Re: pkg/47054 (benchmarks/super_pi: source unfetchable and restricted)

[Robert Elz <kre%munnari.OZ.AU@localhost>]

The following reply was made to PR pkg/47054; it has been noted by GNATS.

From: Robert Elz <kre%munnari.OZ.AU@localhost>
To: John Marino <netbsd%marino.st@localhost>
Cc: gnats-bugs%NetBSD.org@localhost, pkgsrc-bugs%NetBSD.org@localhost
Subject: Re: pkg/47054 (benchmarks/super_pi: source unfetchable and restricted)
Date: Sat, 13 Oct 2012 18:28:39 +0700

     Date:        Sat, 13 Oct 2012 09:41:50 +0200
     From:        John Marino <netbsd%marino.st@localhost>
     Message-ID:  <50791B3E.9040900%marino.st@localhost>
 
   | The client here (I think) is "fetch" which is what pkgsrc dictates.
 
 Actually, why is that?   fetch is certainly an option, but "dictates"
 seems beyond what pkgsrc actually specifies, in fact what I see in
 pkgsrc/mk/defaults/mk.conf is ...
 
 defaults/mk.conf:FETCH_USING?=  ftp
 
 which I think is selecting ftp as the default command.   pkgsrc definitely
 supports using fetch (net/fetch) as an alternitive, but "dictates" might
 be a little strong.   Do you have it set in mk.conf or something, or are
 you not using NetBSD perhaps?   (Just trying to understand why).
 
 I noticed this when I was checking just how fetch works, and discovered
 that I didn't have it installed at all (no command called "fetch" - I do
 have a bash alias I call fetch that just runs ftp for me with appropriate
 args...)
 
 Perhaps this is why you're seeing different results than most of the rest
 of us, we're probably all using ftp, whereas you are apparently using
 fetch instead (my reading of the ftp code in libfetch, which fetch uses,
 suggests that it is very pedantic about what it requires of the FTP server,
 the ftp command is far more forgiving).
 
 kre