I did some testing on FreeBSD 2.2.7-STABLE (i386).  Indexing seems to
work without any noticable errors.  It takes about 25 seconds to index
125 documents.  When I search for anything it runs for a long time and
then receives a SIGKILL from memchr() (according to gdb).  It must have
some sort of trouble reading the index file.  The test index works but a
more complex index doesn't search properly.  I did a backtrace and came
up with this:

#0  0x2009aaf8 in memchr ()
#1  0x7cf in ?? ()
#2  0x4979 in readFileEntry (fp=0x200bc620, line=0xefbf89f4 "",
    docProperties=0x0) at index.c:916
#3  0xbb92 in getindexfilenum (fp=0x200bc620) at merge.c:325
#4  0x6301 in search (words=0xefbfcd44 "dave and daneel",
indexlist=0x44030,
    structure=1) at search.c:222
#5  0xff0d in main (argc=0, argv=0xefbfd938) at swish.c:584

Does that make sense to anyone?

Also, I don't think that the index decoding (swish-e -D) works properly
(at least not on FreeBSD).  Maybe that's related.

I have a mirror of my web site on this machine (Linux 2.2.14).  I'll try
the same test here tomorrow.  Maybe the problem will be more obvious to
me if it works on Linux.

-- 
,David Norris
  Open Server Architecture Project - http://www.opensa.org/
  Dave's Web - http://www.webaugur.com/dave/
  ICQ Universal Internet Number - 412039
  E-Mail - dave@webaugur.com