The test was to check whether some tracking tag javascript was being inserted into the page correctly or not.

I assumed that I could get the page source as it was being delivered to the browser by calling selenium.GetHtmlSource(); and then check that for the javascript string I was expected.

Unfortunately, GetHtmlSource is just a proxy for the browsers DOM.InnerHTML method; and that returns the Html after it has been preprocessed by the browser.

Turns out that preprocessing does a couple of funky things, including

• Changing line-endings (Firefox)
• Changing capitalization (IE6)
• Seemingly random removal / insertion of ” & ‘  (IE6)

So, when I was expecting a string like this:

1
2
3
4
5
6

<script language="javascript" type="text/javascript">
<!--
   var amPid = '206'';
   var amPPid = '4803';
   if (document.location.protocol=='https:')
...[snip]...

IE6 was presenting me with:

1
2
3
4
5
6

<SCRIPT language=javascript type=text/javascript>
<!--
   var amPid = '206'';
   var amPPid = '4803';
   if (document.location.protocol=='https:')
...[snip]...

A possible solution is to ignore case, whitespace and quotes when doing the comparison, with a helper method like this:

1
2
3
4
5
6
7
8
9
10
11
12
13

/// <summary>
        /// Use this to compare strings to those returned from selenium.GetHtmlSource for an Internet Explore instance
        /// (IE6 seems to change case and inclusion of quotes, especially for Javascript.?)
        /// </summary>
        /// <param name="expected"></param>
        /// <param name="actual"></param>
        private static void AssertStringContainsIgnoreCaseWhiteSpaceAndQuotes(string expected, string actual)
        {
            string expectedClean = Regex.Replace(expected, @"\s", "").ToLower().Replace("\"","").Replace("'","");
            string actualClean = Regex.Replace(actual, @"\s", "").ToLower().Replace("\"", "").Replace("'", "");
            StringAssert.Contains(expectedClean,actualClean,
                                  string.Format("Expected string \n\n{0} \n\nis not contained within \n\n{1}", expected, actual));
        }

It was the line endings that really floored me; because they were automatically normalized/corrected by my test runner when displaying the error. Aaargh!

I don’t want to manage another set of usernames & passwds, so I’d like apache2 to authenticate off the standard users on my system, via PAM.

To get this to work, you need to install and configure mod_auth_pam and mod_auth_shadow

aptitude install libapache2-mod-auth-pam libapache2-mod-auth-shadow

Ensure the www-data user is part of the shadow group, so apache2 can read the passwords

usermod -G shadow www-data

And set up the relevent virtual host:

                AuthPAM_Enabled On
                AuthShadow on
                AuthPAM_FallThrough Off
                AuthBasicAuthoritative Off
                AuthType Basic
                AuthName "Restricted to group: sysadmins"
                AuthUserFile /dev/null
                Require group sysadmins

Restart apache, and you’re done!

https://help.ubuntu.com/8.04/serverguide/C/certificates-and-security.html

Create a new virtual host (you can only have one SSL virtual host / IP)

sudo cp /etc/apache2/sites-available/default /etc/apache2/sites-available/ssl

Edit ssl sothat it looks like this:
NameVirtualHost *:443

ServerName webangle-www1.everyangle.co.uk
ServerAdmin webmaster@localhost

SSLCertificateFile /etc/ssl/certs/server.crt
SSLCertificateKeyFile /etc/ssl/private/server.key


Finally, if you want to force redirect of all traffic to a certain folder via SSL (e.g, /phpmyadmin), add the following to /etc/apache2/sites-available/default

#Redirect traffic to /phpmyadmin through https
        RewriteEngine   on
        RewriteCond     %{SERVER_PORT} ^80$
        RewriteRule     ^/phpmyadmin(.*)$ https://%{SERVER_NAME}/phpmyadmin$1 [L,R]

Enable it:

sudo a2ensite ssl
sudo /etc/init.d/apache2 reload

A convienient way to access that remote storage is to configure rsync over sshfs:

sudo aptitude install sshfs
mkdir /mnt/sshfs
mkdir /mnt/sshfs/rsync.net
sshfs **username**@ch-s011.rsync.net: /mnt/rsync.net

First, unmount

fusermount -u /mnt/rsync.net

Then, install autofs, and edit the config file

sudo aptitude install autofs
sudo vi /etc/auto.master

Add the following line 

/mnt/sshfs /etc/auto.sshfs --timeout=30,--ghost

Then,  

sudo vi /etc/auto.sshfs

Add

rsync.net -fstype=fuse,rw,nodev,nonempty,noatime,allow_other,max_read=65536 :sshfs\#**username**@ch-s011.rsync.net\:

And finally restart autofs 

sudo /etc/init.d/autofs restart

Now, when you cd /mnt/sshfs/rsync.net, after a short delay you will automatically be connected to the remote filesystem over SSH.  After 30 seconds of inactivity, the connection will be closed.

For backups, I want to do weekly backups and daily incrementals of the data files, and sync these off to an external backup location.

Broadly, there are two parts to the backup, creating the backed up files, and then copying them offsite.

Creating the backups

I’m using backup-manager 0.7.6-debian1, which handles backing up sets of files and MySQL databases to tar.gz files.

sudo aptitude install backup-manager
sudo /usr/sbin/backup-manager --version

The comments in the config file make editing it quite straight forward.

sudo vi /etc/backup-manager.conf

One minor points:

• Separate multiple backup methods with a space, eg:

  export BM_ARCHIVE_METHOD="tarball-incremental mysql"

To test:

sudo /usr/sbin/backup-manager --verbose

The output folder you specified (/var/archives) should now contain some .tar.gz versions of your data. Hurrah!

Getting the files offsite

Originally I intended to use Amazon’s S3 as a backup store, following Michael Zehrer’s instructions on how to rsync with S3. However, I couldn’t get this to work reliably; so I opted instead for rsync.net which offers standard scp, ftp, WebDav and sshfs access to their geographic backup locations.

Backup-manager can rsync over ssh, which is a quick and efficient way to sync changes over to the remote host..

The first step is get your rsync.net account setup; and set up your ssh so you can access without typing in a password

Then, set the BM_UPLOAD_METHOD to rsync, and configure both the scp and the rsync settings in /etc/backup-manager.conf (pay attention not to prefix remote folders with / ).

Test with:

sudo /usr/sbin/backup-manager --verbose

Once its all working, set up a cron job to call backup-manager daily.

crontab -e

I run backup-manager once per day in the wee hours, and log output to /root/crontab/daily_backup-manager.logs

  0 3   *   *   *    /usr/sbin/backup-manager -v > /root/cronlogs/daily_backup-manager.log

Viola!

The evolution of the giraffe.

In case you were wondering.  More sidesplitting stuff at http://www.flickr.com/groups/farside/pool/

• 40% of all software errors are caused by pressure on developers to complete quicker (Glass 2004)
• Under extreme schedule pressure, code defects increase by 400% (Jones 2004)
• Projects which aim to have the lowest number of defects also have the shortest schedules (Jones 2000)

This makes sense is you consider that good engineering practises are the first to leave the building under pressure to finish, and most teams will revert to quick & dirty hacks to get things implemented, without complete testing etc.

My personal opinion is that the only way to shorted development cycles is to reduce the feature set. Its pleasing for me to see that the research seems to back this up.

When deciding which features will be dropped; I think its worth revisiting the business requirements that are driving a particular set of features. In many cases a simpler “design” could suffice; for example a fancy calendar widget could be replaced with a simple textbox; a little used settings screen could be retired in favour of manually changing config files; or overly complex but little used workflows could be put on the back burner.

I maintain that a lot of “features” can be dropped, without actually impairing the business functionality of the system.

Just remember, what every you do DON’T consider dropping testing or QA in an effort to meet your deadline; unless you want to guarantee that you will continue to miss all future deadlines until the project gets cancelled!

Rusty Zarse blogged about a useful ViewData helper class, which allows you to set properties by passing values to the partial through the ViewData.

I’ve extended this slightly to enable the following syntax:

1
2
3
4
5
6
7
8

<% Html.RenderPartial("YUIDataTable",
               ViewDataDictionaryBuilder.Create(
                      new
                      {
                        DataTableId = "QuoteDataTable",
                        ConfigNamespace = "QuoteDataTableConfig",
                        HideFilter = true
                       }));%>

Which sets properties on a ViewUserControl like this:

1
2
3
4
5
6
7
8
9
10
11

     public partial class YUIDataTable : ViewUserControl
    {
        public string ConfigNamespace { get; set; }
        public string DataTableId { get; set; }
        public bool HideFilter { get; set; }
 
        protected void Page_Load(object sender, EventArgs e)
        {
            ViewDataDictionaryBuilder.SetPropertiesToViewDataValues(this);
        }
    }

Here is the full helper code.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

using System;
 
namespace MvcHelpers
{
    ///
    /// With thanks to http://www.vitaminzproductions.com/technology-blog/index.php/2008/11/12/setting-properties-using-aspnet-mvc/
    ///
    public static class ViewDataDictionaryBuilder
    {
        public static System.Web.Mvc.ViewDataDictionary Create(object data, ModelType model) where ModelType : class
        {
            return (System.Web.Mvc.ViewDataDictionary)CreateInternal(new System.Web.Mvc.ViewDataDictionary(model), data);
        }
 
        public static System.Web.Mvc.ViewDataDictionary Create(object data, object model)
        {
            return CreateInternal(new System.Web.Mvc.ViewDataDictionary(model), data);
        }
 
        public static System.Web.Mvc.ViewDataDictionary Create(object data)
        {
            return CreateInternal(new System.Web.Mvc.ViewDataDictionary(), data);
        }
 
        private static System.Web.Mvc.ViewDataDictionary CreateInternal(System.Web.Mvc.ViewDataDictionary dictionary, object data)
        {
            AddPropertiesToViewData(dictionary, data);
            return dictionary;
        }
 
        private static void AddPropertiesToViewData(System.Web.Mvc.ViewDataDictionary dictionary, object data)
        {
            if (data == null) return;
 
            System.Reflection.PropertyInfo[] properties = data.GetType().GetProperties();
 
            foreach (var property in properties)
            {
                dictionary.Add(property.Name, property.GetValue(data, null));
            }
        }
 
        public static void SetPropertiesToViewDataValues(System.Web.Mvc.ViewUserControl viewUserControl)
        {
            foreach (var property in viewUserControl.GetType().GetProperties())
            {
                if (viewUserControl.ViewData[property.Name] != null)
                    property.SetValue(viewUserControl, Convert.ChangeType(viewUserControl.ViewData[property.Name], property.PropertyType), null);
            }
        }
 
    }
}

Hope that’s useful to you!

The aim of the project is to centralise all those little test helper methods you end up creating into a useful assembly you can use to jumpstart your next project.  Things like:

• Comparers
  • Generic object comparers
  • DataSet comparers
• Test Data generators
  • Builder pattern
• Automocking containers

For example, I’ve just added an “AssertValues” functor; which helps you check whether the values of who object instances are the same. 

One area I keep using asserts like this is in integration tests; where I want to check that the objects I’m persisting to the database via my ORM actually end up in the database in a non-mangled form.  In this case, I new up entityA, persist it, reload it into entityB and then need to check that all the values in entityB are the same as those in entityA.

A standard Assert.AreEqual will fail, because entityA and entityB are different instances.  But, my helper method AssertValues.AreEqual will pass, because it checks the (serialized) string values of entityA and entityB.

Here is another, simpler example to illustrate the concept.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

    [TestFixture]
    public class StandardObjectsTests
    {
        public class StringContainer
        {
            public string String1 { get; set; }
            public string String2 { get; set; }
        }
 
        [Test]
        public void ObjectsWithSameValue_ShouldBeEqual()
        {
            var stringContainer1 = new StringContainer {String1 = "Test String1", String2 = "Test String 2"};
            var stringContainer2 = new StringContainer {String1 = "Test String1", String2 = "Test String 2"};
 
            Assert.AreNotEqual(stringContainer1, stringContainer2);
 
            AssertValues.AreEqual(stringContainer1, stringContainer2);
        }
   }

I’m sure you have a bunch of similar helper methods lying about your projects.

How about contributing them to the TestHelper project?

I really enjoyed Mike Hadlow’s talk on IOC injection; with specific reference to his opensource eCommerce application, SutekiShop.  Clearly an expert on the subject on ASP.NET MVC, Onion architecture, Repositories & Services,  and binding it all together with IOC; he is also a gifted presenter.   If you’re looking for a reference implementation of an ASP.NET MVC application (or indeed just a loosely coupled, TDD driven web application); I’d strongly advise you to check out Mike’s SVN repo.

Toby Henderson gave an interesting demo of how you can run .NET apps under Linux (Ubuntu) using Mono.  Worth bearing in mind when considering your hosting & deployment options.

Sebastien Lambla gave a highly entertaining (if opinionated) presentation of a series of WFP tips and tricks.  My favourite tip (which isn’t really WPF related)

Tired of always checking if your event delegates are null before calling them?  Just declare them with a standard empty delegate.  Then they are never null!

1

  event MyEvent = delegate {};

Recommended book: WPF Unleashed by Adam Nathan

As always it was a great event - remember, if you want to be at DDD8 (2009); sign up early!

See www.developerday.co.uk for slides & videos from all sessions