April 2006 Blog Posts
Visual studio's DataSet visualizer is ok but it's not that great :))
This one is excellent and if you do a lot of db development (which you do if you're reading blogs on SQLTeam :)) you need this.
You can get it here.
As i saw this word being used with blittable and non-blittable datatype
i couldn't help but laugh... who comes up with this stuff??
Blittable types are datatypes that don't require conversion when passed between managed and unmanaged code while
non-blittable datatypes are represented differently in managed and unmanaged code.
Complete explanation here
I've posted an article about WebDAV protocol and my use of it for accessing MS Exchange store. WebDAV is based on HTTP 1.1 and is based on sending XML requests to find, set, remove and search for item properties.
Microsoft recommends that Exchange access from .Net is made with WebDAV. This MS article explains more.
It also works with Office documents...
you paste the code you want to display on the page and it creates the color coded html. preety cool...
there's even a source of the app that does the convert that you can view.
Maybe i'm reinventing hot water but no matter. It's fun for me :)))
These 2 functions round the number up or down on the decimal position we specify.
create function dbo.FloorEx(@num decimal(38, 33), @decimalPosition int)returns decimal(38, 33)as beginif @decimalPosition < 0 set @decimalPosition = 0return floor(@num*power(10, @decimalPosition))/power(10, @decimalPosition)endgocreate function dbo.CeilingEx(@num decimal(38, 33), @decimalPosition int)returns decimal(38, 33)as beginif @decimalPosition < 0 set @decimalPosition = 0return ceiling(@num*power(10, @decimalPosition))/power(10, @decimalPosition)endgo declareselect @num decimal(38, 33), @decimalPosition int @num = 10.123456, @decimalPosition = 3select dbo dbogo @num, .FloorEx(@num...
If you do a lot of managed/unmanaged code interaction with COM objects
then this is a page you must bookmark.
I stumbled on an interesting thing yesterday while playing... :))
Seems that getting all data from a table is faster if we add a "phantom" where part
that forces a clustered index seek instead of a scan.