Forum Moderators: martinibuster
Anyone know what the heck this is? I don't want to get my hand slapped again for something I didn't do. (Kind of reminds me of the AV days when reporting SPAM to them got you banned)
I must be missing something, and heavens knows that is highly possible. My concern is not Mediapartners, it is the fact that it is saying it visited /?CFSplashed=yes. Since I don’t have FastClick and I have suspended all my adwords, about 2 weeks ago, I didn’t see any relevance to the link. What am I missing? I don't have a splash page. Is CF standard for something?
google.com/search?q=%3FCFSplashed%3Dyes&hl=en&lr=&ie=UTF-8&filter=0
Why on earth would CFSplashed=yes be appended to a filename?
I've also noticed that most of the sites that I found indexed with CFSplashed=yes in the query string are government sites.
Well I have back links from NASA
but I doubt they are running adwords or adsense.
See my earlier message in the thread about Opera and post the User Agents for any and all calls to those pages with those query strings, with the exception of those from Mediapartners. Keep in mind the lag between a visit to a page and Mediapartners coming by can be anywhere from a few minutes to many hours so make sure you grep your logs far enough back.
Possibly someone using Opera. Versions of Opera that display AdSense ads within the Opera browser will cause Mediapartners to visit any URL entered, regardless of whether the URL exists.
Even images?
User visits page with the version of Opera browser that shows AdSense ads [opera.com] within the browser. Instead of entering the URL expected by the webmaster of the site the user clicks a link or manually enters a URL with the querystring "?CFSplashed=yes" appended.
If that exact URL has never been visited by the Mediapartners bot (or sometimes if it hasn't visited it recently), it triggers it to visit at some point in the future, between a few minutes and many hours later.
The Mediapartners bot visits and is logged by the web server.
What I am suggesting is digging back farther in the web server logs for all requests for that exact URL to see whether an Opera browser user made the initial visit.
I probably should have said earlier that I've noticed the same thing on one of my sites (page with invalid querystring visited by Mediapartners and Opera was why, though I don't know *why* the user ever manually appended an odd querystring or if it wasn't manual what caused it).
[edited by: richmondsteve at 7:58 pm (utc) on Dec. 3, 2003]
I mean they are so important now that the can change standards like robots.txt. It should be UA the a list of dissallows, but with mediabot, it is 180 degrees out of phase. That alone made me lose some respect for them. They should have said create a google.txt file with x, y, and z commands. Those alone sums up something.
think that has nothing to do with what the original poster brought up.
The same could be said for all of your posts, richmondsteve.
Yea but this was Mediapartners-Google/2.1 not Opera or IE and the IP was G's. So someone would have to do a lot of spoofing to pull a joke on me? Why waste their time. It really would not be that funny.
I doubt anyone would do that, unless they were ridiculously bored, and they decided to do that to other people 4 months prior.
You might just be worrying yourself over nothing, but with Google, you just never know anymore.
Well, not anymore anyway.
The same could be said for all of your posts, richmondsteve.
panic, I wasn't criticizing you - I was just stating that even though files with image file extensions like .jpg could trigger Mediapartners I believe it would have to be the result of a .jpg file really being something other than an image, calling the AdSense JS code, which I assumed wasn't the case since you probably would have mentioned that if it was. It's possible something else could cause it, which is why I said I wasn't sure.
The scenario I described in message #14 has been confirmed to result in Mediapartners visits even if the page has never had AdSense on it. I've seen it, so have others.
Relevant WebmasterWorld thread on AdSense and Opera and the behavior I was describing [webmasterworld.com]
I realize that jim_w doesn't think it was triggered by an Opera user b/c the UA he saw reads Mediapartners-Google/2.1 (because it probably was Mediapartners), but the key is whether another machine accessed the page at some point prior to that and whether it was an Opera user (though if the HTTP_REFERER was set to something else by the user it would be hard to know for sure whether it was Opera).
jim_w, why not humor me and check your logs for *any* visits to that exact URL, with the same query string at some point in the past. I've seen the same behavior happen on one of my sites with 2 different causes - 1. an Opera user (that part I confirmed) who visited a page on my site that never had AdSense enabled and 2. someone who manually added in a stupid query string (best guess since there was no HTTP_REFERER) on a page that *did* have AdSense.
There is no other reference to CFSplashed in my logs going back 4 months.
Here is the actual log entry.
crawler8.googlebot.com - - [02/Dec/2003:14:32:00 -0500] "GET /?CFSplashed=yes HTTP/1.0" 200 29759 "-" "Mediapartners-Google/2.1 (+http://www.googlebot.com/bot.html)"
Ok the number of bytes taken, 29759, is about the size of my index.html w/o graphics. 200 says it found CFSplashed=yes, but it could not have.
Beats me.
The query strings aren't the same as yours, but what you found and what I found could be related.
All were from 64.68.87.66 and 64.68.87.69, UA of Mediapartners-Google/2.1 (+http://www.googlebot.com/bot.html) and query strings like:
?cf0B1E0561=Q0NJQ1xUSmFja3NvbjpyZWEsbW50bG06FC
hMxrQkXgsnCET5sgWo7w==
?cf0A468C36=QkFZNVIxXGV4bmrpOmJheDVyMTq3QDMRMuZeLxHm9+NJ3Hin
I'm stumped. I'd love to know why, but it's very infrequent and it doesn't appear to be hurting anything. Odd.
this thread made me register, try base64_decode the code!
it tells that a user named => Jackson
Using a => HTTP relam
based on => NTLM
and some more....
btw)
i 99% think it is Froogle, and not google!
try:
[froogle.google.com...]
please let me know...
