Hello Simon,
I think that autoconf or imake or configure could in principle make life
easier, if properly setup and maintained. That is a bi if. As it stands
out Makefile requirements are rather simple, and the platforms supported
are rather limited, so neither begs for autoconf.
The emails today about Makefile issues seemed to me to be about a broken
compiler setup on campus, where g77 can't find the appropriate f2c (or
g2c or whatever.) This is a broken compiler setup that would not be
fixed by any fancy scheme. Tong implemented a kludge until whomever on
campus fixes g77.
The only reluctance I have with these fancy make schemes is that it
seems non trivial to get them to work. I for one don't know how to
correctly write an Makefile.in or configure script. And ours would need
special since we need CERNLIBS and ROOT. If you know how to do this, set
it up, and show us how it is done so we can make changes, I think you
have a very strong chance to convince the collaboration to switch to
autoconf.
Cheers,
Maurik
On Fri, 2001-10-05 at 15:04, Simon Sirca wrote:
> > in the weekly software meeting, we talked about having Makefiles only
> > depend on the standard environment variables for paths, LD_LIBRARY_PATH, etc.
>
> I am sorry I could not participate in today's meeting since I was at Bates.
> But based on the e-mails sent after the meeting, let me notice once more
> that I regret that the use of autoconf, which would be a perfect tool
> to match each person's computing environment to the BLAST software
> distribution, failed to become an option. It would solve the trivial
> problems concerning paths, etc. in an unambiguous and elegant way.
> In particular, it is much better to insert the shared-library search path
> into the executable itself using -Xlinker -rpath options during linking,
> instead of defining it through a non-gauge-invariant LD_LIBRARY_PATH.
>
> Simon
>
> --
> Simon Sirca
> MIT-LNS, Room 26-402 Tel: +1 617 258-5438
> 77 Massachusetts Avenue Fax: +1 617 452-5950
> Cambridge, MA 02139-4307, USA URL: http://pierre.mit.edu/~sirca
>
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST