However, these Government departments are not the employers of home care workers and neither are the home care agencies. The individual person requiring the service is often the employer. This greatly complicates the processing of payroll, instead of having a couple of employers with many employees, you now have just as many employers as employees.

Trying to manage a fluctuating mass of hundreds of employers and their current authorizations for invoicing third-parties can quickly become overwhelming, especially for traditional payroll systems.

Payroll Invoicing handles the entire process in a single, easy-to-user client application. Setup as many employers & employees as you want, associate your employees with many different employers and your ready to start generating invoices.

Payroll Invoicing
www.payrollinvoicing.ca
+Payroll Invoicing

For more information contact:
Tim O'Brien
timobrien@payrollinvoicing.ca

Cross Platform Game Development with PlayN - Architecture

2013-07-22T10:39:00.000-07:00

Since PlayN is a cross-platform framework I wanted to have a shared cloud service to provide a data storage for game scores, level data, etc... I would use this service to allow users to design custom levels on the PC via their browser, and publish it for play on any other device running the game.

I create a Google App Engine project called Play-Plus (www.gplayplus.com). The data architecture allows for the hosting and data storage of an unlimited number of games. I plan to reuse this framework for ever game I publish, the service will be expanded to allow user in-app purchases and turn\real-time multi-player action.

Once completed, I will provide an SDK for other developer to use. Initially I explored using an existing API, such as ScoreLoop, Facebook, etc... However, it required me to configure leader boards for each level, etc... Since levels can be created dynamically, I needed leader boards, rating, etc... to all be created instantly when a level was published.

I have separated the java classes for PlayN - Play Plus which I will package into a JAR, it will include the play plus dialogs which use TriplePlay. This will be used inside the core project so there are no platform dependencies. Additional JAR files per platform will connect the core service interfaces with the individual optimize platform implementations.

Next time I will discuss resources to read to get started developing your own game.

Until next time, I would greatly appreciate a like on Facebook or play on Facebook to help it get published in their App Centre.

Join the G+ community for Dirty Worms

View all Kyght Game Releases at http://kyght.blogspot.ca/

Cross Platform Game Development with PlayN - Concept

2013-07-04T11:35:00.000-07:00

The first game I developed was very simple and used the LibGDX framework, it was good experience but I wanted to create a game that was more complex and allowed me to build onto a story. At the time LibGDX only supported Android & Java.

While searching for possible frameworks to use, I came across PlayN which offered Android, iOS, HTML5, Java & Flash. Not to mention the fact that is was developed by people at +Google and used by Rovio for Angry Bird Chrome.

I love 2D physics games and I really wanted to create one. My game concept was born from a really awesome movie which included a cast of legendary actors. However, I needed to make them fictional characters, so I made them worms and of course their enemy is birds.

The idea of what types of birds they would be didn't come until much latter. I live in Newfoundland... and we have lots of seagulls. This spring they were all around my house, perched on the roof, about 20 of them and relieving themselves quite often!! For about 2 week, we would awake to the sound of these gulls, squawking on the roof. Ohhh, how I wanted to get rid of them PERMANENTLY!! However, seagulls are a federal protected species, errrr. So I couldn't kill them in real life, but I can in the game :)

Now with my concept started, I begin into programming. There is a lot to discuss so I'm going to create a post for each sections of the game I've programmed, +Box2D , +Google Cloud Platform integration, graphics, ageing particle systems, triple play, licensing, Facebook integration, and platform specific code.

Until next time, I would greatly appreciate a like on Facebook or play on Facebook, seems Facebook doesn't list your game in App Center unless you already have people liking and playing it.

Join the G+ community for Dirty Worms

View all Kyght Game Releases at http://kyght.blogspot.ca/

Using LINQ to SQL with SQLMetal & WPF

2013-05-16T09:31:00.000-07:00

Introduction

I had an opportunity to evaluate LINQ for use in a production application and was surprisingly impressed. Microsoft often tends to release new versions of Visual Studio & .NET framework that contain few real enhancements to improving the speed efficiency of developing a real world application. The combination of LINQ, SQLMetal & WPF is a major improvement for developing data-driven application.

My next blog post will be related to using SQLMetal to generate serializable classes for use with Window Communications Foundation (WCF). This will be very exciting to investigate and on the surface appears almost completely transparent to the rest of the client application and it's WPF bindings.

Database Setup

First you have to setup your database model before you can use SQLMetal to generate your ORM layer. Create your tables, primary keys, indexes and foreign keys. It's very important to setup your foreign keys as they will be used to generate your object relations in SQLMetal, making it super easy to reference related data from a parent record (Ex. Employee.Province.Code)

SQLMetal Generated ORM Layer

Now we can use SQLMetal to generate the ORM layer to use in the application.

SQLMetal

The SQLMetal file is included in the Windows SDK that is installed with Visual Studio. By default, the file is located at drive:\Program Files\Microsoft SDKs\Windows\vn.nn\bin. If you do not install Visual Studio, you can also get the SQLMetal file by downloading the Windows SDK

For complete usage information of SQLMetal visit Microsoft's website.

My initial command line usage of SQLMetal is pretty simple.

sqlmetal /server:localhost /database:employee /code:employee.cs

Solution

My solution file consists of a WPF application and a Class Library. Add your SQLMetal generated class file to your Class Library and have you WPF application reference that Library.

Connecting to your database from inside your WPF application is as simple as creating your database object with a supplied connection string.

DB = new Employee ("server=localhost; database=employee; Trusted_Connection=True;");

Using LINQ

Now accessing your databases objects in code is super simple. The code sample below retrieves a list of all provinces in the database. The Provinces class is generated by SQLMetal and has properties that correspond with the underline table.

var provincelist = new List<Provinces>(from p in DB.Provinces select p)

WPF Bindings

While accessing the database from code is great, often data changes are performed by a user. So displaying the data, accepting user changes and saving back to database is key. This is where WPF bindings save a lot of time.

In your C# code, get a list of employee's

var employees = from c in db.Employee select c

Pass an employee object to a UserControl for displaying, set the UserControl.DataContext property to the employee object returns from the LINQ call.

this.DataContext = employee;

The XAML below showing binding controls on a UserControl to the employee object.

<TextBox Text="{Binding Path=First}" Height="23" HorizontalAlignment="Left" Margin="0,10,5,0" VerticalAlignment="Top" MinWidth="150" Width="auto" />

<ComboBox Name="provList" DisplayMemberPath="Code" ItemsSource="{Binding Source={x:Static list:GlobalList.ProvinceList}}" SelectedItem="{Binding Path=Provinces}" Width="75" Height="23" HorizontalAlignment="Left" Margin="5,0,0,0" VerticalAlignment="Top"/>

A Text box to display the employee's first name is pretty easy, just bind the Text property to the First property on the employee object.

The combo box is was a little troublesome at first because I was setting the Employee.Province (int) using SelectedValuePath. While this was correctly setting the Employee object I ran into a LINQ issues.

The member 'X' was changed to be inconsistent with the association member 'Y'

This was because the Employee.Provinces (Class) object was still null even though I set the foreign key value. The solution was to set the SelectedItem and let LINQ take care of synchronizing the foreign key value.

This was also true for ComboBox columns inside DataGrid's, I used the following XAML to correctly bind my DataGrid drop-down lists.

<DataGridComboBoxColumn
Header="WorkingType"
DisplayMemberPath="Name"
SelectedItemBinding="{
Binding Path=WorkTypes,
Mode=TwoWay,
UpdateSourceTrigger=PropertyChanged}">

<DataGridComboBoxColumn.ElementStyle>
<Style TargetType="{x:Type ComboBox}">
<Setter Property="ItemsSource"
Value="{Binding Source={x:Static list:GlobalList.WorkTypesList}}" />
</Style>
</DataGridComboBoxColumn.ElementStyle>
<DataGridComboBoxColumn.EditingElementStyle>
<Style TargetType="{x:Type ComboBox}">
<Setter Property="ItemsSource"
Value="{Binding Source={x:Static list:GlobalList.WorkTypesList}}" />
</Style>
</DataGridComboBoxColumn.EditingElementStyle>

</DataGridComboBoxColumn>

Saving Data

Now for the easy part, saving the data back to the database after the user has altered it using the WPF controls.

This one statement applied all inserts, updates and deletes back to the database.

DB.SubmitChanges();

So the above handles querying data and editing it, but what about insert, this is also very easy. Add a new button to your code and call the following function.

employee = new Employee();
DB.Employee.InsertOnSubmit(employee);
UserControl.DataContext = employee

This code add a new employee, allows the user to edit it and when DB.SubmitChanges() is called, the record will be inserted. RemoveOnSubmit is also available for deletes.

Conclusion

Microsoft appears to have a real winning combination here, just imagine its usage with WCF. Creating a responsive data driven application for clients is fast, stable and contains little code. The biggest challenge so far seem to be using the correct WPF bindings for the LINQ generated objects.

Google AppEngine Search API

2013-01-31T12:10:00.000-08:00

Introduction

In the article Geospatial Queries with Google App Engine I discussed my initial attempts to preform geospatial queries using +Google App Engine. The described how I arrived at my solution but didn't provide any details on my implementation.

Datastore Entity vs Scored Document

I store all my data in Datastore entities, they are easy to save, fetch and run queries on. So when I first started to read about the search API and having to create a document for indexing I felt there was a lot of overlay with entities.

I had to build a document with fields I had already saved in an entity, I initially created a document with only the fields I required for searching by location along with keys to my saved entities so I could fetch them. Each time I added\updated an entity I called my index function to keep my index current with my entities.

IndexSpec indexSpec = IndexSpec.newBuilder().setName("locIndex").build();

Index idx = SearchServiceFactory.getSearchService().getIndex(indexSpec);

Locdoc = Document.newBuilder()

.addField(Field.newBuilder().setName("id").setText(eid))

.addField(Field.newBuilder().setName("parentid").setText(RegId))

.addField(Field.newBuilder().setName("regid").setText(RegId))

.addField(Field.newBuilder().setName("active").setText(Active))

.addField(Field.newBuilder().setName("location").setGeoPoint(geoLoc))

.build();

idx.put(Locdoc);

Geospatial Index Search

After indexing my entities with the new search api, I used it to search for entities with a specific distance from a geopoint and it worked great.

String queryStr = "(active:1) AND

distance(location, geopoint(" + lat + ", " + lon + ")) < " + userad;

Results<ScoredDocument> results = getLocIndex().search(queryStr);

Initially, I would get the entity key from the document and fetch the entity using the key. However, I quickly realized this was a bad idea. I check my dashboard quotes and discovered I was hitting two quotes categories "Datastore Read Operations" and "Search API Calls".

Optimization to minimize quote impact

Since I had to use the Search API to locate data by location, I decided to examine all location based searches in my app and encode into the indexed document all the fields I required to handle a user response.

This resulted in much lower Datastore Read Operations and an increase in the response time. Remember to limit the amount of data you push into your indexed document because there are limitations. See Search API quote.

Thanks for reading, follow me on G+ at +Tim O'Brien

+Kyght
www.kyght.com
tobrien@kyght.com

Note: The above icon is the property of Google and is only here to provide readers with

a visual recognition of the Google product.

Geospatial Queries with Google App Engine

2013-01-18T10:24:00.001-08:00

Introduction

While developing an Android app that uses Google Cloud Messaging and integrates with +Google App Engine, I ran into a common App Engine issue. Geospatial queries using High Replication Data Store. In my application I connect users via their location, when a user performs an action in the app, users in their selected location receive a message via GCM.

DataStore vs Rational Databases

Coming for the rational database world of Oracle and SQL server, I attempted to execute a query to locate devices inside the users selected area. I used a bounding box to determine what devices are within a given area. Pleased with my approach, I constructed my filter and updated to App Engine for execution.

query.setFilter( CompositeFilterOperator.and(

new FilterPredicate(COFFEERUN_ACTIVE_PROPERTY, Query.FilterOperator.EQUAL, 1),

CompositeFilterOperator.and(

new FilterPredicate(COFFEERUN_LAT_PROPERTY,

Query.FilterOperator.GREATER_THAN,
boundbox[0].getLatitudeInDegrees()),

new FilterPredicate(COFFEERUN_LAT_PROPERTY,
Query.FilterOperator.LESS_THAN,
boundbox[1].getLatitudeInDegrees())
),

CompositeFilterOperator.and(

new FilterPredicate(COFFEERUN_LON_PROPERTY,

Query.FilterOperator.GREATER_THAN,
boundbox[0].getLongitudeInDegrees()),

new FilterPredicate(COFFEERUN_LON_PROPERTY,
Query.FilterOperator.LESS_THAN,
boundbox[1].getLongitudeInDegrees())

)));

BOOM!!! Exception (IllegalArgumentException: Only one inequality filter per query is supported)

What! this can't be right, this type of query executes all the time in Oracle and SQL, lets check the documentation.

Google Developers - App Engine - Datastore Queries

A query can have no more than one not-equal filter, and a query that has one cannot have any other inequality filters.

Research & Development

Ok, that's fine, time to do some research, GeoHashing seems to be the answer and the GeoModel project has an App Engine implementation. However, it's a lot of work to integrate and I still have to read my matches to determine if they are within my search area.

There must be something in native App Engine to handle this type of query, after all Google does it all the time. I find out there is a native GeoPt datatype in App Engine, I will try using that first to see if it works.

I update the code to store GeoPt values for device locations and change my query to use the new GeoPt value, I upload my code to App Engine and give it a try.

GeoPt geoLocG = new GeoPt((float)boundbox[0].getLatitudeInDegrees(),
(float)boundbox[0].getLongitudeInDegrees());

GeoPt geoLocL = new GeoPt((float)boundbox[1].getLatitudeInDegrees(),
(float)boundbox[1].getLongitudeInDegrees());

query.setFilter( CompositeFilterOperator.and(

new FilterPredicate(COFFEERUN_ACTIVE_PROPERTY, Query.FilterOperator.EQUAL, 1),

CompositeFilterOperator.and(

new FilterPredicate(COFFEERUN_GEO_PROPERTY,

Query.FilterOperator.GREATER_THAN, geoLocG),

new FilterPredicate(COFFEERUN_GEO_PROPERTY,
Query.FilterOperator.LESS_THAN, geoLocL)
)));

Iterable<Entity> entities =
datastore.prepare(query).asIterable(DEFAULT_FETCH_OPTIONS);

Yes it worked, it didn't raise an exception and I get some results in my rather large area query. I move on and go back to working on the Android app.

Then during testing I find that I'm not receive some GCM messages and other devices are getting messages that are outside the visual search area on the map. I then start adding integration with Google Places API and custom POI data from various US cities. I set my testing environment location for my devices to San Francisco, then it becomes very clear. I'm getting back entries that are outside my query area.

I take each matching entries geographical location and plot it using +Google Maps , I add my bounding box and quickly see there is an issue. I guess the lesson learned here is, just because it executes doesn't mean it worked.

Conclusion

So the question is "When searching the datastore by location, you can not use an inequality filter for latitude/longitude. Many older articles suggest using geomodel, however with the newer GeoPt datatype, does datastore correctly handle greater than/least than on GeoPt's or does geomodel still have to be used for location queries?"

Thankfully I recently joined +Google+ , and was able to ask that very question to the +Google Cloud Platform Developers . If you want to view the post directly, click here.

They suggested I use "App Engine Search API (Experimental)", which comes with the following note

Experimental!
Search is an experimental, innovative, and rapidly changing new feature for Google App Engine. Unfortunately, being on the bleeding edge means that we may make backwards-incompatible changes to Search. We will inform the community when this feature is no longer experimental.

I changed all my code to use the Search API and it works awesome. I double checked all matches using +Google Maps and it was perfect.

So if your looking to perform location based searches, have a look at +Google App Engine Search API.

Here is a link to the Java Search API reference doc "Search API - Performing Location Based Searches"

Let me know if you have any issues or need more info, you can find me on +Google+ at +Tim O'Brien

Thanks for reading.

+Tim O'Brien
+Kyght
www.kyght.com
tobrien@kyght.com

Note: The above icon is the property of Google and is only here to provide readers with

a visual recognition of the Google product.

Source Documentation

2011-11-22T08:55:00.001-08:00

Documentation... The most hated word in software development. Programmers can't stand doing documentation and will avoid it at all cost.

So, it's best to make it easy for them to do, and conduct code reviews to ensure it's being completed. I'm not talking about source code comments regarding the logic, I'm talking base classes, interfaces, and components that are used everywhere and demand a specific usage.

One of the biggest reasons why developers don't do this is because source code and documentation are often in two different location and they always get out-of-sync so why invest the effort. This is a very valid reason, however as I mentioned in my post about retaining senior staff this can cause major issues down the road.

It's best to keep your source code and documentation in the exact same spot, this will allow developers to quickly and easily update the documentation based on changes in the source code. This also allows the documentation to be versioned along with the source changes. When other developers have to use or update your code, there is clear documentation of your intentions.

Visual Studio was updated to produce xml formatted docmentation based on special comment tags beginning with \\\ will be processed. You can also use special <XML> tag and have your output formatted differently.

In your source code, MyClass.cs, you can add the \\\ comment above the class, interface, method, etc... that you want to document.

/** <summary>

     * ITranslator provides methods for callers to get a translation. Implementers of ITranslator have to
     * preform the translation requested.
     * </summary>
     * <example>
     *      When receiving a ITranslationItems object when implementing Translate the sender can be
     *      enumerated as shown below.
     *      <code>
     *          foreach(var item in sender.TranslationItems()) {
     *
     *              // Get translatable strings from the class
     *              //implementing ITranslationItems
     *
     *              String lbls = item.Label;
     *
     *              if (lbls == null) continue; //No Label set
     *
     *              // 3. perform translations
     *              lbls = map.GetTranslation(destinationlang, lbls);
     *
     *              //Set Translated Label to be used
     *              item.Label = lbls;
     *          }
     *        </code>
     *     </example>
     */
    public interface ITranslator {

Update your project file to output xml documentation and source code lines beginning with \\\ above will be generated with your assembly myCode.dll and myCode.xml.

Below is a sample of the output generated from Visual Studio.

<?xml version="1.0"?>

<doc>
    <assembly>
        <name>myCode</name>
    </assembly>
    <members>
        <member name="T:Translation.NamespaceDoc">
            <summary>
            The Translation namespace is used to by application link objects with items requiring translation to avialable translation providers. Providers must implement the iTranslator interface.
            </summary>
        </member>
        <member name="T:Translation.ITranslatable">
            <summary>
             ITranslator provides methods for callers to get a translation. Implementers of ITranslator have to preform the translation requested.
             </summary>
             <example>
                  When receiving a ITranslationItems object when implementing Translate the sender can be enumerated as shown below.
                  <code>
                      foreach(var item in sender.TranslationItems()) {
                          // Get translatable strings from the class
                          // implementing ITranslationItems

String lbls = item.Label;

                          if (lbls == null) continue; //No Label set

                          // 3. perform translations
                          lbls = map.GetTranslation(destinationlang, lbls);

//Set Translated Label to be used

item.Label = lbls;

}

                    </code>
                 </example>
        </member>
...

An XML formatted document is hardly usable documentation, and if you only have Visual Studio then that's all you get. You have to download some additional applications to package your work into a pretty and professional outputs. Luckly it's not going to cost the company a dime, first download Sandcastle, it will compile your xml and process your assembiles to generate a great documentation in a number of formats. Then download Sandcastle Builder because sandcastle doesn't have a GUI and it's much easier to setup a project using Sandcastle Builder.

Create a new project using Sandcastle Builder and add documentation sources, your assemblies and your generated xml files. There are four supported HelpFileFormats available, HtmlHelp (.chm), MSHelp, MSHelpViewer (Visual Studio integrated help), and Website.

After you compile your project you will end up with a great looking help file.

Now you can distribute your source documentation to your team, or even better yet!!! Include the documentation building in your build process. Add the following lines (with your paths and filenames) to your CruiseControl.NET installation (ccnet.config) and generate versioned documentation with your versioned application.


<exec>
   <executable>C:\Windows\Microsoft.NET\Framework\v4.0.30319\MSBuild.exe</executable>
   <baseDirectory>C:\Dev\DeveloperDocs\Sandcastle\Projects</baseDirectory>
   <buildArgs>/p:Configuration=Debug CoreFramework.shfbproj</buildArgs>
   <buildTimeoutSeconds>0</buildTimeoutSeconds>
</exec>

<exec>
   <executable>C:\Program Files (x86)\Inno Setup 5\ISCC.exe</executable>
   <baseDirectory>C:\Dev\DeveloperDocs\Sandcastle\Projects</baseDirectory>
   <buildArgs>CoreFramework.iss</buildArgs>
</exec>

Note: CoreFramework.iss is an InnoSetup file, generate your own using the built-in script wizard.

I hope this article was helpful, let me know if you have any questions.

Retaining Senior Staff (Software)

2011-11-18T07:25:00.001-08:00

One of the biggest issues facing many companies is resource retention. I don't believe many companies understand the long term cost of losing a resources, especially senior resource that have been with the company for years.

Product development is simple, write some code and make it work as expected, right? No!!! if your a company that is developing a product that you intent to market and update for years to come, then it's not that simple. Version 1 of the product development cycle is the easiest, there is no existing code base, no existing issues, and time pressures are not customer driven.

Now fast forward 5 years since the release of version 1, with a quarterly release cycle you now have released 20 versions of your product (5 major and 15 minor revisions). In additional to this you have responded to customer demand and released 50 patches of various versions. The source code has been modified by a team of 10 developers (2 seniors, 5 intermediate, and 3 junior). However, it's not the original team you started with, you have had 3 senior, 4 intermediate, and 6 junior developers come and go since you started.

Now you only have 1 senior and 2 intermediate developers that were part of the original team, one day your last senior resource puts in his notice, shortly afterwards 1 of your intermediate resource also gives notice. Unless you have well defined processes and clearly defined resource roles and procedures then your in trouble.

Your long term resources have likely taken on additional work because they loss the resources they trained and depended on for various tasks. They have created scripts, custom exe's, and utilize 3rd-party tools to complete their work load on time, these tools only exists on their computer, the source is not checked into source control and only they know the procedures to use them.

You try to offer your last senior resource some more money to stay but it's too late, the job they are going to has more money, a smaller work load, and more benefits. Your other resources try to hold down the fort while you get another senior resource, you change your release cycle to annual and can only manage minor releases and patches. Your team is struggling and pressures are growing, resources beginning to leave more frequently.

Your company is starting to get a bad name due to turnover, you have to increase your salaries way above your competition. You let your junior resources go to cover the increase in costs. Now with your increased salaries you manage to hire 3 senior and retain 2 intermediate resources. Your new senior resources complete a review of the source code base and inform you that a rewrite needs to be completed. How could this be, the source has been maintained and updated for 5 years. You allow the rewrite because you have no other choice.

You begin to think that your previous senior resources must have not know what they were doing. However, they were not the problem, you were!!! You allowed this to occur! Do you know what happens on a daily basis? What activities are each resource is responsible for? What procedures to they use to complete their assigned tasks?

You should have required all resources to document each activity they preformed! All source code should be well documented! Main activities such as builds should be preformed on separate and controlled environments! Developer documentation of frameworks and products should be maintained so that new staff could understand existing code and not have to rewrite new code!!!

On the human resources management side, when you conducted your annual reviews with your employees did you really listen to what they were saying? Did you ask the right questions? Have you improving the work environment? Do you allow time flexibility? Ability to work remotely? Is vacation time enough for changing employees who now have families? Did you allow salaries to remain stagnate?

Employee retention is not about always increasing salaries, and companies that operate REAL pension plans have a massive advantage over other companies. As employees grow older they have families, expenses grow, they begin to plan retirement. There was a time when companies treated employees like family and looked after them, this created very loyal employees. Over the years many companies focused on rapid and massive growth, employees became numbers and when growth slowed the work force was cut without a second thought. Employees in turn became less loyal and no longer stay with a company for very long.

Not everything is about money, employees who are treated with respect, given help when needed, and have their long term objectives supported will likely remain with that company for years. If they move on for other reasons you may find them returning.

The most critical task is to not let your company fall into risk, don't let risk build. If you lack the knowledge then hire a right hand man and treat him as such. Capture what your business does, who does what tasks, how do the tasks get completed. Define and document your processes and procedure, store & version them in a document repository, and stay on top of your businesses daily operations. Once you have all the information, use it to remove risk, automate major activities, host them in environments that you control, take away the guess work when completing task so that employee turnover has less of an impact.

nANT C# Assembly Labeller (CruiseControl.NET)

2011-11-15T09:03:00.001-08:00

If your using CruiseControl.NET for your build automation then your going to implement it's biggest advantage: Controlled and consistent version labeling.

The example below shows hows to add version labeling to CruiseControl.NET and then access the label in nANT for processing AssemblyInfo.cs files.

Step 1: Add the CruiseControl.NET assemblyVersionLabeller to your ccnet.config file. This will provide access to the CCNETLabel environment variable for the rest of your project.

NOTE: In the example below, major, minor and build are hard coded into ccnet.config file to provide a single point of control, revision will be generated from the source control system.

CCNET.CONFIG

<labeller type="assemblyVersionLabeller">

       <major>5</major>
       <minor>6</minor>
       <build>2</build>
       <incrementOnFailure>false</incrementOnFailure>
   </labeller>

Step 2: Create your nANT script to search and update all AssemblyInfo.cs files in your source code. You will have to update the <items basedir="..\..\"> based on the location of your nANT script and your source code.

nANT_AssemblyLabeler.build
<?xml version="1.0"?>
<project name="setversion" default="setversion">
<target name="setversion" description="Sets the version number to CruiseControl.Net label.">

        <references>
            <include name="System.dll" />
        </references>
        <imports>
            <import namespace="System.Text.RegularExpressions" />
        </imports>
        <code><![CDATA[

[TaskName("setversion-task")]

             public class SetVersionTask : Task
             {
              protected override void ExecuteTask()
              {
               StreamReader reader = new StreamReader(Project.Properties["filename"]);
               string contents = reader.ReadToEnd();
               reader.Close();

string replacement = "[assembly: AssemblyVersion(\"" + Project.Properties["CCNetLabel"] + "\")]";

string newText = Regex.Replace(contents, @"\[assembly: AssemblyVersion$"".*""$\]", replacement);

string replacement2 = "[assembly: AssemblyFileVersion(\"" + Project.Properties["CCNetLabel"] + "\")]";

string newText2 = Regex.Replace(newText , @"\[assembly: AssemblyFileVersion$"".*""$\]", replacement2);

System.Console.WriteLine("Updating Version: " + Project.Properties["CCNetLabel"] + ", AssemblyFile : " + Project.Properties["filename"]);

StreamWriter writer = new StreamWriter(Project.Properties["filename"], false);

writer.Write(newText2);
writer.Close();

}

             }
             ]]>
        </code>
    </script>

        <in>
            <items basedir="..\..\">
                <include name="**\AssemblyInfo.cs"></include>
            </items>
        </in>
        <do>
            <setversion-task />
        </do>
    </foreach>
</target>
</project>

Step 3: Add the following nANT task inside your <tasks> node before compiling the source code to update all AssemblyInfo.cs files.

CCNET.CONFIG

<nant>

    <executable>C:\AutomateBuild\nant-0.91\bin\nant.exe</executable>       <baseDirectory>C:\Source_Trunk\ReleaseManagement\Scripts</baseDirectory>
    <nologo>false</nologo>
    <buildFile>nANT_AssemblyLabeler.build</buildFile>
    <buildTimeoutSeconds>6000</buildTimeoutSeconds>
</nant>

Build Automation

2011-11-09T05:46:00.000-08:00

One of the most important parts of any development shop is release management, yet it's amazing that most companies have a fly by the seat of your pants approach. As developer we create our own misery by ignoring this critic task. Nobody cares that the code works on your computer; if there is a problem it should be easy to fix.... AND!!! fixing the issue isn't adding a line or two of code, checking it in and praying.

Release Management has more to do with outputs rather then code. Constantly producing the same well tested product using the same method each time and tracking all code, symbols, and binary related to that output is critical.

Fortunately there are tools available to assist developers with full build automation.

NOTE TO PRODUCT\PROJECT MANAGERS: DO NOT!!! expect a junior or intermediate resource to complete this complex task. Also, this is not a one time task it's a job role to maintain the build process. Changes to the output product that require modifications to the build process have to be completed as required.

Like most companies these days, nobody wants to pay money on time, resources, and products if they feel they have been getting along just fine without a real release management process. One of the best tools available is CruiseControl (also CruiseControl.NET). I will be discussing the .NET version of CruiseControl available at http://www.cruisecontrolnet.org/projects/ccnet

Companies will still have to spend money on a senior resource and the time it takes to setup a continuous integration server, however, time spend on packaging of future releases should be 50 - 70% less. This allows a company to reevaluate their current release schedule and produce releases more frequently. Another major bonus is it removes any possible dependencies on specific resources that have limited availability.

CruiseControl.NET is easy to install and config, the installer is a simple next... next... done process. CruiseControl.NET installs as a windows service and also installs a web dashboard to monitor and control the service (Ensure IIS is already installed on the target computer). Another consideration is security, CruiseControl.NET by default installs under the local host process. I recommend establishing a network domain account and have your administration grant rights to network resources, under windows services, find CruiseControl.NET, click properties, and update the LogOn tab with the network username and password. This will also CruiseControl.NET to operate as a separate identity, publishing outputs to network drives, send status emails, getting\commiting with source control, etc...

CruiseControl.NET's main configuration file "ccnet.config" is installed in the "server" sub folder. This is where you config all your project, the primary basics of you project are

Defined your project - name, description, category, working directory
Source Control - type of source control, locations, security
Versions Labeler - type of labeler, major, minor, build
Task - building source code, copying files, etc...
Publishing - logger, destination locations, email

The url below contents information on all the possible xml tags available and how to config each one.
http://www.cruisecontrolnet.org/projects/ccnet/wiki/ConfiguringTheServer

Here is a good example of a basic project configuration using SVN, Visual Studio, InnoSetup.
    - Pulls code from SVN
    - Defined a version 1.5.3 (The SVN revision number is appended, Ex: 1.5.3.3873)
    - nANT script to apply CruiseControl.NET version to AssemblyInfo.cs files
    - Compiles debug and release mode using Visual Studio 2010
    - Executes some batch command to move newly compiled files into release structure folders
    - Compiles release into self extracting installer using InnoSetup with a digital signs .exe
    - Publishes installer to network drive under the version folder \1.5.3.3873\MyInstall.exe

(Note: Project compiles and saving symbol files for future support issue, the source code is also tagged to allow remote debugging of issues on non-developement environments.)

<cruisecontrol xmlns:cb="urn:ccnet.config.builder">

<project name="MyProduct">
      <category>Products</category>
      <description>My awesome product is awesome</description>
      <webURL>http://localhost/ccnet/MyProduct/ViewLatestBuildReport.aspx</webURL>
      <workingDirectory>C:\Dev\MyProduct</workingDirectory>

<sourcecontrol type="svn">
      <executable>C:\Program Files (x86)\CollabNet\Subversion Client\svn.exe</executable>
      <trunkUrl>https://192.168.1.2/MyProduct/trunk</trunkUrl>
      <workingDirectory>C:\Dev\MyProduct</workingDirectory>
      <username>svnUsername</username>
      <password>svnPassword</password>
      <tagOnSuccess>true</tagOnSuccess>
      <tagBaseUrl>https://192.168.1.2/MyProduct/tags</tagBaseUrl>
</sourcecontrol>

<labeller type="assemblyVersionLabeller">
         <major>1</major>
         <minor>5</minor>
         <build>3</build>
         <incrementOnFailure>false</incrementOnFailure>
</labeller>

<tasks>

         
          <nant>
                   <executable>C:\Dev\Tools\nant-0.91\bin\nant.exe</executable>
                   <baseDirectory>C:\Dev\MyProduct\ReleaseManagement\Scripts</baseDirectory>
                   <nologo>false</nologo>
                   <buildFile>nANT_AssemblyLabeler.build</buildFile>
                   <buildTimeoutSeconds>6000</buildTimeoutSeconds>
          </nant>

            
            <devenv>
                  <solutionfile>MyProduct.sln</solutionfile>
                  <configuration>Release</configuration>
                  <executable>C:\Program Files (x86)\Microsoft Visual Studio 10.0\Common7\IDE\devenv.com</executable>
                  <version>VS2010</version>
            </devenv>

            
            <devenv>
                <solutionfile>MyProduct.sln</solutionfile>
                <configuration>Release</configuration>
                <executable>C:\Program Files (x86)\Microsoft Visual Studio 10.0\Common7\IDE\devenv.com</executable>
                <version>VS2010</version>
         </devenv>

         
         <exec>
                  <executable>C:\Windows\System32\cmd.exe</executable>
                  <baseDirectory>C:\Dev\MyProduct\ReleaseManagement\Debug\1.x\Products</baseDirectory>
                 <buildArgs>/c CopyFilesForSupporting.bat</buildArgs>
         </exec>

          
         <exec>
               <executable>C:\Windows\System32\cmd.exe</executable>
               <baseDirectory>C:\Dev\MyProduct\ReleaseManagement\Releases\1.x\Products</baseDirectory>
               <buildArgs>/c CopyAssembliesToReleaseFolder.bat</buildArgs>
         </exec>

         
         <exec>
               <executable>C:\Program Files (x86)\Inno Setup 5\ISCC.exe</executable>
              <baseDirectory>C:\Dev\MyProduct\ReleaseManagement\Releases\1.x\Products</baseDirectory>
              <buildArgs>"/SMYSignCode=C:\Program Files (x86)\Microsoft SDKs\Windows\v7.0A\Bin\SignTool.exe sign /a /f C:\Dev\MyProduct\ReleaseManagement\My_Code_Signing.pfx /v /t http://timestamp.verisign.com/scripts/timstamp.dll /p mypassword $f" MyInnoSetup.iss</buildArgs>
           </exec>

    </tasks>

      <publishers>

          
          <buildpublisher>
                <sourceDir>C:\MyProduct\ReleaseManagement\Debug\1.x\Products\Output</sourceDir>
                <publishDir>\\MyFileServer\1.x\Debug</publishDir>
                <useLabelSubDirectory>true</useLabelSubDirectory>
                <alwaysPublish>false</alwaysPublish>
          </buildpublisher>

          
          <buildpublisher>
                <sourceDir>C:\MyProduct\ReleaseManagement\Releases\1.x\Products\Output</sourceDir>
                <publishDir>\\MyFileServer\1.x\Release</publishDir>
                <useLabelSubDirectory>true</useLabelSubDirectory>
                <alwaysPublish>false</alwaysPublish>
           </buildpublisher>

           <xmllogger logDir="C:\Dev\MyProduct\ReleaseManagement\1.x\BuildLogs" />

      </publishers>

    </project>
　

</cruisecontrol>

Reading COM TypeLib GUID at Runtime

2011-11-04T10:16:00.000-07:00

/*
* Every need to read a type library dll to update other reference files. This is often a requirment when working with languages
* that don't control the GUID or you just want to create a generic process. The example below can easily be turned into a console app
* that takes the DLL and Replacement file as command arguments. I'm using Visual Studio 2010.
*/

using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;
using System.Text.RegularExpressions;
using TYPELIBATTR = System.Runtime.InteropServices.ComTypes.TYPELIBATTR;

namespace GUIDReader
{

class Program

{

[DllImport("oleaut32.dll", CharSet = CharSet.Unicode, PreserveSig = false)] static extern UCOMITypeLib LoadTypeLib(string szFile);

const string mydll = @"D:\mydll.dll";
const string globalasa = @"D:\global.asa";

static void Main(string[] args)
{

    TYPELIBATTR tlibvalue = new TYPELIBATTR();
    IntPtr ptr;
    LoadTypeLib(imASPdll).GetLibAttr(out ptr);
    tlibvalue = (TYPELIBATTR)Marshal.PtrToStructure(ptr, typeof(TYPELIBATTR));
    var guidstr = tlibvalue.guid.ToString();
    Console.WriteLine(guidstr);

    StreamReader reader = new StreamReader(globalasa);
    string contents = reader.ReadToEnd();
    reader.Close();

    var newvalue = "NAME=\"NAMEOFLIB\" UUID=\"{" + guidstr.ToUpper() + "\"}";
    var newText = Regex.Replace(contents, @"NAME=""NAMEOFLIB"" UUID=""{.*}""", newvalue);

    StreamWriter writer = new StreamWriter(globalasa);
    writer.Write(newText);
                writer.Close();

}

}

}

HTML Style Sections to Inline Style Tags

2008-05-09T07:12:00.000-07:00

//***** LargeEdit Scripting *****
//
// Title:  HTML Style Sections to Inline Style Tags
// Author:  Tim O'Brien
// Date:  May 8, 2008
// Purpose: Performs a JScript regular expression search on the HTML document for
//          style attributes and replace class references.
//
//*******************************

function Run() {
  var scanlines= "(.+)\\s\\x7b([^\\x7d]*)";
  var scansection="<(style).+>(.+)[^\/\1]*"

    LargeEdit.ResultLog(' HTML Style Section to Inline Style Tags (Start)');

  var styletext;
    var inpStr = LargeEdit.CurrentFile.Text;
  var newtext = inpStr;

    //Update InpStr to only Style Sections
    var arrStyle;
    var oReSections;
    var oRe;
    oReSections = new RegExp(scansection, "gmi");

    var retStr;
    var objtag;
  var objval;
  var classidx;
  var classname;
    while ((arrStyle = oReSections.exec(inpStr)) != null) {
     styletext = arrStyle[0];

    // Get the Matches collection
     oRe = new RegExp(scanlines, "gmi");

     retStr = '';
     objtag = '';
   objval = '';
   classidx = -1;
   classname = '';
     while ((arr = oRe.exec(styletext)) != null) {
    //retStr = arr[1] + ' = ' + arr[2];
    objtag = arr[1];
    objval = arr[2];
       //LargeEdit.ResultLog(retStr);
       
       //Do Replacement of class ids
       classidx = objtag.indexOf(".");
       if (classidx >= 0) {
        classname = objtag.substr(classidx+1);
          LargeEdit.ResultLog('Class Tab Found: ' + classname);
        newtext = newtext.rep('class="' + classname + '"', 'style="' + objval.trim() + '"');
       }

     }
  } //Each Style Section

  LargeEdit.New();
  LargeEdit.CurrentFile.Text = newtext;
    LargeEdit.ResultLog(' HTML Style Section to Inline Style Tags (Finish)');
}

function StringReplace(findstr, replacestr) {
  var regEx, strout;
  regEx = new RegExp(findstr, "igm");
  strout = this.replace(regEx, replacestr);
  return strout;
}
String.prototype.rep = StringReplace;

function StringTrim()
{
 return this.replace(/(^\s*)|(\s*$)/g, "");
}
String.prototype.trim = StringTrim;

//copyright 2008 All rights reserved

Delimited Line Template

2008-05-09T06:50:00.000-07:00

//***** LargeEdit Scripting *****
//
// Title:  Delimited Line Template
// Author:  Tim O'Brien
// Date:    May 8, 2008
// Purpose: Merges lines of text containing delimited fields into a template.
//          Usage: Create a SQL insert statement and merge the fields of text
//                 into the correct position for inserting into a database.
//          Example: Insert into table(id, name, date) values ( {F1}, '{F3}', {F2} );
//                   the field tags {F1} will be replaces with the actual values.
//*******************************
var Delimiter = 9;
 //Tab = 9,  Comma = 44, Pipe = 124.
 //(see LargeEdit | Tools | ASCII Codes | View ASCII Code, for a complete list.

function Run() {
 LargeEdit.ResultLog(' Start - Delimited Line Template ');

 var tmp = '';
 tmp = LargeEdit.InputBox('Delimited Line Template','Enter Template:','')
 if (tmp.length <= 0) {
    LargeEdit.ResultLog(' Canceled - Enter a Template ');
  return;
 }

 //Variables used
  var line, newline;
  var flds, fld, fldidx, lineidx, findtxt;
  var newtext = "";
  
  //Loop for Lines of text
 for (var i = 0; i < LargeEdit.CurrentFile.LineCount; i++) {
   line = LargeEdit.CurrentFile.Line( i );
   lineidx = i + 1;
   flds = line.split(String.fromCharCode(Delimiter)); //Tab default

  //Replace Line Index Reference
  newline = tmp; //Reset newline to base template
  findtxt = '{LIDX}';
  newline = newline.rep(findtxt, lineidx);

  //Loop for Fields to do replacements
  for (var j = 0; j < flds.length; j++) {
   fld = flds[j] //Get Field Value
   fldidx = j + 1;
   fld = fld.rep("'", "''");
   findtxt = '{F' + fldidx.toString() + '}';
   newline = newline.rep(findtxt, fld);
  }
  newtext += newline + "\r\n";
  
 }

 LargeEdit.New();
 LargeEdit.CurrentFile.Text = newtext;

  LargeEdit.ResultLog(' Finish - Delimited Line Template ');

}

function StringReplace(findstr, replacestr) {
  var regEx, strout;
  regEx = new RegExp(findstr, "igm");
  strout = this.replace(regEx, replacestr);
  return strout;
}
String.prototype.rep = StringReplace;

//copyright 2008 All rights reserved

Reverse Delimited Field Order

2008-05-08T06:36:00.001-07:00


//***** LargeEdit Scripting *****
//
// Title:  Reverse Fields
// Author:  Tim O'Brien
// Date:   May 7, 2008
// Purpose: Reverse the order of Delimited Fields in a document.
//     Useful for switch lists like City, State, Country so Sorting
//            by Country, State, City can be preformed.
//
//        From:  Field1, Field2, Field3
//        To:    Field3, Field2, Field1
//
//Note: Delimiter var below is defaulted to "Tab" char 9.
//*******************************
var Delimiter = 9;
 //Tab = 9,  Comma = 44, Pipe = 124.
        //(see LargeEdit | Tools | ASCII Codes | View ASCII Code, for a complete list.

function Run() {

  LargeEdit.ResultLog(' Reverse Fields ');

  var line;
  var flds;
  var newtext = "";
 for (var i = 0; i < LargeEdit.CurrentFile.LineCount; i++) {
   line = LargeEdit.CurrentFile.Line( i );
   flds = line.split(String.fromCharCode(Delimiter)); //Tab
  for (var j = flds.length-1; j >= 0; j--) {
   newtext += flds[j]
   if (j > 0) newtext += String.fromCharCode(Delimiter);
  }
  newtext += "\r\n";
 }

 LargeEdit.New();
 LargeEdit.CurrentFile.Text = newtext;

}

//copyright 2008 All rights reserved

Productivity

2008-04-25T09:05:00.000-07:00

Productivity is number one priority! In IT the productivity of the business depends on the personal productivity of the Staff.

As programmers we optimize repetitious code by creating libraries, classes, functions, etc... and call that code over and over. However, most of us don't optimize the tasks and activities we do repeatably.

There are many activities that fall into this category;

Scanning log files for specific data.
Editing blocks of text to be compatible with other programs or languages.
Analysing code or data for specific patterns
...

I'm sure we can all remember a time typing on the keyboard and feeling like a monkey. This is not only a waste of time, it's boring!!! Wouldn't it be more rewarding using your brain to find a way to automate that task and complete the work in a fraction of the time?If you have Microsoft Windows, you already have a powerful automation tool installed;

WScript.exe - This program is installed in your Window\System folder and all *.js and *.vbs files a associated with it. JScript and VBScript are powerful scripting technologies that can be uses to improve your efficiency.

Visit: http://www.microsoft.com/technet/scriptcenter/default.mspx

There is a scripting guide, tools and lots of examples. Check out the script repository, I'm sure you will find something there that would be useful.

Not everything requires scripting, sometimes a quick, simple keyboard macro will do the trick and stop you from feeling like a monkey. Many programs support keyboard marcos, I suggest LargeEdit ;)

Introduction

2008-04-25T07:12:00.000-07:00

My name is Tim O'Brien, I've been a professional IT consultant since 1999. However, my experience with computers started in 1987 and I've been programming since 1993.

I develop and publish applications such as LargeEdit under the company name "Netlegger Systems Inc" (NSi). I've been working contracts since incorporating NSi in 2005, so I haven't had much time to work on the NSi website.

I recently created the brand "Kyght", which I use to release mobile applications and games under.

The primary purpose of this blog is for myself. I spend lots of time doing research to resolve specific programming issues, then years later I'm faced with similar issues.

The postings will vary from analysis to code and will include many different languages and technologies.