Looks like the interesting things to demand are: UIPermission, RegistryPermission, ReflectionPermission, FileDialogPermission(?), FileIOPermission and EnvironmentPermission.
With FileIO stuff we will probably need to demand access to parts of the filesystem in code as we don't know ahead of time what we need access to, and probably shouldn't demand all access. Thus we will not be able to use declarative security with FileIOPermission stuff with a demand or assert. Comments are welcome. I would love to hear comments and suggestions from anyone who has built projects that use .net security stuff. ----- Original Message ----- From: "Scott Hernandez" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, September 19, 2003 2:43 PM Subject: declarative security > I'd like to start littering our code with declarative security statements. > It would be good to start to lock down what nant needs from a security point > of view so we can easy make a statement like our code is secure. > > Does anyone have much experience with .net declarative security and the best > approach to the goal of making a statement like "our app is secure". Or at > least be able to identify the unsafe (with respect to security) sections of > code that need special permissions. I would like to see 3rd party tasks > follow some security guidelines as well. > > We run pretty liberal right now with access to reflection and other internal > access in the framework I would guess. > ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ nant-developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/nant-developers
