Re: config.status[12]: export: SHELL##: is not an identifier

Bob Friesenhahn <[email protected]> Sun, 4 Aug 2024 17:33:14 -0500
Newsgroups gmane.comp.sysutils.autoconf.general
Message-ID <[email protected]>
On 8/4/24 16:39, Paul Eggert wrote:

>
> ... and the "export SHELL" line ends at byte offset 4096, then I'd 
> guess it's a buffering bug in the shell. If the shell is actually 
> ksh93 I wouldn't be surprised.

I am going to assume that you are correct about the the byte offset.  I 
had vague recollection of this being mentioned before on the mailing 
list, but did not encounter it until now.

The very benign seeming changeset at 
https://foss.heptapod.net/graphicsmagick/graphicsmagick/-/commit/7ff15be04d174ccdeea0004ffa4a45dd4a9e81aa 
(see only configure.ac) triggered it.

Running the configure script directly using bash or exporting 
'CONFIG_SHELL=/usr/bin/bash' allows configure to succeed.

I will take this as the opportunity to simply remove the archaic feature 
(Display Postscript) I re-enabled with the changeset since I have not 
encountered a system with it for a great many years.

> That also would explain why an older GraphicsMagick doesn't see the 
> problem: the line in question happens to not appear exactly at a 
> buffer boundary.
Bad luck!
> Of course I'm just guessing.

Good guess. Thank you for your help.

Hopefully a future stable Autoconf will detect the broken shell and set 
CONFIG_SHELL=/usr/bin/bash in order to succeed.

Bob

-- 
Bob Friesenhahn
[email protected], http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/
Public Key,     http://www.simplesystems.org/users/bfriesen/public-key.txt