Welcome to MSDN Blogs Sign in | Join | Help

Google Desktop - Part 2

Nathan Weinberg apparently spoke to Google's Marrisa Mayer and comments on my concerns.

Nathan claims that #10 and #1 are by design. Yes, apparently they are by design - it would be quite hard to incidentally hook up the WinInet.dll or to install a BHO. That doesn't answer my concerns about them, though.

#10 - The BHO is probably the way Google Desktop intercepts the Google.com search results and injects the desktop results in the page. But it also means that GD also has access to any page I browse, even if it's https: and my IE is set to not cache on the hard drive secured pages. I hope this doesn't turn out to be same as the AIM chats case.

It's interesting though that the only indication there is a BHO installed was a dialog box when GD installation hit IE with disabled BHOs. Why isn't this mentioned somewhere else, like in the documentation? Why isn't there an option to not install the BHO?

#1 - Why does a desktop search tool needs to hook up WInInet.dll? My guess is that this is how they inject the desktop search results in other browsers. But why not do it this way in IE as well? Or if it does it, then what's the BHO for? Why isn't there an option to not do hook up WinInet.dll?

And then comes the matter of AIM, which for the sake of the discussion I shall name Number Twelve and which everybody conveniently skips.

Published Thursday, October 21, 2004 11:13 AM by Franci Penov
Filed under:

Comments

# re: Google Desktop - Part 2

Hello Mr Penov,

It's probably just the begenning of a little saga! Many question are pending; Mr Weinberg don't answer them, and nobody does.

Sometime our discussions are probably paranoia, but possibly not. But what's really interesting is that many people are interesting in the possible treat; they talk of it and express their fears about it. Possibly that this attitude will force in some way Google to take into account our fears to build their product. It's what it's really interesting in the whole story.


Salutations,
Thursday, October 21, 2004 12:06 PM by Fred

# re: Google Desktop - Part 2

I wonder why they do it this way instead of just putting a control in the HTML of normal Google search result (via <object>). If they put a control the HTML they could do the local search as soon as the search results come back from the standard search. It wouldn't happen simultaneously, but it also wouldn't be as invasive to the user's system as it sounds it is right now.
Thursday, October 21, 2004 1:40 PM by Drew Marsh

# re: Google Desktop - Part 2

I can't claim to understand the technical underpinings of the web browser, but I would assume the reason for 1 and 10 are because Google Desktop does not work by indexing the web cache, but by intercepting traffic and reading; any more than it doesn't index some sort of AIM log, but instead intercepts AIM chats with the now-infamous googleim:/ protocol.

As for the AIM question I just respond to the numbered ones. I can't see anything if it doesn't have a number :-)

But seriously, the AIM things seems like a non-issue. If you want Google to index something, let it index it. If you don't, don't. There's no reason to accuse Google of trying something dirty by having a feature. Google Desktop is supposed to keep track of everything. That's the benefit. If that's not a benefit to you, why would you install it?
Thursday, October 21, 2004 2:46 PM by Nathan Weinberg
New Comments to this post are disabled
 
Page view tracker