While working on RegexConverter I've started learning things about XML. I disliked it before, but now... I loathe it completely. It feels like the XML people are from another planet and I have to twist my brain at straight angles just to understand what they meant.

One of the problems I met was "how to declare in an XML schema a unique constraint/index on the entire document when there are nested elements of the same type". In other words, something like this:

<regex>
<group index="1">
    <group index="2">
    </group>
    <group index="3">
    </group>
</group>

Bottom line:

1. Use the declared namespace
2. Declare the index in the root element
3. Use the name of the nested element (in my case group) in the selector
4. Use the one or more levels deep selector (.//)
5. Use the declared namespace AGAIN.

.

Concrete example:

<xs:element name="regex" type="regexType">
    <xs:unique name="uniqueIndex">
      <xs:selector xpath="mstns:group|.//mstns:group"></xs:selector>
      <xs:field xpath="@index"></xs:field>
    </xs:unique>
  </xs:element>

I tried to solve this for two hours only to find the solution was this idiotic. Neither one of "group","mstns:group","mstns:group|.//group" work. The only one is "mstns:group|.//mstns:group".

Grrrr...

Update: two more days of work and a huge text file of regular expressions and I know a lot more about the syntax of .Net Regex than even MSDN :)
The third (Apr11) release of the library has a huge load of bug fixes and new features and it is now... [Tadaaaaa]... a stable version. At least I think of it that way :) Go download it.!

I've been working for two days on an idea. What can I do to make those long regular expressions that I always leave in my code more readable and easy to understand without having to compile automatons in your head?

I have first researched on Google regular expression converters. Nothing was even close to what I wanted. Also, on the more scientific side, people were talking a lot about BNFs, as a superset of regular expression syntax. I looked at BNF. Yes, it describes anything in a human readable form, it is used in RFCs but I hate it even more than XML! So XML it is. Most of the inspiration for the code came from this link: Regular expressions and regular grammars

I give you, RegexConverter. It is a library+demo that transforms a Regex into an XML and then back again. The demo application demonstrates (duh!) this by having two panes, one in which to write regex and the other in which to write XML. Changing one, also changes the other. It warns you of errors in both Regex, RegexConverter and then checks if what it got can be safely converted into your changed string!

Please tell me what you think. I believe it can be a real help in understanding regular expressions, some specific ones or regular expression syntax in general, whether one is a pro or a complete noob.

I've worked hard to design the library source in a way that is understandable, I also added comments everywhere. I tried to implement all the specifications of .NET regular expressions from the MSDN site so if you have a regex that is valid but can't be turned into XML or the conversion is not perfect, let me know.

The link is: RegexConverter on Github.

I will update this post with more science and links to my places of inspiration for this, so stay tuned.

Update: Ok, so I haven't updated this post much. Shame on me. I was reminded of this project when I got an email from the FamousWhy site which said RegexConverter "has been granted the Famous Software Award". I know it's probably an automatically generated message and went to many or all of the Codeplex people, but still: automated attention is still attention :)

Just a small mention of a thing I've learned today: what is the FullName of a Type of a class that is nested into another?

Usually, you would use it inside Visual Studio as ParentClass.ChildClass, but if you ever want to use the full name of the ChildClass Type (as when declaring it in web.config) you have to separate them with the + (plus) sign.

So let's assume you have this MyHandler class in the MyNamespace namespace which has a nested class Configuration in it. In the code you do a

var conf = new MyNamespace.MyHandler.Configuration();

while when using the handler type name to declare it in web.config you do it like this:

type="MyNamespace.MyHandler+Configuration"

.

First off, I have to say something about forums: stop copying content from one another, jerks!. I have been trying to find a solution for this problem and I found a zillion forum pages with the same "problems" and the same "solutions" again and again and again!

Much better! Now, I have been trying for an hour to understand why setting customErrors="Off" in the web.config of my ASP.Net application would not work. I tried just about anything, including the bloody forums. customErrors Off did not work!.

In the end I found one little comment for a StackOverflow question: set the retail setting in the machine.config file to "false"!! So, go to %WINDOWS%\Microsoft.NET\Framework\v2.0.50727\CONFIG\machine.config and set retail to false!. Setting it to true means it will NEVER show you any useful debugging message.

A while ago I wrote a post detailing a fix for Very slow UpdatePanel refresh when containing big ListBoxes of DropDownLists. I have restructured the fix into a ControlAdapter and placed it on Github.

Enjoy!

Today Internet Explorer 8 appeared to me in the Automatic Updates list. I have been using IE8 for months now and so I was glad that the official release finally came out. So I downloaded and installed IE8.

The installation process has several steps. First is the removal of any previous version of IE, then a reboot, then several small steps of the setup program: Downloading IE8, Downloading IE8 updates, Installing IE8, Installing updates, Finishing installation. Well, for me, at the Installing updates step it threw an error that said the installation cannot complete because the station is shutting down, then my computer restarted.

I did have Internet Explorer available, though, so I tried a few pages. After the pages loaded, I was invariably getting an error and IE closed. "Internet Explorer has encountered a problem and needs to close. We are sorry for the inconvenience.". Well, so was I!

I was trying the installation of IE the third time now and suddenly a Java update traybar icon appeared. I updated Java, then I was amazed to see that IE was no longer crashing! So, my solution for Microsoft IE errors: update Sun Java! :)

Well, the Java update probably completed some steps that the installer failed to. But still :)

And interesting link I found regarding any IE error that causes the browser to excuse itself and leave is on Sandi's Site

I've gathered the strength to defeat my laziness and put another of my projects on Github. I am talking about the Sift3 algorithm, described here.

The URL for the project is https://github.com/Siderite/Sift3/ where you can download the library and sources in .Net 3.5 C#. A class that also implements Levenstein and Length string distance algorithms is provided.

Please let me know if you are interested in the project, have any suggestions or are even using the algorithm in your projects.

This error can happen in several situations. One of them is when you are trying to access a web service, another is when trying to call a classic asp page. Most of the time, this happens in situations related to URL rewriting. It may happen in Windows XP or Windows 2000, on IIS 5.0 or IIS 5.1.

Well, first of all, in order to do URL rewriting you need to make ASP.Net process ALL URLs, not only .aspx pages. To do that in IIS5, you need to go to the ISAPI extensions and add a new one for '*' that maps to the ASP.Net dll (aspnet_isapi.dll). This process is detailed in this Microsoft page: HOW TO: Use ASP.NET to Protect File Types. What that means is that when you see a GIF image, it will pass through the ASP.Net engine, firing all the usual events.

However, after you do that, you see that web services start behaving strangely. Why is that? One explanation says that "405 mostly comes about when you try to POST against a URL that is not considered dynamic by IIS". It doesn't much makes sense to me.

I have searched a lot for an elegant solution. The only one that actually applied was using a piece of code in the BeginRequest event in Global.asax (or maybe in a HttpModule that one has to register in web.config). It came from this forum: HTTP verb POST not allowed. Here is the code:

//The BeginRequest event is fired for every hit to every page in the site
    void Application_BeginRequest(Object Sender, EventArgs e)
    {
        var extensions = new[] {".asmx", ".svc"};
        foreach (var ext in extensions)
        {
            var index = Context.Request.Path.IndexOf(ext, StringComparison.CurrentCultureIgnoreCase);
            if (index <0) continue;
            
            var path = Context.Request.Path.Substring(0, index + ext.Length);
            var pathInfo = Context.Request.Path.Substring(index + ext.Length);
            var query = Context.Request.Url.Query ?? "";
            if (query.StartsWith("?")) query = query.Substring(1);
            Context.RewritePath(path, pathInfo, query);
            break;
        }
    }

I needed a dynamic menu control for my site. So, of course, I tried to use the ASP.Net Menu control (with its many failings when following CSS standards). It was a painful failure. It didn't work in either Internet Explorer 8 or FireFox 3! That was especially strange since I had used the control in a bunch of sites and it worked back then!

Long story short:

<DynamicMenuStyle CssClass=adjustedCssIndex />

where adjustedCssIndex is a CSS class that specifies the z-index property:

.adjustedCssIndex { z-index:100; }

Long story, it seems that the control assumes there is a default z-index value set by the browsers; Bertrand Le Roy from Microsoft says as much in his blog, and discloses a patch fix.

However, as you can see in that post's comments, there is also a very simple CSS fix to all of this, by specifying the z-index.

The finale of what seemed to become my favourite sci-fi series ever (at its beginning) left me with a bitter taste in my mouth. Not only it makes no sense, but it is overall incredibly idiotic. If you haven't watched the end of the series, don't read further, because this is the mother of all ranty spoilers ever.

What makes it so emotional for me is not only that I really liked the show, but that this was not a show that was ended because of lack of planning or budget or the economical crisis, so nothing was rushed or changed. This was "meant to happen". And it sucked! Sucked worse that a vampire caught in the event horizon of a black hole that is falling into another black hole!

Not only did it not explain anything in a manner that would make sense to me, but instead it went completely overboard on all the things that I hated in the show. God exists, he somehow planned all this (oh, yeah, real modesty here, mr. Moore!), the model 6 in Gaiuses head was an angel, so was Kara Thrace, in the end they all reach Earth (this Earth) and decide to leave all technology behind (they throw the ships in the sun!!!) in the hope that starting anew would make them "break the cycle" and Hera became the chromosomal mother of all future humans. I guess leaving all that technology behind wasn't a good survival strategy for the rest of the 38000 people left alive, was it?!

If everything was God's plan, then there was no cycle except in its brain!! Forgetting mistakes is NOT a step towards not repeating them. Leaving behind technology is just as stupid! And ending the show with a couple of angels walking on Earth now and making bets on if we repeat the mistakes again or not, with background videos of the latest developments in robotics was.... there is no word in the English language for it. It is dumber than creationist! And the last half of the last episode was all about people saying goodbye to one another then going to live alone (read DIE!!) somewhere!

There is a glimmer of hope left though. The centurions were given their freedom and the last baseship. I will be looking at the sky hoping for them to return, nuke Moore and then air an all Cylon TV show about how they didn't repeat any mistake and just carried on!! Gods, this was frakking retarded!

And, of course, there is one more good thing in the series, and that is the Bear McCreary's remix of Bob Dylan's/Jimmi Hendrix's All Along the Watchtower. I am embedding the video with the cool transition from simple piano to all the instruments. Pretty cool!.

Guess what? F***ing YouTube removed the video because of a copyright infringement. What? One minute and a half of a movie scene? Geez! Couldn't find the same scene, so I am embedding All Along The Watchtower.

[youtube:qMo7WybtTWI]

The sound bit of the scene, sans the scene, can be found here. You can also see the live performance of the song here. You might also want to try Bob Dylan's original song.

Update: check out this Google event with McCreary playing the BSG theme with Raya Yarbrough as the vocalist.

This is how Dexter, the animated character in Dexter's Laboratory, begins his days. His enthusiasm for his work is amazing and inspirational, but real people rarely feel this way. Today is one of those great days for me! I woke up late, I took a long shower, I've eaten some good food, I've watched one of my favourite anime shows while getting dressed and I came to work in a beautiful sunny spring morning.

What amazing energy and how great a feeling! I feel I can do anything! Throw your legacy ASP applications at me, make me do stupid changes in even dumber web CRMs, give me the most menial degrading tasks you can think of, I will eat them whole and spit out gold! I am unstoppable!

Ok, you first must know that Microsoft added a specific namespace for working with ActiveDirectory in .NET 3.5. It is called System.DirectoryServices.AccountManagement. The principal objects in the Account Management API include computer, group and user objects and it provides a means for applications to extend the object model to include custom schema object types.

Now that this is out of the way, I want to take my simple Authenticate, ListUser and ReadUser methods (working fine with the old DirectoryEntry method) and translate them into this new way of doing things.

I did some code and I got this error message: "80075000". That is the actual error Message property! Even more remarkable, a google for 80075000 resulted in only 246 results!! none of them explaining what I did wrong. Apparently, I had sent a string in the format "LDAP://ComputerName" as the computer name. Maybe it helps someone.

But this didn't solve it. I changed it with just the computer name, with or without a "\\" prefix, and I got a more clear text, but just as vague UnauthorizedAccessException: "General access denied error".

I am still working on it, but damn, how can a programmer not think about the error message he passes to other programmers?!

Ok, made it work, here is the code for the user authentication in an Active Directory domain.
Classic DirectoryEntry:

bool authentic = false;
try
{
   var entry = new DirectoryEntry("LDAP://ComputerName", 
    "Domain\\username", "password");
   object nativeObject = entry.NativeObject;
   authentic = true;
}
catch (DirectoryServicesCOMException ex){}
return authentic;

As you can see, not the most elegant approach.

The .NET 3.5 way:

using (var context = new PrincipalContext(ContextType.Domain,"ComputerName"))
{
    return context.ValidateCredentials("username","password");
}

Much better, isn't it? Pay attention that in the first case you need the domain in the username and in the second you need it not to be part of the username!

There is a third way (System.DirectoryServices.Protocols.LdapConnection), but seems too complicated to address right now.

The new spring heat seems to have brought not only flies out of hibernation, but also some cats! Their behaviour is strangely similar to that of the flies, though, hmm :)

Recently I have been working on this Sharepoint project. I took it more out of curiosity as I didn't know anything about this piece of software. Now I know a lot more, like how hellish it is to code against it :) But it is also not a bad idea.

In case you don't know (as I didn't) Sharepoint is something like an ASP.Net site designed to work within a company, as an internal tool, allowing a lot of customizations and security from the web interface, with no code required. The desired end result is something looking like the IGoogle or Yahoo home pages, with web parts that can be configured, moved around, minimized, closed, made to interact one with another. Sharepoint Services is in itself a free software, but it only works on a Windows 2003 server or higher, so it sucks that way. Also, there is no real Sharepoint support for Visual Studio and most of the tutorials you find online are either too specific (blogs and such) or too vague (Microsoft style).

Also, there is a lot of confusion regarding the use of the interfaces in the Sharepoint dll, most of which have been obsoleted when the web part engine from .Net 2.0 was introduced.

Ok, short list of steps on how to start making a Sharepoint project in Visual Studio, assuming you code in Windows XP:

1. Install Windows Server 2003 on another machine (virtual or not)
2. Download and install Sharepoint Services 3.0 SP1 on it
3. Get the Microsoft.Sharepoint.dll file and copy it on your XP machine somewhere
4. Download and install the Sharepoint SmartTemplates for Visual Studio
5. Update the WSPBuilder application and some batch files in the template files
6. Start Visual Studio and create new project from installed templates
7. Add a reference to the Microsoft.Sharepoint.dll library
8. Code!

.

Now for the long list.
In order to install Sharepoint Services 3.0 SP1 you need to also install .Net 3.5 SP1. Actually it is a good idea to install this as well as the Visual Studio 2008 SP1 before you do anything (Sharepointy or not). Here is a link.

The Sharepoint SmartTemplates actually create a folder structure that is then used by the WSPBuilder utility to create the WSP file that installs a web part in Sharepoint. You can either import it in the site (upload) or use the setup that is provided with the templates. The problems I met when using it are linked primarily with the version of WSPBuilder that is included in the templates I've downloaded (version 0.2).
So first locate the installed template files: you can usually find them in My Documents/Visual Studio 2008/Templates/Project Templates/ as two zip files. First step is to download the latest WSPBuilder and replace it in the archives. The next step is to change the WSP/createwsp.bat file like this:

@ECHO OFF
DEL .\$safeprojectname$.wsp
ECHO Copying DLL ...
XCOPY /Y ..\BIN\$safeprojectname$.dll .\80\BIN\
ECHO Copying ASCX files ...
XCOPY /Y ..\*.ascx .\12\TEMPLATE\CONTROLTEMPLATES\$safeprojectname$\
ECHO Building WSP ...
..\WSPBuilder\WSPBuilder.exe -WSPName $safeprojectname$.wsp -BuildCAS false -SolutionID $guid2$ -DLLReferencePath "[the folder path where you copied Microsoft.Sharepoint.dll]" -TraceLevel Verbose
ECHO Copying WSP file ...
XCOPY /Y .\$safeprojectname$.wsp ..\SETUP\

The bold parts you must add to the file. The delete because otherwise you might be able to compile the project using the old WSP file if the WSPBuilder run fails. The others is in order to be able to compile the WSP using the sharepoint library and see any errors that might occur.
Alternatively, you can change the WSPBuilder/WSPBuilder.exe.config file with the DLLReferencePath and TraceLevel options.
Ok, now repack the folders into the archives and copy them back.

Now, after you build the project, you will have a Setup folder in the bin folder. That you must copy to the Windows 2003 computer and run. It will install the web part(s) in the project. In order to add more webparts to the project and make them compile in the setup project you need to alter the WSP\12\TEMPLATE\FEATURES\SmartPartTemplate\manifest.xml file and describe the files you add to the project.

After you run the setup project, you have to open the Sharepoint site and go to Site Settings -> Site collection features and activate the web part. Only then you can actually add it to a page.

Sounds complicated? Well, read more :)

Make sure that when you have finished with a web part you DO NOT DELETE THE SETUP PROJECT, but run it to remove the web part first! In order to remove an install web part from a Sharepoint site you must delve into the hell of command line utilities! Well, it's natural to me, but I am an old guy! ;)

Just supposing that you have done the undoable and you managed to delete a setup project with the part still installed, you must use the stsadm tool.
First find it on the Windows 2003 computer (with Find Files) then add the containing folder to the path. Then run cmd in the Start/Run menu and use the following commands:

stsadm -o enumsolutions

will enumerate the installed solutions. Remember the name of the solutions you want to remove.

stsadm -o retractsolution -name "[name of solution]" -immediate

will retract the web part project and allow you to delete it.

stsadm -o deletesolution -name "[name of solution]" -override

retracting doesn't always work, so the override option will force a delete.

Sometimes you manage to change the GUID of the project and you get an error like A solution with the same name "SomeName.wsp" or id "Some GUID" already exists in the solution store.. You delete it, but you still have this error. Try to install the solution with the setup project. Wait for the error, exit the setup project. Use stsadm -enumsolutions to see what the GUID of the project is, copy it, replace the SolutionId GUID in the setup.exe.config file with this one. The setup should then work.

This is about it. I've wasted a few hours to learn all of this. I know it's not terribly organized, but writing something is better than sharing nothing.

I haven't been writing for a while, but that is because I was working! Amazingly so, as I am not known for my willingness to work. But that also has its boons, you know, as I will not only gain material wealth for my wife's shoes, but also material for the blog! :) You will have to wait a bit for that, though.

Instead, I will talk about three little gems I found while browsing ShellCity. In case you don't know, ShellCity is a blog dedicated to the tools, not the result. Every day four free utilities are being presented in this retro looking site. Anyway, without further ado:

Fences. This allows you to organize your many desktop icons by grouping them into labeled transparent folder like structures. Not only does it make your desktop look better and feel better, but when you change your desktop resolution, it also remembers the location of these groups so that when you revert, you get them in the same position! Great thing to have and my favourite in this post.

MSVDM, or the Microsoft Virtual Desktop Manager. This is pretty old stuff, but I've only recently discovered it. It is NOT an exe file and it will not be installed in the Start menu! Instead it is a taskbar toolbar. It shows four buttons which allow you to switch between four different desktops. The desktop icons remain on all of the desktops, but you can define a different background for each and the opened windows are different from desktop to desktop. So, what you use it for is to open a group of utilities based on context. As an example, open a Visual Studio and some browser windows regarding a certain project, then open another Visual Studio and some other browser windows for another project that you work on simultaneously.

WizMouse. This one is not something you immediately go Wow! about. It sits in the traybar and does only one thing: it scrolls windows when you move the scrollwheel. But it doesn't scroll the active window, but the window directly under the mouse pointer! A lot of annoyance is saved by this.

Hope it helps you all. Till next time!