🚀 go-pugleaf

I tried to make a new special target, with different behaviour for
long and short fingering.  I then discovered that I always got the
short answer, independent of the flags given.  I traced this down to
in.fingerd.c where a check is made to look for special target scripts
with an s- or x- prefix.  This attempt is made irrespectively of the
value of the is_long_finger flag, and before the special code for long
fingers are executed.  I solved the problem by moving the test later
in the code, but I'm not sure if may I have introduced some other
problem with this fix.  Anyway, this is how I made it:

--- in.fingerd.c.org	Tue Oct 27 04:14:19 1992
+++ in.fingerd.c	Thu Nov 12 16:34:33 1992
@@ -371,10 +371,6 @@
       return;
     }
 
-  if (maybe_special_target (user, stream, 's')
-      || maybe_special_target (user, stream, 'x'))
-    return;
-
   /* Is this a local finger?  If so, just call the local client finger
      daemon to get packets, and print them out. */
   if (xstricmp (user, ".local") == 0)
@@ -453,6 +449,10 @@
 
       return;
     }
+
+  if (maybe_special_target (user, stream, 's')
+      || maybe_special_target (user, stream, 'x'))
+    return;
 
   /* If the user wants to know about free machines then look the information
      up in the database kept on the availability of hosts.  Print statistics