if you have access to the filesystem where the files are stored, is there some 
advantage to using the spider at all?

otherwise you could do:

   swish-e -i /path/to/site1 -c config1

and

   swish-e -i /path/to/site1/site2 -c config2

which would be both faster and create two different indexes for searching.

fh oregon scribbled on 4/10/05 6:02 PM:
> My goal here is to have the main site and the virtual site(s) indexed 
> and searchable.  The more I mull this over I came up with a way to fake 
> out the indexer.   As a test, I placed a (hidden) link on the main page 
> directly to the /SFCC directory and !!!  It looks like it is all working 
> now.  I need to do more testing.
> 
> -fh
> 
> Bill Moseley wrote:
> 
> 
>>On Sat, Apr 09, 2005 at 11:10:12AM -0700, fh oregon wrote:
>> 
>>
>>
>>>The root of the site (frankhunt.com) is /web/httpd/htdocs  Within that 
>>>directory is the main index.html as well as a few other html documents 
>>>and directorys for other parts of the site.  One of those directorys is 
>>>/web/httpd/htdocs/SFCC which is the root of the 
>>>siliconforestcorvetteclub.com domain.
>>>   
>>>
>>
>>Again, the spider has NO knowledge of your directory structure.  If
>>you spider frankhunt.com and there's no pages in frankhunt.com in CFCC
>>then it won't spider them.
>>
>>Try it yourself.  Go to frankhunt.com and only click on links that
>>include frankhunt.com as the host name.  That's all that will be
>>indexed.  That link to CFCC is not the same host name.
>>
>>Look, you also link to http://www.fs.fed.us/gpnf/volcanocams/msh/ --
>>do you expect that to get indexed?  And everything it links to, also?
>>
>>Sounds like you are not clear on how web servers map directories.
>>
>> 
>>
> 
> 

-- 
Peter Karman  .  http://peknet.com/  .  peter(at)not-real.peknet.com