Re: pkg/49782 (textproc/mdocml won't build on Solaris)

[Brad Forschinger <bnjf%bnjf.id.au@localhost>]

The following reply was made to PR pkg/49782; it has been noted by GNATS.

From: Brad Forschinger <bnjf%bnjf.id.au@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: solaris-pkg-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost, 
	pkgsrc-bugs%netbsd.org@localhost, bnjf+netbsd%bnjf.id.au@localhost
Subject: Re: pkg/49782 (textproc/mdocml won't build on Solaris)
Date: Mon, 7 Sep 2015 12:47:29 +1000

 I keep getting pestered by GNATS, so I thought I'd better reply.
 
 The patch still hasn't been included, so it still doesn't build.
 
 On Fri, May 29, 2015 at 6:40 PM, David Holland
 <dholland-pbugs%netbsd.org@localhost> wrote:
 > The following reply was made to PR pkg/49782; it has been noted by GNATS.
 >
 > From: David Holland <dholland-pbugs%netbsd.org@localhost>
 > To: gnats-bugs%NetBSD.org@localhost
 > Cc:
 > Subject: Re: pkg/49782 (textproc/mdocml won't build on Solaris)
 > Date: Fri, 29 May 2015 08:36:14 +0000
 >
 >  On Fri, May 29, 2015 at 12:05:01AM +0000, Christos Zoulas wrote:
 >   >  |  > I believe that __foo__ symbols are the new and preferred way, where
 >   >  |  > __foo are historical...
 >   >  |
 >   >  |  GCC defines both, but I believe Studio only defines __sun.
 >   >  |
 >   >  |  Just use __sun, as we already do in the vast majority of patches.
 >   >
 >   >  Doing both is safer... There are many symbols where __foo__ is
 >   >  defined and __foo is not. I can think of one important example
 >   >  (__STDC__). Having a unified rule for all of them is less thinking
 >   >  when you need to decide what to code.
 >
 >  That's why pkglint is there to keep track of it for us...
 >
 >  --
 >  David A. Holland
 >  dholland%netbsd.org@localhost
 >