diff --git a/en/gallery/gengallery.pl b/en/gallery/gengallery.pl
index 68f226b4ce..c5a38a78dd 100755
--- a/en/gallery/gengallery.pl
+++ b/en/gallery/gengallery.pl
@@ -1,70 +1,82 @@
 #!/usr/local/bin/perl
 #
 # gengallery.pl - script to convert tab-delimited gallery db file to
 # a section of an SGML list (not a full SGML file!) for inclusion
 # into another SGML file where the <UL></UL> list element pair is
 # already in existence.
 #
-# Syntax: dump.pl type < galleryfile.db > galleryfile.inc
+# Syntax: gengallery.pl type < galleryfile.db > galleryfile.inc
 # where type is one of: commerical, nonprofit, personal
 #
 # yymmdd own comments
 # ------ --- ------------------------------------------------
 # 980311 nsj First pass
 # 980312 jrf Added sorting
 # 980313 nsj Wrapped file input routine with error checking
+# 981229 nsj Tried to be more intelligent in eliminating any
+#	     possible blank entries in the .db file and 
+#	     malformed url entries (missing http://).
 
 # Setup
 # Which sort program are we using?
 $sort	=	"/usr/bin/sort";
 
 # What gallery are we processing?
 $type	=	$ARGV[0];
 if ($type =~ m/commercial/i) {
 	$type = "commercial";
 } elsif ($type =~ m/nonprofit/i) {
 	$type = "nonprofit";
 } elsif ($type =~ m/personal/i) {
 	$type = "personal";
 } else {
 	die "I don't understand type $type";
 }
 
 # Open a pipe from a unix sort of the db file (case insensitive)
 if (-f $ARGV[1])
 {
 	$infile = $ARGV[1];
 } else {
 	die "File $ARGV[1] is unreadable or does not exist";
 }
 
 # Open the (sorted) file
 open(DBFILE, "$sort -f $infile |");
 
 # Iterate through each line, throwing out those that don't
 # go in this gallery.  Output each entry as list elements.
 while (<DBFILE>)
 {
 	# We only want entries of type $type; throw the rest out.
 	next unless m/^$type/;
 
 	# Translate out characters special to SGML (and HTML)
 	s/&/&amp;/g;
 	s/</&lt;/g;
 
 	# Split the db line into its component parts.
 	($dummy, $name, $url, $description, $email, $dateadd, $datever) = 
 	m/([^\t]+)\t([^\t]+)\t([^\t]+)\t([^\t]*)\t([^\t]+)\t([^\t]+)\t([^\t]+)/;
 
+	# Skip bogus entries (in lieu of actually checking to
+	# eliminate them from the .db file.)
+	next if ($name =~ m/^$|^\s+$/);
+	next if ($url =~ m/^$|^\s+$|^http:\/\/\s+$/);
+
+	# Be semi-intelligent about entries without "http://" in front
+	# by prepending that to any that are missing it.
+	$url = "http://" . $url unless ($url =~ m/^http:\/\/.*$/);
+
 	# Dump it out to the file, in SGML <LI> format
 	if ($description ne "")
 	{
 		print "<LI><A HREF=\"$url\"><STRONG>$name</STRONG></A> -- $description</LI>\n";
 	} else {
 		print "<LI><A HREF=\"$url\"><STRONG>$name</STRONG></A></LI>\n";
 	};
 
 };
 
 # Close the pipe like good little daemons.
 close(DBFILE);