Hi,

I am hoping someone has a clue on this. I have code that uses
CompileAssemblyFromDom to compile a DLL on the fly. Works fine here.

On a client machine however, no DLL is produced, using identical input.
There are no exceptions, no compile time errors coming back. And no file.
The code checks for the file's existance immediately afterwards, and it is
not there. Setting GenerateInMemory to false on the params object and
setting the OutputAssembly correctly.

Tried giving all sorts of permissions on the folder. Though usually this
problem would result in an exception - not fail silently.

Anyone have any ideas on why the method would execute with no errors or
exceptions, but yet not produce a file on this one machine? It's a 2003
server, if that matters. But like I said, works fine on other 2003 server
machines as well as my XP development machine.

Turns out this was related to the account not having sufficient permissions
to access the VB compiler.

This sounds like something worthy of an exception? Not just failing silently
without producing the expected result.

"Marina" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> Hi,
>
> I am hoping someone has a clue on this. I have code that uses
> CompileAssemblyFromDom to compile a DLL on the fly. Works fine here.
>
> On a client machine however, no DLL is produced, using identical input.
> There are no exceptions, no compile time errors coming back. And no file.
> The code checks for the file's existance immediately afterwards, and it is
> not there. Setting GenerateInMemory to false on the params object and
> setting the OutputAssembly correctly.
>
> Tried giving all sorts of permissions on the folder. Though usually this
> problem would result in an exception - not fail silently.
>
> Anyone have any ideas on why the method would execute with no errors or
> exceptions, but yet not produce a file on this one machine? It's a 2003
> server, if that matters. But like I said, works fine on other 2003 server
> machines as well as my XP development machine.
>

I'm a bit surprised that no exception was generated. Perhaps there's a
try catch somewhere that swallows the error up!

No, there was not. I am 100% sure. In fact, there was code immediately after
the call to this method that checked for the existance of the file (which
returned False in this scenario). The method executed completely and with no
error.

"Chris Dunaway" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> I'm a bit surprised that no exception was generated. Perhaps there's a
> try catch somewhere that swallows the error up!
>