Re: CVS, bootstrapping and DJGPP

[Richard Dawe <rich at phekda dot freeserve dot co dot uk>]

Hello.

Alexandre Duret-Lutz wrote:
> 
> >>> "Richard" == Richard Dawe <rich@phekda.freeserve.co.uk> writes:
> 
> [...]
> 
>  Richard> I found the problem. It was due to $SHELL (in the
>  Richard> environment) not being set to bash.
> 
> Is this something we should set in Automake?
> 
> For instance
>  $ENV{'SHELL'} = '@SHELL@'
> where @SHELL@ is the shell computed by ./configure

Yes, but I don't know whether you should do this for all platforms or just
DJGPP (by checking for defined($ENV{'DJGPP'}) before defining $ENV{'SHELL'} to
be @SHELL@).

> Or is this something DJGPP users should stick in their
> environment?

That would be the easy solution. But this hasn't been adopted. This problem
has come up a number of times before.

$SHELL hasn't been set to bash for DJGPP for the following reasons:

* It would require bash to be present. What happens if it isn't? It breaks all
system() calls.

* bash is quite a large program. DJGPP doesn't have shared libraries => big
executables. DJGPP may be used on systems where disk space and memory is
tight.

* The code for system() handles most cases without requiring bash to be
present.

> The issue doesn't seem limited to automake.  For instance
> autoupdate will calls autoconf using arguments like
> "--trace define:'AC:\$f:\$1'".
> 
> [...]

Yes, I caught cases like that in the DJGPP port of autoconf. It was only with
automake that I didn't.

Thanks, bye, Rich =]

-- 
Richard Dawe [ http://www.phekda.freeserve.co.uk/richdawe/ ]