Google is working on a new language named "Go" and they call it a "systems programming language". It is not the intent to make a review or to compare it with other existing language, far from that, but it worth having a look at that new language.
First, the syntax is similar to the languages that most of the people are used to. It is more or less the same as C or Java. A bit closer to C anyway for some reason described further in this post. One notable difference is that the type is defined after a variable name, for example.
One of the goal of the language is to simplify the code or to make it less verbose. But, as a tradeoff, it makes the code reading more difficult, or maybe it is just a matter of getting used to the syntax. For example, there is only one keyword for the loops : for
A thing I find nice is the possibility for a function to return more than one value and therefore they introduced multiple return functions.
The other features of the language are summarized here :
Anonymous functions
Slices; a part of an 0-based array
Maps; other word for dictionary
Object orientation using structures; which makes it closer to C than Java and in my opinion not the best thing the language could have
Interfaces
No inheritance; Instead of supporting object inheritance, the creators of the language prefer to use composition of structs
Real-Time flavour; My ADA background reveals to the surface when I see the constructs of Go to support multitasking. What I liked a lot in ADA was the possibility to write methods that were callable from other tasks and how easy it was to create a multi-task program. With Go, it seems that simple. You define a method and rather than calling it normally, you prefix it with the keyword "go" :

go MyParallelTask;

Communication in a multi-task program can also be a challenge, being just to exchange information or data or just to synchronize two tasks. With Go, you need to instantiate a channel and use arrows (channelname <- value and variablename := <- channelname) to send to or received data from the channel. That channel can be synchronous or in other words blocking enabling the synchronization between two tasks or can by asynchronous (under certain limits given by the buffer size of the channel).

On the side of what is not in Go, we have :
Generics or templates
Exceptions

For those two points, the Go team has not yet decided if they would add these features for the reasons that can be read here : http://golang.org/doc/go_lang_faq.html#generics

In conclusion, it is too early to say if this language will succeed and gain many market shares in the programming world. It will mainly depend on the built-in libraries and the platform that will support it. Less important for such system languages is the interoperability features and how a Go program will be able to communicate with the external world. Another success factor is the developer community. Today, Go is only available for Linux or Mac OS X and probably some tools or IDEs would be needed in the future, but not sure because as it is mentioned a number of times, the aim of the language is system programming language and therefore not playing in the same field as C# and other high-level languages.

Title : Foundations of BizTalk Server 2006

Author : Daniel Woolston

Summary :
BizTalk Server 2006 was until last year, the latest version of the middleware product from Microsoft. Since, the new version, BizTalk Server 2009, has been released. Despite BizTalk Server 2006 is not the first or second version of the product, it really needs documentation on how to use and develop on this platform as it is absolutely difficult or impossible to start on such product like this.
The book is structured around the different component or pieces of BizTalk : Messaging, Schemas, Maps, Ports, Orchestrations. It covers all the aspects of the product even tackling the application deployment.
It is full of useful screenshots and the explanations are clear enough even if you don't have the software at hand.

Review :
Working with BizTalk since couple of years, I found the book quite basic. Ok, the title contains "Foundations", so don't expect advanced explanation but rather take it as a first look at BizTalk. A good example is to read it before an intermediate or advanced training. At the end, it gives an excellent overview of what can be done and how can specific problems can be solved with BizTalk Server 2006. Another good point is that it is not a big pad to read and can be absorbed in a week-end :-)

Summary :
F# is a new language that is coming in the pipe of Microsoft for the Visual Studio platform. It aims to tackle the functional programming paradigm eventhough it is possible to use the imperative or object oriented programming.
Robert Pickering starts his book by explaining the basics of F#, how to get and how to use the tools. Then, the book describes the F# syntax to be used in the three language paradigm, functional programming, imperative programming and finally object oriented programming. Among other things, the notion of type inference is presented. Once the syntax is presented, the book describes the way to develop web, windows or even WPF applications using the .NET framework. Data access is also addressed using the current technologies available, such as ADO.NET or LINQ. Then, a quick look at DSLs, compilation and code generation is given, presenting the lex and yacc tools coming with the language. Finally, a full chapter is dedicated to the interoperability between .NET and F#, because even if F# is based on the CLI, the language introduces several types that are not available in the other .NET languages (C# or VB.NET).

Review :
Discovering a new language is really interesting and with F#, it is the occasion to see a new paragigm, functional programming. In really short, with F#, everything is a value, even a function. It means that you can use a function as a function parameter. The concept of type inference is also very attracting. The book is very easy to understand and a lot of little examples are explained in details, making the reading very fast. The first half of the book is dedicated to the language itself. The second half is more on using the .NET framework and I would say that it is the less interesting of the book. Indeed, during the first part, you have came across various examples using types and classes of the framework and user interface development being web or windows, or data access meaning that the second part does not bring a lot a information. Once you know these topics from the .NET documentation or from another book and once you have read how to access the .NET BCL from F#, then this part is pretty straightforward and not really useful. Moreover, the examples used to depict the topics are more explaining how to use the BCL classes than the language itself. Nevertheless, the last parts discussing the interoperability and the possibility of generating DSLs are more interesting.
My final words are that it is a very intersting book if you want to see another land (functional programming). Unfortunately, on my bookshelf, I also have "Expert F#" that I just opened to see what is inside and I saw that it takes the explanations and descriptions of the language from the beginning. If I had knew that before, maybe I would have bought this one instead. So, if the goal is just to scratch the surface of F#, "Foundations of F#" is the best suited, otherwise, if the goal is to go really deeper in the topic, then prefer "Expert F#" (a review of that one will be posted).

I have been needing a BPMN and a UML 2 stencil for Visio 2007, and, truth said, I think I found my happyness in the following stencils :

For UML 2, I found one at softwarestencils. There are plenty of different stencils, even for Visio 2007. It offers all the shapes, but for some of them, you need to change the properties of them to have them appearing on your design surface. So, for UML 2, go there.

Regarding BPMN, the one proposed by Tyner Blain is a goodfit and it seems, at a first glance, it fits all my current needs. It offers all the BPMN shapes, the only thing missing, maybe, are the swimlanes.

It is done !

I successfully passed the 70-631 exam and I am now MCTS on Sharepoint, more precisely "Configuring" SharePoint.

It was not so easy (that is why it is a certification) and it required a bit of work in order to be able to see the different aspects of the SharePoint configuration. MeasureUp helps, but when you are in front of the screen for the exam, you realize and you thank the experience you have with SharePoint...

But, it is done, and now, go for the next one...

The coming weeks, I am going to have a look at the latest Beta of Visual Studio 2008 Team Foundation Server and Microsoft has done all it was possible to make it easier (or almost).

First, they released Virtual PC hard drives (vhd) with all the software included. So, as I have the luck of having a MSDN subscription, I first downloaded the 9 parts of the VHD file. Yes ! A total of 4.2GB to download, taking the whole night to get it. But, once deflated, it takes more or less 13GB ! That's ok for me, on an external 420GB hard drive, I can afford it.

The only problem I had was that when I launched the VPC image, it simply failed, telling me that "The parent virtual hard disk has been deleted or has been moved from its previous location. Please select the new location of the parent virtual hard disk". That was strange, because, on the MSDN subscriptions site, there is no mention about a parent virtual hard disk. Or, at least, not enough visible for me.

To solve this little issue, it is better to download the VHD from the Microsoft download site. There, you have all the instructions you would need, and, the most important, where to download this parent virtual hard drive.

Link to the Visual Studio 2008 Beta 2 Team Foundaction Server download

I found a very interesting post about "How to build a SharePoint development machine". I think there is a lot of good points there, along with a list of useful tools to develop SharePoint application smoothly.

Thanks to Eli Robillard (MVP SharePoint) for his excellent post

When you have a lot of pressure on a project regarding the planning especially, sometimes, just a little thing that makes your developer life tougher is not supportable.

It was the case this week, during a migration from an application written in ASP to .NET. I migrated several dialogs that were supposed to be displayed as modal dialogs. During the developments and the testing, everything went fine. No bugs.

But, when integrated in the whole application, it simply did not work at all, just because my .NET web forms were displayed using the showModalDialog javascript function. In one of them, I had a datagrid with a column containing hyperlinks. What a surprise when you click on the link and that a page with the url javascript:__dopostback(...) opens ! And you know why ?

Just because you have to add a <base> tag in your head section of the page, like below :

<head>
   ....
   <base target="_self">
   ....
</head>

I spent almost half a day to fix this, but I also learned it.... :-)

In this project, I had to work on an Excel file getting its data from a database in order to generate a document usable for the end-user. The problem was that a lot of pivot tables were included everywhere in the document making the manual update of these tables really painful. So, my problem was to automate the process of this update in order to refresh all the pivot tables in all the worksheets. In finally wrote a little piece of code that could be triggered either manually or automatically after an event, for example.

As it could be also useful for you, here is the VBA code :

' Scans all the Pivot tables contained in the file and refresh the data
Public Sub RefreshPivotTables()
   Dim aWorksheet As Worksheet
   Dim aPivotTable As PivotTable

   ' Find the pivot tables in all worksheets
   For Each aWorksheet In Worksheets
      ' If a worksheet contains more than one pivot table, refresh all of them
      ForEach aPivotTable In aWorksheet.PivotTables
         ' Updates also the range of data to be sure that the pivot table uses all the data
         ' If the range of data never changes, it isnot necessary tocall the sub below.
         RefreshPivotTableRange aPivotTable
         aPivotTable.PivotCache.Refresh
      Next aPivotTable
   Next aWorksheet
EndSub

' Adapts a pivot table data range
PrivateSub RefreshPivotTableRange(aPivotTable As PivotTable)
OnErrorResumeNext
   Dim newRange As Range
   Dim WorksheetName AsString

   ' Get the worksheet name
   WorksheetName =Left(aPivotTable.SourceData, InStr(aPivotTable.SourceData, "!") - 1)
   ' The GetDataRangeForWorksheet is a sub that returns the new range of data in the given worksheet.
   Set newRange = GetDataRangeForWorksheet(Worksheets(WorksheetName))
   aPivotTable.SourceData = WorksheetName & "!" & newRange.Address(ReferenceStyle:=xlR1C1)
EndSub

Couple of days ago, writing an ASP page enabling users to download attachment generated on the fly, I struggled a bit on a nice "Internet Explorer was not able to open this Internet site. The requested site is either unavailable or cannot be found" error. As I know that GIYVBF (for those of you who are wondering what it is stands for, try this)

First, I checked several times that I used the correct caching expiration and attachment settings without any success. Finally, the only thing I did not find was this :

If, during the generation of your attachment, there is an non-handled error or exception, the attachment genaration will abort and you will get this famous error message. Thus, this is another way of getting the nice error box....In my case, there was some code accessing a database and when it failed it displayed this message.

I remember a discussion I had with a colleague about workflow engine developments and moreover, about the real needs of starting such developments. But, the topic was not only and not really about the development itself, but more about the existence of such engines.

Last June, Gartner published a document presenting the major actors of BPM solutions. Among them, we can find K2 with a coming product called K2.NET, based on the new Windows Workflow Foundation from Microsoft. We can also find FileNet which has been bought by the giant IBM in the meanwhile, but also Fuego and last but not least, Microsoft with BizTalk Server 2006.

Then, on one side we have ready-to-use products, proprietary and requesting customization, and on the other side, Microsoft with the WF. Thus, the question that has come on the table was : When to use the WF and when to use BTS ?
Since this time, I tried to figure out the different criteria needed to help to decide. But the list is absolutely not finished and I am sure that a lot of points could be added.
In fact, what needs to be understood is that the WF has never been designed to replace BTS, not at all. In one case, we have a framework on top of which our application has to be developed and which will be part of the developed application. In the second case, it is a licensed product which needs a bit of understanding (I would say that the learning curve is quite big...) to implement orchestrations between existing applications, building the final solution. Nevertheless, these technologies are complementary.

First, BTS comes with a lot of features and functionalities that are not necessary needed for the majority of the applications.
But, if your application needs to interact with one or more other applications, hosted on different servers or using different technologies (such as Java, for example), BTS is certainly an advantage. With this, BTS allows document transformations quite easily, some of them or some of the file formats are already implemented or just need the installation of an add-in (HL7, SWIFT for example).
With BTS, it is also easy to know what is happening in the system using the BAM (Business Activity Monitoring) and when you need to share credentials or authentication informations between systems whether they are Windows or not, the use of SSO (Single Sign-On) is a big plus over the WF.
At the opposite, if the application has to contain the workflow, the use the WF. And finally, if none of the BTS features is required, of course, then use of the WF is quite obvious.

Again, this list is not exhaustive and would need more points to add, but, in my opinion, it is an interesting start when we think about using a workflow engine.

Comments are welcome...

The project I am currently working on involves a lot of Javascript code to validate values entered by the users and to ensure that business rules are followed in complex forms. Unfortunately, and as everyone knows, it is tough like the hell to develop and debug Javascript code.

However, I recently found an article talking about a tool that was just under my nose and that I was not aware of. Here, I speak about the Microsoft Script Debugger. And, in this, it is a real little pearl. Working with Visual Studio since long time, I knew it was possible to debug Javascript code by attaching the Internet Explorer process, but the files needed to be local. Otherwise it is impossible.

But, the Microsoft Script Debugger tries to find all Internet Explorer sessions and, by selecting the instance you want, it is then possible to do a "Break At Next Statement". The next line which will be executed will switch you to the debugger window allowing you to execute the application step by step. It is even possible to watch the values held by the variables. You don't need to have the files locally and it allows you to debug a production environment as well where it is not possible to access the source code.

Well I would not say I am going the best friend of Javascript, but at least, it makes this task less tough. This application is in the Start menu, under Programs, Accessories, Microsoft Script Debugger. With Windows 2000, if it is not there, it just need to be installed from the Windows 2000 version, as a Windows component.

For the other Windows versions, please follow this link.

If you are in the SOA (Service Oriented Architecture) field, you should go to the ZapThink's 5th SOA Practitioner's Conference in Geneva, at the Crowne Plaza Hotel.

This event, sponsored by my Company, amongst some others, will held several interesting topics such as the basics of SOA (Fundamental SOA Concepts) and real world application of SOA (Real implementation cases of SOA).

Go there and have fun with SOA...

Currently developping a little application using the Framework 2.0 and the System.IO.Ports.SerialPort class, I got this nice exception when trying to set the StopBits property of my SerialPort instance to StopBits.None :

Quite strange, isn't it ?

Knowing that GIMF, nothing to read about this problem, or a little in the mono version of .NET. So, I launched one of my favourite tool, Reflector, to see inside the System.dll (Version 2.0.0.0 in the Framework version v2.0.50727) assembly. And here is the result when we look at the set_StopBits sub :

As we can see, before setting the property to the value passed in parameter, a little check is done on its value. And if that value is less than StopBits.One, it throws the ArgumentOutOfRangeException. Maybe the guy who wrote this code forgot to add a "N" at the beginning of the enum value....

Or, if it is what they really wanted to implement, nothing in the documentation is telling it is forbidden to set the StopBits property to None.

If anyone reading this post has an answer, I would appreciate at its own value and I thank you for your comments....

C'est bon, j'ai reçu la confirmation de ma participation aux TechDays 2006 qui se tiendront les 5 et 6 avril à Genève.

Au menu :

Windows Vista, WinFX, Sharepoint et BizTalk Server 2006, entre autre.

A mon avis, les conférences sur LINQ, Atlas/AJAX et Aero risquent d'être fort intéressante

Au détour d'une rencontre avec un guru du SQL, j'ai pu admirer le code suivant, permettant d'effacer tous le doublon d'une table. Imaginons une table (aTable) avec une seule colonne (aNumber), aucune clé, mais avec plusieurs lignes contenant la même valeur :

1
1
2
1
3
3
3
1
2
1

Le but est du supprimer tous les doublons, ceci en une seule requête. Vous avez la réponse ?

Voici le code SQL :

set rowcount 1

select 1
while@@rowcount > 0
delete test
where 1 < (selectcount(*) from test t2 where test.id = t2.id)

set rowcount 0

Some couple of weeks ago, I had the opportunity to meet SQL guru, and I have seen a very nice SQL code, deleting duplicates rows in a table. Let's imagine a table (aTable) with only one field (aNumber), without any key but multiple rows with the same value :

1
1
2
1
3
3
3
1
2
1

The goal is to delete duplicate values in only one SQL request. You have the answed ?

Here is the code :

set rowcount 1

select 1
while@@rowcount > 0
delete test
where 1 < (selectcount(*) from test t2 where test.id = t2.id)

set rowcount 0

Une fois un SQL-Server et un BizTalk Server 2004 installés, il est loin d'être trivial, ensuite, de modifier le nom du serveur hébergeant ces services.

Le service de Single Sign-On est d'ailleurs le plus embêtant à remettre sur ses pieds. Le plus simple restant encore à rechercher les occurences de l'ancien nom du serveur dans la base de registre et de les remplacer par le nouveau nom (en tout cas là où cela est pertinent...).

En revanche, l'exécution de l'outil de configuration de BTS2004 (ConfigFramework.exe) va échouer sur la modification des jobs que BTS2004 installe dans SQL-Server. La cause de l'erreur provient du fait que le serveur ayant généré ce job n'est plus le même que celui qui essaye de le modifier. Pour éviter ceci, il est possible d'exécuter la commande SQL suivante (dans le cas où il n'y aurait que des jobs BizTalk. Dans le cas contraire, il faudra filtrer la requête) :

UPDATE msdb.dbo.sysjobsSET originating_server = @@SERVERNAMEWHERE
originating_server <> @@SERVERNAME

Ceci à pour effet de changer le nom du serveur original du job.

A noter que changer le nom d'un serveur hébergeant déjà des services tels que BizTalk n'est pas recommandé et ceci fait donc déjà partie des bidouilles...

Ecrire un composant COM en .NET est relativement simple, pour peu que l'on n'essaie pas de faire des spécialités.
Il y a néanmoins quelques petites choses à se remémorer si l'on veut éviter de perdre un temps précieux, et une de celles-ci concerne les tableaux. Imaginons que nous voulions écrire un composant COM en .NET, utilisable depuis le code VBA d'une feuille Excel. Le code pourrait être :

Public Sub mySub
   Dim myArray() asDouble
   ReDim myArray(1 to 10) asDouble

   Dim comObject As myCOMObject

   for i = 1 to 10
      myArray(i) = i
   next i

   Set comObject =New myCOMObject

   myCOMObject.MyMethod(myArray)

   Set comObject =Nothing
EndSub

Si vous exécutez ce code, une exception sera levée :
System.IndexOutOfRangeException: Index was outside the bounds of the array.

Ce qu'il faut savoir, c'est que sauf sur mention explicite avec la clause Option Base, les tableaux VBA sont indexés à partir de 0. Or, dans notre cas où l'on veut redéfinir la taille du tableau myArray, il faut se veiller à ne pas indiquer la borne inférieure, et ainsi la laisser à 0.
Du côté du composant COM, les valeurs sont bien entendus disponible à partir de l'indice 1. Ne pas essayer de lire les valeurs à l'indice 0 :

Double [] myLocalArray =new Double[myReceivedArray.GetLength(0)];

for(int i = 1; i <= 10; i++)
{
   myLocalArray[i] = myReceivedArray[i];
}

Writing a .NET COM component is fairly simple. As far as we do not try to do specialties. There are come little tricks to remember before spending lot of time. One of them is the arrays. Let's write a .NET COM component that should be used from a VBA code of an Excel sheet. The code could be :

Public Sub mySub
   Dim myArray() asDouble
   ReDim myArray(1 to 10) asDouble

   Dim comObject As myCOMObject

   for i = 1 to 10
      myArray(i) = i
   next i

   Set comObject =New myCOMObject

   myCOMObject.MyMethod(myArray)

   Set comObject =Nothing
EndSub

If you run this code, an exception will raise :
System.IndexOutOfRangeException: Index was outside the bounds of the array.

What is important to know about is that VB defines arrays using a 0-based indexes unless you specify the Option Base clause. In that case, I redefined the myArray size, but I had to not specify the lower-bound, letting that bound to 0.
On the COM component side, values are available only from the second cell (index equal to 1). Do not try to read the cell with index 0 :

Double [] myLocalArray =new Double[myReceivedArray.GetLength(0)];

for(int i = 1; i <= 10; i++)
{
   myLocalArray[i] = myReceivedArray[i];
}

Travaillant actuellement pour un client désirant utiliser un composant .NET depuis Office 2000, il a donc fallu s'attaquer à COM Interop. Et, bien que la documentation le dise, il est possible d'avoir quelques surprises lorsqu'il s'agit de gérer les exceptions et les erreurs COM.

Prenons donc le code suivant :

[ComVisible(true)]
publicint myMethod()
{
  // snip
  thrownew myException("Description", 2500);
  // snip
  // Do something here
  return 0;
}