Re: diffutils release imminent

[Jim Meyering]

Mike Frysinger wrote:
> On Friday 05 February 2010 05:07:37 Jim Meyering wrote:
>> For those not subscribed to the bug-diffutils mailing list,
>> you may be interested in this snapshot:
>>
>>   http://thread.gmane.org/gmane.comp.gnu.diffutils.bugs/11
>>
>> I expect to make a release next week, so any
>> patches, testing or feedback would be most welcome.
>
> the diffutils website on gnu.org says to use bug-gnu-utils ... it doesnt
> mention bug-diffutils anywhere

Thanks.  The bug-diffutils list is newer than the preceding
release (2004!), and once I make a release, part of that procedure
will mostly-automatically update the web page.

> the referenced snapshot passes `make check` on my Gentoo x86_64 system
>
> there's a long standing multibyte patch that a lot of distros have been using.
> it can be found in the Fedora src rpm for example.

If it's like the multibyte patch for coreutils,
I'll probably defer that one.

> Fedora also seems to have a trivial typo patch you'll probably want to
> integrate:
> --- diffutils-2.8.1/src/analyze.c.badc        2004-01-08 10:05:59.094598972 
> +0000
> +++ diffutils-2.8.1/src/analyze.c     2004-01-08 10:06:22.294841900 +0000
> @@ -983,7 +983,7 @@
>       changes = briefly_report (changes, cmp->file);
>        else
>       {
> -       if (changes | !no_diff_means_no_output)
> +       if (changes || !no_diff_means_no_output)
>           {
>             /* Record info for starting up output,
>                to be used if and when we have some output to print.  */

Can you point me to an example of how that bug causes
diff to misbehave?  I'd like to add a test for the fix.