Re: cygwin breakage

[Peter O'Gorman]

Hi Chuck,

Charles Wilson wrote:

> The combination of Peter's recent two patches:
>
> 45-peter-eval_show.patch
> 46-peter-revert_eval_show.patch
>
> Which, in the end, results simply in
>
> --- ltmain.m4sh.1       2004-10-15 10:25:26.140625000 -0400
> +++ ltmain.m4sh 2004-10-15 10:25:59.390625000 -0400
> @@ -5135,7 +5135,8 @@

> -         $show "$cmd"
> +         func_quote_for_eval "$cmd"
> +         eval \$show \"$func_quote_for_eval_result\"

> manages to break these regression test on cygwin

> +FAIL: mdemo-make.test

> +FAIL: mdemo-dryrun.test

> +FAIL: mdemo-make.test

> Peter, what EXACTLY were you trying to fix?  Was something actually 
> broken, or did the previous WORKING code offend your sense of elegance? 
> (No, I'm not real happy).

This successfully fixed the (purely aesthetic) bug recently reported by the 
sleepycat folks on Mac OS X :(.
Sorry that you are unhappy, I will revert the change after work tonight, but 
I am curious as to how exactly it is failing. Could you send the relevant 
bits of VERBOSE output?

> This patch DOES harm.  Please revert or fix it.
Yes, yes. Okay.

Peter
--
Peter O'Gorman - http://www.pogma.com