Date:	Sun, 16 Jan 2000 14:41:43 -0500 (EST)
Message-Id:	<200001161941.OAA18690@indy.delorie.com>
From:	Eli Zaretskii <eliz AT delorie DOT com>
To:	snowball3 AT bigfoot DOT com
CC:	djgpp-workers AT delorie DOT com
In-reply-to:	<200001161614.SAA14425@is.elta.co.il> (snowball3@bigfoot.com)
Subject:	Re: Bash 2.03 prerelease now available
References:	<200001141850 DOT NAA10705 AT delorie DOT com> <200001161614 DOT SAA14425 AT is DOT elta DOT co DOT il>
Reply-To:	djgpp-workers AT delorie DOT com
Errors-To:	dj-admin AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

> My local copy of gnu/bash-2.03/config.h and my copy of bash203s.zip both 
> contain a seemingly good config.h with a size of 16,222 bytes. If the 
> version of config.h you extracted from the archive isn't that size, then 
> somehow you got a corrupted version. I have tested the archive by 
> extracting it and building the source on another machine without problems, 
> so I'm curious how you got a bad config.h.

The file is okay inside the zip archive, but the moment I run "sh build.sh",
config.h is replaced with the one that has only a single comment line.

The problem seems to be that stamp-h is not in bash203s.zip, so
Makefile tries to create it.  And here's the *real* problem: the
command that gets invoked says CONFIG_HEADERS=config.h instead of
saying CONFIG_HEADERS=config.h:config.h-in.  This is why config.h gets
nuked instead of being recreated.

I don't know why did it work for you; perhaps stamp-h was already
there, or perhaps you had config.h.in on that machine.

- Raw text -