[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Scheme-reports] command-line

On Wed, Apr 4, 2012 at 9:18 PM, Aaron W. Hsu <arcfide@x> wrote:
I know that you can get the argv[0] from the environment, but without
being tied somehow to the program execution, it is not apparent to me
whether COMMAND-LINE even makes sense.  That is, I am imagining an
embedded Scheme that runs as the scripting language of some other
application. Here one might not have explicit access to the argv
structure, but then would it even make senses to populate the
COMMAND-LINE with anything?

 Is the unreliability of /proc/self/cmdline really a problem ? AFAK, "ps" uses /proc/<pid>/cmdline to access the command line arguments of processes.
 If you need arguments, you may provide a "<your scheme>-init" function like this for instance:
 (define (<your scheme>-init . args) ...)

 which let your users to arbitrarily fill the command line passed to your underlying Scheme.
 I don't know if other people agree with this. But again, you are free not to implement the "process-context" library.
Scheme-reports mailing list