From: ahoward <ahoward@...>
Date: 2003-05-10T02:00:21+09:00
Subject: Re: extern "C" of prep_stdio

On Sat, 10 May 2003, ts wrote:

>
> [[ spamassassin definetely think that you are a spammer :-)) ]]
>
> >>>>> "S" == Simon Strandgaard <0bz63fz3m1qt3001@sneakemail.com> writes:
>
> S> 	VALUE w = rb_class_new_instance(2, params, klass);
>
> S> 	rb_defout = rb_stdout = rb_stderr = w;   // Is this correct ???
>
>  no, you must call IO#reopen

in several of the recent threads on io redirection i have been admonished for
suggesting the use of reopen and told to use assignment

  $defout = $stdout = $stderr = redirected_io_object

instead.

now, i personally have had much better luck using #reopen, but i'm curious as
to your rational.  since using reopen can cause a broken pipe to
feeder/feeding processes, and it makes it harder to restore io objects to
original condition (though not by much).

am i the only one out here who thinks an RCR for a suite of common redirect
methods seems reasonable?  there are few topics which have caused so much
confusion/discussion on this list - well, maybe 'class << self'... ;-)

-a

--
  ====================================
  | Ara Howard
  | NOAA Forecast Systems Laboratory
  | Information and Technology Services
  | Data Systems Group
  | R/FST 325 Broadway
  | Boulder, CO 80305-3328
  | Email: ara.t.howard@fsl.noaa.gov
  | Phone:  303-497-7238
  | Fax:    303-497-7259
  ====================================