On Thu, May 19, 2005 at 09:55:40PM -0700, Ben Pfaff wrote:
John Darrington <[EMAIL PROTECTED]> writes:
> So the question is: Should we
>
> a) emulate SPSS's ability to read this faulty system file (for the
> sake of 'compatibility')?
>
> b) prefer 'correctness' and leave these guys to fix their own buggy
> (and proprietary) product?
Let's make it work. What advantage is there in being
incompatible, except for some putative "moral high ground"?
Software should be liberal in what it accepts, strict in what it
produces.
Done. I've changed sfm-read.c so that it accepts these broken files.Also, in line with the "liberal in what it accepts" philosopy, I've changed a few of the error messages regarding unexpected records to be warnings. It's possible that this could mean a corrupt system file could crash pspp, but we can cross that bridge when we come to it. J' -- PGP Public key ID: 1024D/2DE827B3 fingerprint = 8797 A26D 0854 2EAB 0285 A290 8A67 719C 2DE8 27B3 See http://pgp.mit.edu or any PGP keyserver for public key.
pgpzZunPG6gRJ.pgp
Description: PGP signature
_______________________________________________ pspp-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/pspp-dev
