Currently I have 2 exe files. app.exe and updater.exe. When app.exe finds out that there is new version available it runs updater.exe, which downloads and replaces it.
I'm wondering if it's possible to build updater.exe into app.exe. On app start it should check if there is updater.exe in the directory and if it's not than extract it. Any help appreciated.
You can integrate updater.exe as a resource into app.exe. The following SO post shows how to extract it at run-time:
Embedding an external executable inside a C# program
Note that, if your application is located in the default application directory (C:\Program Files, or, more generally, %ProgramFiles%), you will not have permissions to create a file in the same directory (which is a good thing). Thus, you might need to extract updater.exe into a temporary directory that the current user has write permissions to (such as Path.GetTempPath).
When I'm understanding you correctly you need to pack you updater.exe into you app.exe as ressource. Extract it, when App.exe starts.
But I can tell you, this idea with replace is bad. You should choose a MSI package to deploy your application.
For traversing and listing directory contents, see http://msdn.microsoft.com/en-us/library/system.io.directory.aspx
For accessing your embedded resources, see http://msdn.microsoft.com/en-us/library/xc4235zt.aspx
For writing binary data to disk, see http://msdn.microsoft.com/en-us/library/system.io.binarywriter.aspx
For running an exe from disk, see http://msdn.microsoft.com/en-us/library/system.diagnostics.process.aspx
Related
I created setup file using Inno setup.I have exe,dll and one xml file in my setup.
When I install on 64 bit machine it works fine means it take xml file from directory where exe is present.
But When I install same setup on 32 bit machine it take dll path but while accessing xml file it takes path of desktop where shortcut of exe is present and showing FileNotFoundException.
Thanks
Your application is most likely not specifying pathnames on the files it is trying to open, so it is expecting to find them in the current directory. Inno Setup by default does not set the "Start In" field on shortcuts its creates; this causes Windows to pick a directory itself, which usually won't be the directory containing your application.
In virtually all cases, this is something that should be corrected at the application level. Properly designed GUI applications should not expect to be started from a particular directory; they should always specify full pathnames on files they open. In Delphi or C++Builder, for example, it's possible to get the full pathname of the directory containing the application EXE by calling: ExtractFilePath(ParamStr(0)). To get the full path of a file named "File.txt" in the application directory, use: ExtractFilePath(ParamStr(0)) + 'File.txt'.
To get path of working directory of application while loading xml file in code.
string WorkingDir=System.AppDomain.CurrentDomain.BaseDirectory
XDocument temp_xdocument= XDocument.Load(WorkingDir+"file.xml");
It works for me.
I'm planning to build my winform into a .exe file. I'm just wondering what to do with the XML files that my application needs?
I did some research and found out that I can add the XML files in the Resource folder before creating a .exe file.
Or I need to create a setup file? When the user runs the setup file, the XML files will be installed into their pc.Now I wonder which one is the best way to go for,
Note: XML files might get modified by the user.
If you want to ship the XML files as seperate to the .EXE then you can set the Copy to Output Directory to Copy if newer. (click on file and then go to properties).
OR if you want it as part of the .EXE I think you can change the Build Action to Embedded Resource.
I personally would create a Setup as per your edit and include the XML files. I usually just add everthing from the bin/release folder that is needed when I create a setup file.
You could either deploy the necessary files along with the executable in the same folder or embed them as resources (if they are read-only). If you need to modify them do not embed them as resources into the executable.
The correct way depends on how you intend to use the files. If the files always are deployed together with your application, the application never writes to them and they are never upgraded without upgrading the application, you can go with them embedded as resources.
If you need to update them separately from the application, you need to have them as physical files.
You don't necessarely need a installation package, unless you need to apply some logic during setup, such as updating the content of the setup based on user input or to check preconditions. The application can just be copied into place regardless of if you have embedded the files or not.
I'm developing a Web application that uses a couple XML files to store data. I have their Build Action set to Content, and on install the files are copied successfully to the Applications Virtual Directory:
C:\inetpub\wwwroot\ApplicationName\
The problem I'm having is that writing to these XML files (in order to save settings and things like that) causes a lot of write permissions issues. Therefore to get around it, I'm trying to copy these files from the virtual directory they're installed to to a new directory under the C drive, using the following PostBuildEvent in the Web Deployment Project:
xcopy "$(TargetDir)*.xml" "C:\CompanyName\ApplicationName\" /y
However, this does nothing. I'm not sure if this is because PostBuildEvents in the installer are not actually fired on install, but only on building the installer, or if TargetDir represents the bin directory:
C:\inetpub\wwwroot\ApplicationName\bin\
instead of the root application virtual directory:
C:\inetpub\wwwroot\ApplicationName\
Does anyone have any ideas? Has anyone dealt with this sort of thing before? I'm really stumped on this one.
Update:
I included a PostBuildEvent that should give full permissions to all users:
icacls "$(TargetDir)" /grant Users:F
But it doesn't seem to have resolved the problem.
I'm also unsure where exactly $(TargetDir) if pointing to, if it would be ..\ApplicationName\ or ..\ApplicationName\bin\
All you have to do is give full permissions to the user under whose context the Web Application Pool runs.
You can find this user by starting IIS manager and look at the Application Pools and Identity column
Something like
cacls C:\inetpub\wwwroot\ApplicationName /G Users:F
will give all permissions to this subdirectory to all users on the computer, If you are running under ApplicationPoolIndentity refer here
Turns out there's a much easier way to do what I was trying to do.
Instead of using PostBuildEvents to create a new directory structure and move files there, I added the new directory structure into
InstallerProject > View > File System
after which I located the Content Files from SomeProject entry in my installer project, and changed the Folder value to the newly specified directory structure, in this case
C:\CompanyName\ApplicationName\
This seems to work fine, I'm now able to access these files freely as I originally intended.
Assuming i have an executable called mainprogram.exe and another called Update.exe. What I want to acheive is: when Update.exe is started, it replaces mainprogram.exe with, for example, C:\Program Files\MyProgram\mainprogram.exe. The point is to include the executable that needs to be replaced in Update.exe so it would be an all-in-one solution.
Also note that I can't create Update.exe that will download the mainprogram.exe from internet, it has to be already included in Update.exe.
In building:
copy UPDATE.EXE to be embedded into the MAIN.EXE as a resource
At runtime (update time)
drop UPDATE.EXE from resource to the hard drive
start UPDATE.EXE with appropriate parameters (command line arguments, depending on the situation)
have UPDATE.EXE try to delete common file that MAIN.EXE will create and leave it opened
when MAIN.EXE exits, handle to the file will be released, allowing UPDATE.EXE to continue doing what it should - replace the MAIN.EXE with new version
start MAIN.EXE from UPDATE.EXE
(optional) delete UPDATE.EXE from MAIN.EXE, as you don't need it for now.
You can use .NET resources and access them via Properties.Resources. If you want to copy the content of a resource to the hard disk you can use System.IO.File.WriteAllBytes.
Adding and Editing Resources (Visual C#)
What you want is called a "Setup Project". This will allow you to bundle any files necessary for such an update to be deployed in such a manner.
Here is the MSDN link for doing this.
In a nutshell, you are creating an installer package, but you can limit the functionality of the installer to replace the designated files only.
You could just use an SFX archive - works pretty much like a setup program and "self extracts". Many archiving programs allow it, or you could just write a setup project in .NET which will do the same thing if you're having to include all the files in the updater anyway.
I have a simple C# console application developed on my local machine using VS2008 Pro. I want to know how to deploy this solution onto a network share folder?
A similar Java console program is already placed (as a JAR file) in the same network share folder. Users simply open command prompt, navigate to shared folder and type "java -jar programName.jar inputParameter1 inputParameter2"
How can I achieve the same with .NET?
You can copy the exe over yourself, go to the bin folder in the directory your source code is in and copy it there.
or you can click the BUILD menu and use the PUBLISH menu item. This will allow you to enter the path to your network share and visual studio will copy the built app to the folder for you.
If your application is really "simple", you should be able to just copy the files to a shared folder and run it from there. However, if your "simple" application tries to do things that are restricted by the permissions you might have to configure them with caspol. Assemblies loaded from a shared drive have much fewer permissions than the ones loaded from a local drive.
It would be mostly the same process as the Java program. To deploy, compile the program and copy the exe from the bin folder (along with any dependencies) to the network share.
To run the program users would open the command prompt, navigate to shared folder, and type "programName.exe inputParameter1 inputParameter2"
You can use Publish feature of VS. Note that you can change settings in the Publish section of the console application project to remove some features that you don't need. For instance the renaming of .dll and .exe files by appending the '.deploy' extension to the name of the files or publishing in a new 'version' folder each time. Go to "Project Properties"->"Publish" and remove "Automatically increment revision" checkbox at "Publish Version", click "Options..." button and clear all checkboxes there too.
Right click your project, select publish which will make an executable, you can put that in your shared drive, similarly users can go into the command prompt and run it and give some args.
In the exact same way assuming they have the proper dependencies installed (.net, 3rd party assemblies, etc). copy the bin folder then have them execute the exe file.
Take a look at ClickOnce deployment:
ClickOnce is a Microsoft technology
for deploying Windows Forms or Windows
Presentation Foundation-based
software, also called Smart clients.
It is similar to Java Web Start for
the Java Platform.
MSDN
Wikipedia