Enum to List
I’ve found this method useful lately so I thought I should add it here where everyone can see it.
It is meant to be used with the .NET Compact Framework. In the full framework this would have been much easier -
without reflection.
What the method does is that it adds the enum’s fields to a generic list of the same type T as the enum.
public static List EnumToList( )
{
Type enumType = typeof( T );
if( enumType.BaseType != typeof( Enum ) )
{
throw new ArgumentException( “Not an enum type!” );
}
List enumFieldList = new List( );
FieldInfo[ ] enumFields = typeof( T ).GetFields( );
foreach( FieldInfo fieldInfo in enumFields )
{
if( !fieldInfo.IsSpecialName )
{
enumFieldList.Add( (T) Enum.Parse( enumType, fieldInfo.Name, false ) );
}
}
return enumFieldList;
}
C++ UNREFERENCED_PARAMETER
I am currently developing a driver for one of our custom made devices under Windows CE and I came across this macro -
UNREFERENCED_PARAMETER in one of the driver examples I took as reference for developing my own. The definition itself does not say much:
#define UNREFERENCED_PARAMETER(P) (P)
All it does is to layout the parameter or expression it has been passed. After some browsing on the internet I’ve found this post by Paul DiLascia:
http://msdn.microsoft.com/en-us/magazine/cc163805.aspx which I think greatly explains what is meant by compiling with Warning Level 4 – it made me a believer anyway.
Indeed, an unreferenced parameter would cause an error when compiling using level 4 and this is where UNREFERENCED_PARAMETER comes in handy.
For instance you have a function like this:
void foo( int arg1, float arg2 )
{
UNREFERENCED_PARAMETER( arg1 );
}
Since you won’t be using arg1 and also you will compile with level 4, you just fool the compiler by passing arg1 as parameter to a macro that just adds the line:
arg1;
to your code. And since it is not used the compiler does not generate code for it and no efficiency or space is lost.
I will not repeat here stuff written in that post, just thought I’d mention it as a great post on C++ I’ve found.
Install an application for all users
In this article I present techniques based on Visual Studio 2005 Deployment, C#, Windows Registry.
I have recently encountered one of those errors you should expect but you don’t. I have an application for which I have created an installer project, under Visual Studio 2005. The application works fine after installation for the administrative role user that installed it, but when switching to a less priviledge user on the same computer the install/ repair program starts for that user prompting for the installation kit path. As an aditional detail, the setup kit creates the application registry keys it requires under HKEY_CURRENT_USER key. And this is the flaw…
When creating registry entries for an application that might be used by more than one user, each of which having different access rights on the computer, we should place the registry key under the “user/ machine hive” node in the Registry view of the deployment project. This way we ensure that the keys get into the right place when the application is first installed. According to Microsoft’s definition, the keys placed under user/ machine hive will be placed either under HKEY_CURRENT_USER when the application is insalled only for one user and under HKEY_LOCAL_MACHINE when the application is intalled for all users. Here is a screenshot :
The problem with keys under HKEY_LOCAL_MACHINE is that they are read only for restricted users, for example user belonging only to the “Users” group. It is though required to assign special permissions for those keys a user not only to read but also modify. The reason I needed this was to save some user settings in the registry, so when a less privileged user saves the settings, he/ she must also have proper permisions.
To solve my problem I’ve created a custom installer class implemented an event handler for the Commited event. In this handler, write access permissions are given to users belonging to the “Users” group.
I give bellow the implementation of my event handler.
// Event handler for ‘Committed’ event.
private void CustomInstaller_Committed( object sender, InstallEventArgs e ){
RegistryKey hkCustom = null;RegistryKey hkTester = null;
// depending on the installation options, we will search for the application’s registry keys :
// 1. – if installation is performed just for the current user, the keys will be under HKEY_CURRENT_USER
// 2. – else if the installation is performed for all users, the keys will be under HKEY_LOCAL_MACHINE
hkCustom = Registry.CurrentUser.OpenSubKey( “Software” ).OpenSubKey( “CustomInstallerTester” );if( hkCustom != null ){
hkTester = hkCustom.OpenSubKey( “Tester99″, RegistryKeyPermissionCheck.ReadWriteSubTree /*required*/ );}
if( hkTester == null ){
hkCustom = Registry.LocalMachine.OpenSubKey( “Software” ).OpenSubKey( “CustomInstallerTester” );
}
if( hkCustom != null ){
if( hkTester == null ){
hkTester = hkCustom.OpenSubKey( “Tester99″, RegistryKeyPermissionCheck.ReadWriteSubTree /*required*/ );}}
if( hkTester != null ){
System.Security.AccessControl.RegistrySecurity regSec = new System.Security.AccessControl.RegistrySecurity( );// grant write permissions to the users group :
regSec.AddAccessRule( new RegistryAccessRule( “users”,
RegistryRights.ReadKey | RegistryRights.WriteKey | RegistryRights.Delete,
InheritanceFlags.ContainerInherit,PropagationFlags.None,
AccessControlType.Allow) );hkTester.SetAccessControl( regSec );}
if( hkTester != null ) { hkTester.Close( ); }
if( hkCustom != null ) { hkCustom.Close( ); }
}
Canceling the Load Form event in .NET
I was trying to find a way not to load a form when some error condition occured.
Briefly, the whole cancel logic should be placed in the constructor of the form and an exception should be thrown whenever canceling is necessary. The Show or ShowDialog methods must be included in the same try-catch block as the constructor.
Thus, when the exception is thrown, no form is displayed. There is no reliable way of canceling the Load event from within the Load event handler : if you call Close( ), you get an exception of the type : “Cannot acces a disposed object” on the Show or ShowDialog form methods.
See the following discussion also :
http://discuss.fogcreek.com/dotnetquestions/default.asp?cmd=show&ixPost=1286.
Ubuntu .ICEAuthority problem
Recently I’ve rebooted my VMWare Ubuntu machine and after the login sequence I’ve received an error whose log I give bellow.
/etc/gdm/PreSession/Default: Registering your session with wtmp and utmp
/etc/gdm/PreSession/Default: running: /usr/bin/X11/sessreg -a -w /var/log/wtmp -u /var/run/utmp -x “/var/lib/gdm/:20.Xservers” -h “” -l “:20″ “wes”
/etc/gdm/Xsession: Beginning session setup…
_IceTransTransNoListen: unable to find transport: tcp
_IceTransmkdir: ERROR: euid != 0,directory /dev/X will not be created.
_IceTransmkdir: ERROR: Cannot create /dev/X
_IceTransPTSOpenServer: mkdir(/dev/X) failed, errno = 13
_IceTransOpen: transport open failed for pts/ubuntu:
_IceTransMakeAllCOTSServerListeners: failed to open listener for pts
_IceTransISCOpenServer: Protocol is not supported by a ISC connection
_IceTransOpen: transport open failed for isc/ubuntu:
_IceTransMakeAllCOTSServerListeners: failed to open listener for isc
_IceTransSCOOpenServer: Protocol is not supported by a SCO connection
_IceTransOpen: transport open failed for sco/ubuntu:
_IceTransMakeAllCOTSServerListeners: failed to open listener for sco
** (gnome-session:23826): WARNING **: Unable to lock ICE authority file:
The solution I’ve found on a forum( http://www.linuxquestions.org/questions/showthread.php?s=ef11204ac2b7cacdd6cfc0b651512838&t=403160&page=3 ). To save you the trouble of digging for it in the posts, I’ll just mention what I did to make Gnome start :
Open the failsafe terminal and edit the file /etc/gdm/Xsession. Add the following lines right under the PROGNAME = Xsession line :
ICEAUTHORITY=”tmp/ICEauthority-${root}”
export ICEAUTHORITY
This worked fine for me since Gnome is now normally started.
Set PATH environment variable
This post may look shorter than it’s title. In order to set the PATH environment variable to include your custom path you should type the following on the command line :
> PATH=$PATH:/your_dir
That should do the trick.
Create an ODBC connection from VS2005 setup
I am developing an application in the medical field with an Sql Server database ( let’s name it “MyDB” ) . Recently some of our clients have been asking for a way to easily export the data from our database in order to be used into some other analysis oriented tool.
That would have been easily achieved through an ODBC connection in the created into the Control Panel. I looked up for an easy way to create the ODBC connection through the Visual Studio 2005 Setup project. It turned out to be as easy as that. All you have to do is create registry entries for your ODBC connection. In my case, I’ve created a new System DSN and afterwards just copied the registry entries into my own setup project, I erased the DSN I’ve manually created and then I’ve executed the setup kit to test whether it works or not. Everything went just fine …
Let me give an example for a connection that requires an SqlServer driver.
In the Setup Project, open the registry view and browse to the key HKEY_LOCAL_MACHINE\Software. Under this key you will have to create the ODBC.INI key. Then, under this key, you will create a key with your database name, in our case that would be MyDB. Here is how an export from the registry editor would look like, for the database name key :
[HKEY_LOCAL_MACHINE\SOFTWARE\ODBC\ODBC.INI]
[HKEY_LOCAL_MACHINE\SOFTWARE\ODBC\ODBC.INI\MyDB]
“Description”=”My database”
“Default”=”"
“StatsLog_On”=”Yes”
“Driver”=”C:\\Windows\\system32\\SQLSRV32.dll”
“Trusted_Connection”=”Yes”
“Server”=”(local)”
“LastUser”=”"
“Database”=”MyDB”.
You can easily copy these entries into your setup project under the database name ( “MyDB” ) key I’ve mentioned before.
Now, you must get back to the upper level ODBC.INI key and under it, create the key :”ODBC Data Sources”. You will then set the key : “MyDB” with the string value “SqlServer”. Here’s how the export looks like :
[HKEY_LOCAL_MACHINE\SOFTWARE\ODBC\ODBC.INI\ODBC Data Sources]
“MyDB”=”SQL Server”
So, now you have created, into your setup project, the registry entries you need in order to automatically create an ODBC connection upon application setup. If you want to test this, erase ODBC entry from your ODBC administrator tool in Control Panel, build up your kit, and when the setup finishes, check the registry editor again and view the entries right where they were before. Now, check the ODBC administrator tool and notice the ODBC entry.
On the other hand, when you need to uninstall the application, there is no need for you to do anything special as the uninstall process will remove the registry entries for you.
Please note that the example I’ve given works with an Sql Server database. For other types, you have to repeat the same procedure, because the registry keys will not be the same as in this case, for, let’s say, a Microsoft Access database.
-
Archives
- May 2009 (1)
- April 2009 (4)
- February 2008 (1)
- November 2007 (1)
- August 2007 (1)
- July 2007 (2)
- June 2007 (2)
- May 2007 (1)
-
Categories
-
RSS
Entries RSS
Comments RSS
