Webinator 2 and admin script

[Thunderstone]

Webinators getting better each day..
I've started to modify the default search vortex script to clean up some of
the look so I can use top.html and bottom.html a little more.
Now the one thing I really need is to limit the rewalk to the organelle walk
command. The details on the walk are a little sketchy so maybe you could
answer this.

If I use the -jURL option in the original walk will the rewalk honour the j
option? I'll need this in my next venture as a search engine. You see I
want to index only a certain category of web home pages. In other
words I want to only index portions of the web sites I'm indexing. In total I
have about 400 sets of pages to index. These pages may be on about
300-350 different websites. By the looks of things this isn't what rewalk
will do for me. If not can we expect this sort of feature?

The next question is about the admin script for the life of me I can't get it
to work. I managed to mung the Security on ip address (The script
doesn't return the correct IP address since it assumed it could truncate
(susbstr function) the IP address to 11 characters instead of the 13
characters in my IP address, IE I'm at 192.168.192.254 and it reports
192.168.192.2, Problem is it still reports that the database doesn't exist
and that it can't create the database even though it is there. I've verified
the Variables:

DB=
$DefDB=
$GwPath=

I know the def DB shouldn't have the leading / in it.
The default DB doesn't of the last slash in it either.

If I change the code in the admin script so that the DB is hard coded it
passes the exists test. However the rewalk option doesn't work. It says
it is going to do it and goes to the view logs screen but the temp directory
is never created and the walk never starts. The webinator directories
are set as the web server user with rwx priv. The gw is set as setuid.
Every thing works manually just not the web admin interface.

Looks as if we have some coding problems in the base script or the
Netscape Enterprise Server 3.0 is not returning the document root.
Any help would be greatly appreciated.

[Thunderstone]

Rewalk walks all of the urls specified over time to build the database.
It uses all of the options from the last run like when you do a -recall.
If you want see what options are in there use the following SQL:
"select Profile,Name,String from options"

So it will re-use your options unless you did several walks, each
with different options.

If it doesn't work the way you want it to, you can still do it the old
way of walking to a new database and swapping it with the live one
when you're finished.


Yes. That was example code that needed to be modified to suit your site.


If you followed the standard install procedure, the only change you should
have to make is setting <DB=> and <$GwPath=> to the correct full paths.

It sounds like you have inconsistent permissions. The database must
always be managed by the same userid. Make sure that what you did
manually agrees with what will happen, permission wise, when run
from the web server. Try starting with a database that was created
from the admin interface and never touched manually.