Forcing SYSTEMROOT (opinions needed)

Christopher Faylor
Thu May 3 10:35:00 GMT 2001

On Thu, May 03, 2001 at 11:19:26AM +0200, Corinna Vinschen wrote:
>On Wed, May 02, 2001 at 10:28:50PM -0400, Christopher Faylor wrote:
>> On Wed, May 02, 2001 at 06:55:19PM -0500, Parker, Ron wrote:
>> >>So we have to trade the possibility of someone wanting complete control
>> >>of his environment versus the possibility of someone not specifying
>> >>SYSTEMROOT but needing it for the program that is about to be run.
>> >>
>> >>Should I flip a coin?
>> >
>> >Since the Winsock dll is dynamically loaded via LoadLibrary, would it
>> >be possible to fill out SYSTEMROOT just prior to the load?
>> Hmm.  If we could be guaranteed that the program being loaded was a
>> cygwin program we could.
>> Or, maybe we don't care...  This is a creative approach to this problem.
>> I like it.
>What about adding a CYGWIN env setting "[no]pamper" with default
>setting "pamper"? We could add a function to Cygwin which is only
>called when "pamper" is set. That function could be filled with
>functionality which we _think_ are comfortable for users which
>simply want to have a functioning Cygwin under all circumstances
>and don't give a damn for purism.
>The first entry into this function could be to add always
>"SYSTEMROOT" and "SYSTEMDRIVE" to the environment.
>I'm pretty sure we would get lots of further entries over the time.

I'm not sure if you're 100% serious but this but I think that the number
of CYGWIN environment variables is already uncomfortably high.

This doesn't strike me as a CYGWIN setting.  It's something that a
programmer wants to be able to set in his own code.  If I'm calling
execl and only want four things in my environment, I should be able
to do that without being overridden by a user's environment variable

That's why I suggested some kind of API to control this behavior that
could be used by a savvy (?) programmer.


More information about the Cygwin-apps mailing list