[ic] to Mike Heins: mv_tmp_session?

Dan Browning interchange-users@icdevgroup.org
Fri Sep 6 19:51:03 2002


At 12:41 PM 9/6/2002 -0700, you wrote:
> >> I want to make sure I understand what mv_tmp_session will do.  If I
> >> understand you from a previous thread correctly, it will disallow id= and
> >> mv_pc= in the URL and store all session information in a cookie?  Will
>page
> >> caching be a problem at that point?  Will mv_pc=abc still be able to be
>used
> >> to record [data session source]?
> >
> >  ( NOTE: I don't like messages without context, because they usually raise
> >  questions. In this case, whether you have read and acknowledged that this
> >  is 4.9 only. )
> >
> >  I suggest you read the message I posted, which says that no session is
>written
> >  to disk at all. Therefore there is obviously no way Interchange can
>record
> >  anything about source in the way you seem to think might be possible.
> >
> >  Nothing says you couldn't set a *cookie*, though, that you could later
> >  retrieve to get a source.
>
>
>I apologize if my questions seem stupid.  I am definitely a beginner.  I'm
>trying to figure out three things:
>
>#1: Will mv_tmp_session solve my search engine indexing problems when 4.9
>comes out or should I start digging into one of the other solutions that has
>been mentioned?  It WILL solve those problems if it keeps id= and mv_pc=
>from being generated into the URL.  Is that the case?
>
>#2: Will using mv_tmp_session still allow me to set [data session source] to
>a variable, evaluate the variable, and then set [data session source] to a
>cookie?
>
>#3: Also, will the absence of mv_pc=### in the URL cause any problems as far
>as page caching?  For example: with a single user, would an item displayed
>as In Stock still be displayed as In Stock later on when it is Out of Stock?
>
>If I can figure these three things out, I will be able to make an informed
>decision about my search engine indexing dilemna.  Thank you!

I think we've beaten this horse every which way possible, Grant.  I'll give 
it a crack one more time:

  * Everything works like it does now

  * Except on spider_entrance.html, which is a new page that you've
  created
         - spider_entrance sets mv_tmp_session right off the bat
         - It then proceeds to have links to all the content that you want
           the spider to spider on your site. (Why did the fly fly?  Because
           the spider spied her.)
         - All of these links will be void of ? and &, due to mv_tmp_session

  * Note that when you submit your site to search engines, you need to
    use /spider_entrance.html as the URL.

  * Some search engines will spider your site from various locations anyway,
    which is why CFM recommended checking for robot user agents (UA) and
    sending them to the spider_entrance.html (and crossing fingers that
    they're smart enough to avoid recursively circling the site).

Anyone feel free to expound (or correct) that (note that I haven't yet 
played around with mv_tmp_session).  Cheers,

/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\
| Dan Browning, Kavod Technologies      <dan.browning@kavod.com> |
| (360) 882-7872 x7, 6700 NE 162nd Ave, Suite 210, Vancouver, WA |
\~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/
"Stealing a rhinoceros should not be attempted lightly."