[ic] No session id wanted, even if no cookies

Bryan D Gmyrek gmyrek at U.Arizona.EDU
Fri Jul 30 15:45:15 EDT 2004


I've seen this discussed before but looking through the docs
can't figure out how to implement it.  The idea is that since
over 99% of people have cookies turned on then why not just
require cookies for everyone to avoid the ?id= stuff in URLs.
The major advantages are that the URLs look nicer, will show up
with the correct PageRank in the google toolbar and of course the
search engines will never see the ?id=.   Now I know your're
thinking that the RobotUA solves this.  While it's a beautiful
feature it doesn't totally solve this problem.  For one thing
sometimes people copy a url from the browser window and place the
link to your site with the ?id= in the url.  The search engine
will now index 2 identical pages, the one with and the one
without (and possibly more with different ?id= in them).
This sort of thing can get you into trouble with the search
Also, sometimes spiders will check for cloaking by spidering
their standard useragent (or at least I'm paranoid enough to
think so) ... in this case they might get the ?id= stuff.  Doing
a search on Google with
site:www.mysite.com -asdlfkj
I can see that googlebot does indeed pick up some of my pages
with the ?id=, even though I use RobotUA and RobotIP.
So is there some way to set a variable in catalog.cfg or
interchange.cfg to turn uff ?id= (and mv_pc= while we're at
it..)?  Tried:
ScratchDefault  mv_no_session_id  1
ScratchDefault  mv_no_count       1
ScratchDefault  mv_tmp_session    1
in catalog.cfg
but that seems to have no effect.


If you've read this far here's a tidbit in case you haven't heard
Make sure to set up a 301 redirect from your non-www domain to
your www domain.  I.e. in your .hatccess file in your web
directory put:

Redirect 301 http://mysite.com/ http://www.mysite.com/

This will help you avoid being the victim of a google bug called
the SLOW DEATH whereby all of your pages slowly disappear for 'no
apparent reason'.

More information about the interchange-users mailing list