In these days we're totally involved with the launch of our Navision addon (I can't talk too much about it, but it's a complete module for business analysis) and the demo we're doing was the chance to keep in contact with a new future Navision module (that Microsoft will launch with Navision 4 Service Pack 1): Business Analytics for Microsoft Navision.
This tool is a powerful data analysis tool that, by using Online Analytical Processing (OLAP) from Microsoft SQL Server 2000, permits you to organize all your business data into information units (cubes) and (by using a familiar Microsoft Outlook-style interface), helps you to presents this information as you want on your desktop or on reports.
We've received a test version from Microsoft and today I've played a bit with the tool to organize a demo (we want to interact with this tool from our addon).
First, Business Analytics requires that you have a SQL Server with Analysis Services installed (Service Pack 3). When you've installed the requirements and start the tool, you can start a new data analysis by pressing the New Analysis button.
At the left side you've the data menus (Outlook 2003 style) and by pressing the Source Data menu you can see the SQL Server cubes available on your system.
I've selected the Sales (Invoicing) cube, that has the structure below:
What I want to do is a simple multidimensional analysis: I want to analyse the Sales Invoice amount per Business Posting group (companies) by Posting Period (Years). I've created a new multidimensional table (by pressing the Table menu, New Cross Tabulation) and on it you can easily drag'n'drop the dimension you want on your analysis and launch the calculation: the tool calculates the multidimensional data and shows you on the table:
We've the Sales Invoice amount for every company calculate by years (as you can see).
Now I want a graph with the Sales Invoice Amount per day of week. On the toolbar, I press the Object menu, New Pie Chart and a graph is added. I've to set the dimension for this graph, so I choose (with a simple drag'n'drop from the cube on the first figure) the Posting Period (Year) and the Sales Invoice Amount measure. Now launch a calculation for the graph and the result is this:
What is really powerful is that the multidimensional table and the graph are linked together (drill down). If I want to see how the Sales Invoice Amount for the year 2003 is distribute on the week, I've only to select the Year 2003 column on the table and the graph below will be automatically calculated. Here you can see the result:
The same if you want to see how the Sales Amount for a company is distribute on the week. You've only to select the company (a row) on the multidimensional table and the graph shows you the new result:
And if you want how the Sales Invoice amount for a certain day of the week is distribute during the year for the companies? You've only to click on the righ place on the graph (for example on Monday) and the table values will be recalculated accordingly:
Really powerful and easy to use... 
But now arrives the problem for me... after this analysis, I've tryed to interact with Navision by building a cube from the Navision environment and work with it with Business Analytics. Unfortunately there's not documentation on how to perform this task and I've see that it's not so easy.
To work with this preview version of Business Analytics with Navision 4, you've to import on Navision some objects (.fob) for the tool administration. You'll see the new menu on Navision in the Administration menu, under Application Setup, General, Business Analytics Setup.
Here you have a tool to create a cube by working with the Navision tables (you can create how many cubes as you want):
and after you've create your cube, you can save it on your OLAP system. To perform this task, you've to do 2 steps:
- Under the Functions menu, you've to press Update XML. This will create the structure.
- You have to set the configurator path (it's the program called BAConfigurator40.exe under the program directory installation, normally C:\Program Files \Business Analytics\Configurator\) and press the Start Configurator button. This process will launch the tool that creates the new OLAP cube on SQL Server.
For me this works all good... and now? I've opened my Business Analytics to work with my new cube but I've not find a way to view the new just created cube. 
Business Analytics seems to ignore it. Why this? In the next days I'll try to discover this thing... 
However, the tool is really powerful and opend a new way to perform complex data analysis. Microsoft has say us to promote the tool and I think that it will not be difficult to do.
I'm not an RSS Aggregators fans because usually I love to read my favourite
blogs via web, but on my Notebook I've an RSS aggregator installed (sometimes
could be useful when I'm in travel) and my choice lots of times ago was to
install Sauce
Reader.
Sauce Reader was a good aggregator for my usage, with lots of
features, with the nice interface of Office 2003 and totally written with .NET
Framework.
Today I've checked the product site for the new version availability and I've
discovered a curious thing: SauceReader 2.0 (the new version) is extremely
more fast than the previous version and the new program is totally rewritten
without using the .NET Framework.
These are the official reasons: "Sauce Reader 2.0 is a complete rewrite
and no longer requires the .NET framework. Using .NET for Sauce Reader v1 we
were never able to achieve the level of performance we considered appropriate
for a heavily used productivity application. .NET is a compelling and powerful
platform, but currently unsuitable for widely adopted client side
applications".
Lots of times we've talked about performances in .NET and this is a real
case: if you want an high performance application, maybe you've to use languages
such as C++, but I think that you've also to check your code in an extremely
accourate manner... sometimes little things not well optimized slow down your
application incredibly. I'm sure that the Sauce Reader staff has worked a lot on
optimization and they have the right reason for a total rewritten of the
program, but these things sounds like an alarm.
.NET Framework must be optimized?