Re: [PATCH] Fix deftypefn in fopen_unlocked.c

["Eli Zaretskii" <eliz@gnu.org>]

> Date: Sun, 15 May 2005 19:52:01 +0000 (UTC)
> From: "Joseph S. Myers" <joseph@codesourcery.com>
> cc: aj@suse.de, ian@airs.com, gcc-patches@gcc.gnu.org, 
>     gdb-patches@sourceware.org
> 
> > I verified the former before I submitted the patches.  As for the
> > latter, unless someone is going to preview the DVI output and make
> > sure it looks okay in print, "make info" can be regarded as a
> > good-enough test for "make dvi" as well.
> 
> On the contrary, many times patches have broken "make dvi" but passed 
> "make info".

I didn't say it was a perfect test, just a good-enough one.

> See the last such breakage in libiberty 
> <http://gcc.gnu.org/ml/gcc-patches/2005-03/msg02853.html> for an example.  

I didn't change any places that could cause similar problems in this
specific manual.

> "make info" and "make dvi" detect different subsets of invalid Texinfo; 

In general, yes; but in the case of libiberty, not really.

P.S.  If you have such stringent standards for accepting docs patches,
how come what I found needed so many fixes?  I found those problems by
simply looking at the index; any reasonable QA should have discovered
that long ago.