Google has reintroduced their Google Desktop with a vengeance. It was evil enough the first time around, but this time it’s downright scary. My original complaint was that Google Desktop ignores basic practices amongst RSS readers for saving bandwidth on the sites it is polling. It was pinging my site every 5 minutes asking for updates without caching the results and thus was using an unreasonable proportion of my bandwidth.

Since a new version was recently released,  I decided to try it out to see if the issue had been fixed since I sent them mail. I installed Fiddler to monitor the traffic of the application and what I found out surprised me a great deal. Google Desktop not only pings sites every 5 minutes in a manner inconsiderate of their bandwidth but it also does so without the users direction. Below is a screenshot of some of the HTTP traffic generated by Google Desktop

The highlighted requests are requests to URLs of Atom & RSS feeds that were in my browser cache by Google Desktop. I did not configure the application to fetch these feeds. So not only does Google Desktop flood websites with feed requests in a manner bordering on the behavior of a malicious application, it also does this automatically without the end user explicitly subscribing to the feed.

That's messed up.


 

Friday, 04 November 2005 22:50:14 (GMT Standard Time, UTC+00:00)
Not that I think it acceptable behavior to ship with a knob that goes down to five minutes, much less to ship with it set there by default, but there must be a knob: I've had two Google Desktop users knocking at my feed today, one fetching precisely every 30 minutes, and another which fetched three times at one hour intervals and then stopped.
Friday, 04 November 2005 23:17:22 (GMT Standard Time, UTC+00:00)
If this were Microsoft, we could pull an Om Malik and claim that Google is doing this to undermine desktop aggregators in order to push rss readership to their personalized portal. Alas it's Google and they're geniuses so we'll conveniently forget about this.
Anti
Friday, 04 November 2005 23:33:43 (GMT Standard Time, UTC+00:00)
(The image points to http://localhost/... instead of your server's URL.)
Friday, 04 November 2005 23:39:20 (GMT Standard Time, UTC+00:00)
mardoen,
Thanks for pointing that out. It's now fixed.
Saturday, 05 November 2005 01:54:37 (GMT Standard Time, UTC+00:00)
The link to "the first time around" also points to localhost instead of your blog entry.
Saturday, 05 November 2005 09:14:42 (GMT Standard Time, UTC+00:00)
What is Google Desktop's user-agent?

bryan
Saturday, 05 November 2005 09:17:51 (GMT Standard Time, UTC+00:00)
okay, just saw it in the first time around. Will assume it is the same.
bryan
Saturday, 05 November 2005 14:23:35 (GMT Standard Time, UTC+00:00)
Quote:
...I quite frankly am turned off by people obsessing about competitors. If we want to build kick ass products then we should obsess about our customers.

Why don't you STFU, stop politicking and actually practice what you preach? It's disgusting that this visibility crap seems to get unqualified kudos at MSFT.
john
Saturday, 05 November 2005 15:31:13 (GMT Standard Time, UTC+00:00)
Oooh, BURN!!!!! John, I came from Dare's comment's on Scoble's blog (that you posted above) - so totally apropos. I actually don't mind anyone obsessing over anything in particular, I'll just drop them off my RSS if it's not something I care about. But damn your comment fit the situation. :-)
Chris
Saturday, 05 November 2005 15:39:54 (GMT Standard Time, UTC+00:00)
Nonetheless, after wandering around your site I added it to my RSS feed - look forward to hearing more from you for a bit.
Chris
Wednesday, 09 November 2005 20:27:38 (GMT Standard Time, UTC+00:00)
And even more google "badness" I had imported my OPML inot the beta newsreader at Google, and then noticed a page I setup on my site for reading my email as rss feed was getting pinged ALOT. I went onto Google reader and "unsubscribed" the feeds.

The pages still were getting hit so I removed the pages as I no longer use the page.

From my week stats emailed and today looking at my logs I see requests for the page and the resulting 404 error (since I removed the page). I looked up the ip 72.14.199.** and it says it is in Google owned range. SO they are still pinging my site even though I removed from my subscriptions. So they have my data saved somewhere.

This is not nice.
Comments are closed.