Friday, September 9, 2011

important starting facts for Spider & Co

ToDo:

1. simple setup + download server for deployment. registrar server for licenses.
this server shall provide all existing componnets. also spider & Comoponents on that distribution server.

2. the website shall have multi user security and a blog+support forum. I suggest wordpress+bbpress.
we have to select a theme and get clear about the identity, company etc. who will moderate the site.

3. test scenarios to get robust to deploy and work. which operating systems to test deployment and which load tests are perfromed automated. 1-5 cpu in a simple intranet. 5-10 cpu in a routed intranet. 20-30 cpu on a distributed net all at maximal load capacity. maximal number of connect/ disconnect need to be tested.
lost connection and reconnect.

4. set up a web server on HOST_EUROPE to run the web server demo app for visualisation of demo data. a spider shall run on that server, which allows anyone to connect and work with it.

5. demo app data simulation and real IO system in gaggenau testcell


Sunday, August 7, 2011

C# coding conventions

nice idea:

http://www.centerspace.net/the-elements-of-csharp-style/

Fastest and most comprehensive charting

looks like this is the most massive one.

Lightning charts

these guys use the NMath library for their math:

CenterSpace

FloatVector v1 = new FloatVector(length);
FloatVector v2 = new FloatVector(length);

// Summing v1 and v2 will yield a vector of all 1's.
FloatVector sum = v1 + v2;

// Vector scalar operators are also provided.
float a = 2;
FloatVector asum = v1 + a * v2;

Filtering data code:


int numberLeft = 4;
int numberRight = 5;
DoubleVector filterCoefficients = MovingWindowFilter.MovingAverageCoefficients(numberLeft, numberRight);
MovingWindowFilter movingAverageFilter = new MovingWindowFilter(numberLeft, numberRight, filterCoefficients);

DoubleVector filteredSignal = movingAverageFilter.Filter(signal, MovingWindowFilter.BoundaryOption.PadWithZeros);

Summary:
I like the vecto math style for arithmethics. I assume this library could be used to allow performant math processing

the NMath is based on BLAS