Blog

Whilst carrying out some tests on my blog after upgrading my blog platform from BlogEngine 1.6 to 2.0, I noticed an error occurring when a comment was in process of getting submitted. The error that cropped up was: “There was an error in callback”. The strange thing about this error was that it only occurred for Internet Explorer users. In other mainstream browsers (I do not class Internet Explorer as a mainstream), my comments were successfully submitted.

As it turns out, the issue was due to one of my BlogEngine extensions. When I disabled the “Recaptcha” extension all worked as expected. I still need to investigate what is causing the “Recaptcha” extension to error since this is one of the most important extensions to void all spammers out there.

So if you are also experiencing a similar error, disabling the Recaptcha extension may help.

I will post further updates if I manage to get the commenting system working using Recaptcha. But since BlogEngine is now supporting the Disqus commenting platform, I may decide to make the move.

I needed to create a web part that simply output all information from an Announcements list. Sounds simple doesn’t it? But I came across a compiler error when using the “Created By” and “Modified By” columns. The error was telling me that the columns I am trying to read in my code do not exist.

After some Googling, I found out that SharePoint has something called “Internal Field Names”, whereby the display name of a column isn’t the actual column name. The full list of internal field names can be found at Michael Yeager's MSDN Blog.

If ever in doubt, you can find out the internal field names within SharePoint, by carrying out the following:

1. Go to your list and select “List Settings” from the ribbon.

2. Navigate to the where the list of columns are displayed and right click on the column in question. Select “Properties”.

3. When the link property window opens, scroll to the end of the “Address: (URL)” section. You will find a query string parameter called “Field”. This query string parameter contains the real field name.

Confusing isn’t it? Well I guess it wouldn’t be SharePoint if the most simplest task failed to confuse a developer.

One of the best development features in making customisations within a SharePoint environment is the SPMetal tool. The SPMetal tool generates entity classes in order to use LINQ syntax to retrieve items from lists. The SPMetal.exe resides in the 14/bin folder. I will show you a quick demonstration on how I have generated a DataContext class using the SPMetal executable.

You will need to create two files:

1. XML file – containing the list or library name we are interested in generating our class from. In the code snippet (below), the list I am interested in is called Quicklinks.

<?xml version="1.0" encoding="utf-8"?>
<Web xmlns="http://schemas.microsoft.com/SharePoint/2009/spmetal">
  <List Name="Quicklinks" Member="Quicklinks" />
  <ExcludeOtherLists />
</Web>

Save the XML file into the 14/bin director and name it “SPMetalConfig.xml”.

2. Batch File – will contain the SharePoint command prompt needed to fire the SPMetal executable. Creating a batch file is not required. You can run the command as normal through Windows Command Line. I just find it a lot easier to manage and make changes using a batch file.

SET SPMETALDIR="C:\Program Files\Common Files\Microsoft Shared\Web Server Extensions\14\BIN\" 

cd %SPMETALDIR% 

spmetal.exe /web:http://myintranet /code:QuicklinksOMEnitites.cs /parameters:SPMetalConfig.xml

Once this has run, a class called “QuicklinksOMEnitites.cs” will be generated in the 14/bin folder.

One thing I did find when using SPMetal for retrieving list data to be output to a custom webpart was that it is only able to retrieve data from a site collection the list resides. But this is probably to be expected. So if the webpart was used in a separate site collection, no data will be shown. In order to get around this, SPMetal will not suffice. You will need to go through the code route and do the following:

SPSecurity.RunWithElevatedPrivileges(delegate()
{
    using (SPWeb intranetSite = new SPSite("http://myintranet/sites/depts/").OpenWeb())
    {
        SPList quickLinks = intranetSite.Lists.TryGetList("Quicklinks");
        
        if (quickLinks != null)
        {
            //Do something with list data.
        }                           
    }
}); 

From one of the projects I have been working on, I came across a snippet of code that used the XmlDocument.Load method. What alarmed me about this piece of code was the fact that there was no error handling. If for some reason the XML file could not be found or a node was missing, the whole page would have crashed. Not good.

I must admit, I am not exactly the best person to speak about implementing wide-scale error handling in every facet of code. But I do ensure the core foundations of an application or website do incorporate sufficient error handling.

So back to the matter in hand. This is the original code using the XmlDocument.Load functionality:

XmlDocument doc = new XmlDocument();

doc.Load(Server.MapPath("/xml/storeGB.xml"));

XmlNode countryNode = doc.SelectSingleNode("//countries");
foreach (XmlNode node in countryNode.ChildNodes)
{
    //Do something with the elements
    Response.Write(node.Name + node.InnerText);
}

I changed the code to the following:

XmlDocument doc = new XmlDocument();

//Check if language XML file exists
if (File.Exists(Server.MapPath("/xml/storeGB.xml")))
{
    try
    {
        doc.Load(Server.MapPath("/xml/storeGB.xml"));

        XmlNode countryNode = doc.SelectSingleNode("//countries");

        if (countryNode != null)
        {
            foreach (XmlNode node in countryNode.ChildNodes)
            {
                //Do something with the elements
                Response.Write(node.Name + node.InnerText);
            }
        }
        else
        {
            //Output error message if there is no node
        }
    }
    catch (XmlException ex)
    {
        Debug.WriteLine(String.Format("XmlException for countries: {0}", ex.Message));
    }
}

I am sure you will agree that this is the better approach to using XmlDocument.Load.

Whenever I work with strings whilst programming in .NET, somewhere along the lines I always find myself using the awesome “string.format”. I think all of you will admit its the most useful and easiest way to build up your strings.

The only downside to using the “string.format” method is that it lures you into a false sense of security and you find yourself lost without it when it comes to working with strings in other languages. This happened to me when I had to build up a long string in JavaScript. It was long and tedious…or maybe I am just lazy.

Luckily, there have been a few developers who extended the string object in JavaScript to include “string.format”. Amazing! Its goes along the lines of adding this to your own JavaScript code:

String.format = String.prototype.format = function() {
    var i=0;
    var string = (typeof(this) == “function” && !(i++)) ? arguments[0] : this;

    for (; i < arguments.length; i++)
        string = string.replace(/\{\d+?\}/, arguments[i]);

    return string;
}

Here are some other useful links I have found on how to implement “string.format” into your JavaScript code:

• http://www.geekdaily.net/2007/06/21/cs-stringformat-for-javascript/
• http://frogsbrain.blogspot.com/2007/04/javascript-stringformat-method.html

The “Remove Format” button () within FCKEditor, only removes valid inline elements such as: strong, span, strike, font, em, etc.

If you want to be able to make the Remove Formatting function more flexible so that it removes block elements, you can do so by modifying the “fckconfig.js” file found within the FCKeditor folder.

Search for the “FCKConfig.RemoveFormatTags” line, which will look something like this:

FCKConfig.RemoveFormatTags = 'b,big,code,del,dfn,em,font,i,ins,kbd,q,samp,small,span,strike,strong,sub,sup,tt,u,var';

All you need to do now is add any additional elements you wish to remove from your content. In my case, I wanted the Remove Formatting button to remove all header tags. So I carried out the following:

FCKConfig.RemoveFormatTags = 'b,big,code,del,dfn,em,font,i,ins,kbd,q,samp,small,span,strike,strong,sub,sup,tt,u,var,h1,h2,h3';

I have been developing custom web parts and SharePoint customisations for a couple years now. During the early stages of SharePoint development I found a great deal of confusion when trying to retrieve information from different areas of an intranet through using the SPSite and SPWeb methods.

I think any other developer who is starting out SharePoint development may encounter the same issue. What I found useful was the “Site Architecture and Object Model Overview” diagram from the MSDN site.

It nicely breaks down the architecture of a SharePoint site. I highly recommend that you take look at the following links containing more diagrams:

• http://msdn.microsoft.com/en-us/library/ms473633.aspx
• http://technet.microsoft.com/en-us/library/cc263199.aspx

Whilst I am talking about SPSite’s and SPWebs’s in this post, I’ll give you quick overview on how they work using the (above) diagram. Basically, the top level site collection (SPSite) contains your root web (SPWeb) and subsites (SPWebs under SPWebs). So, a subsite will be any site (SPWeb) under the top level web site in a site collection (SPSite).

In light of what has happened recently with some 150,000 Google Account holders loosing their information due to a mishap at Google HQ over the weekend really reinforces the fact that our data is not safe…even in the “cloud”.

At the end of the day our information is stored on hardware that can fail. I think that this whole “cloud computing” malarkey has got all lured into a false sense of security where we think we don’t need to take measures to ensure our data backed up on a regular basis. I have to admit, I too have become a bit tardy when it comes to backing up my online data. If a large company like Google can get it wrong, what hope is there for other companies offering the same thing?

I practically live on the “cloud” in terms of what Google has to offer. I use their email, calendar, document and notebook applications. Even their mobile phone OS: Android! Luckily, there are steps we can take to ensure our data is backed up on your own terms:

	Google Calendar Google Calendar is the one application I use the most. If I lost all my data, I would quite annoyed to say the least (and be very disorganised). You can backup all your calendar entries by opening your calendar settings, click on Calendars and select “Export Calendars”. A zip file will be created containing your calendars in a .ical format.
	Gmail This a simple one. Use an desktop email client such as Thunderbird (or any other client you prefer) to download all your emails directly to you computer through POP access.
	Docs If you only store a handful of documents in your Google Account, you could just download them one-by-one. Understandably, if you have a long list of documents a more automated approach is required. Lifehacker.com shows a really great script you can use to that allows you to download documents in whatever format you require. Take a look here.

Hooray! Our data is saved!

In ASP.NET you would think when you use the “.Count” method that it would be able to simply return the total number of elements within a collection. In majority of cases this is right. Well, apart from when you use the “.Count” method against a collection of profiles within SharePoint. For example:

UserProfileManager profileManager = new UserProfileManager(myContext);

//Get total number of profiles
int numberOfProfiles = profileManager.Count;

I found that I came across two issues when using the code above:

1. The incorrect number of profiles was returned.
2. For some reason, when I deployed the code to a live server environment I kept on getting errors from the line where the count was being returned.

From researching this issue on various blog posts and forums, it seems that UserProfileManager.Count does indeed have issues in returning the count correctly. The only way to get around this is to enumerate through the UserProfileManager:

UserProfileManager profileManager = new UserProfileManager(myContext);

int counter = 0;

IEnumerator profileEnumerator = profileManager.GetEnumerator();
while (profileEnumerator.MoveNext())
{
    counter++;
}

//Number of profiles
int totalNumberOfProfiles = counter;

This will give us an accurate number of profiles that are stored within SharePoint and without any silly errors.

I have been building a custom .NET web part page to use in my SharePoint intranet. The .NET page has quite a lot of custom HTML and jQuery design elements, so using CSS and JavaScript files were essential.

As you know, when we want to use elements from our CSS and JavaScript files we normally add the following lines of HTML at the top of our page:

<!-- CSS -->
<link type="text/css" rel="stylesheet" href="site.css" />

<!-- JavaScript -->
<script src="jQuery.js" type="text/javascript" />

If you added those lines of code in a custom SharePoint page, you’ll find that the page will ignore them. Thankfully, SharePoint has given us some controls to add these references.

At this point its worth stating that I stored all my required JavaScript and CSS files within the “Style Library” directory situated in the root of any SharePoint 2010 intranet. In order to get these files I used controls called”CssRegistration” and “ScriptLink”:

<!-- CSS -->
<SharePoint:CssRegistration ID="CssRegistration1" Name="/Style Library/Home/CSS/jcarousel.css" runat="server" After="corev4.css" />

<!-- JavaScript -->
<SharePoint:ScriptLink ID="ScriptLink1" Name="~sitecollection/Style Library/Home/JS/jquery-1.4.4.min.js" runat="server" />

If you have stored your CSS and JavaScript within the physical file directory situated in the 14 hive folder, you will need to modify the above example to the following:

<!-- CSS -->
<SharePoint:CSSRegistration Name="<% $SPUrl:~SiteCollection/Style Library/Core Styles/jcarousel.css%>" runat="server"/>

<!-- JavaScript -->
<SharePoint:ScriptLink ID="ScriptLink1" Name="<% $SPUrl:~SiteCollection/Style Library/Core Styles/jquery.js%>" runat="server" />

The only difference between this example and our earlier example is that when we have just added “SPUrl” to get files relative to the current site collection.