Re: Interface for SCSI transactions ?

[Samuel Thibault]

Thomas Schmitt, le Tue 13 Sep 2011 13:01:45 +0200, a écrit :
> Please ignore this sentence in my previous mail:
> 
> > Where was this assigned to mach_port_array_t resp. to init_port_set ?
> 
> It is a residue from obsoleted reasoning. I first found "^ array[]"
> in gnumach/include/mach/mach.defs but later noticed that
> device_set_status() itself has a variable length array.
> 
> So my theory question is why strangely named type mach_msg_type_number_t
> is used as counter for dev_status_t and other variable length arrays. 

Overengineering^WNice type names, I'd say.

Samuel