After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 602196 - Middle-click on main window triggers Google search
Middle-click on main window triggers Google search
Status: RESOLVED FIXED
Product: epiphany
Classification: Core
Component: Interface
2.29.x
Other Linux
: Normal normal
: ---
Assigned To: Epiphany Maintainers
Epiphany Maintainers
Depends on: 604087
Blocks:
 
 
Reported: 2009-11-17 13:39 UTC by Josselin Mouette
Modified: 2012-10-07 15:26 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Josselin Mouette 2009-11-17 13:39:09 UTC
When there is some text in the clipboard and I middle-click on the main window, it immediately starts a Google search with the contents of the clipboard.

This is a security issue since it can leak sensitive information to Google. Even worse is the fact the search engine cannot be configured.
Comment 1 Reinout van Schouwen 2009-11-17 14:29:35 UTC
Agreed, there should be a possibility to configure the default search engine.
Comment 3 Reinout van Schouwen 2009-11-17 14:51:16 UTC
Xan,

Loading an URL from the clipboard is not _quite_ the same as feeding clipboard contents to Google. Perhaps we should allow the former, but not the latter.
Comment 4 Xan Lopez 2009-11-17 14:53:18 UTC
It just passes whatever is in the clipboard to epiphany. Epiphany will search in google when the data is not an URL, same thing that happens in the entry location.
Comment 5 Dan Winship 2009-11-17 15:18:13 UTC
ugh, i always turned that behavior off in firefox. in my experience, most of the time when i middle-click on the window content, it's because either (a) i missed the link I was trying to open-in-new-tab, or (b) my finger slipped while scroll-wheeling

i agree that being able to middle-paste a URL would be useful, but this always seemed like a bad way to do it to me. But Reinout's suggestion makes sense; if the string is syntactically a URL (or even a hostname maybe), it was probably an intentional paste
Comment 6 Xan Lopez 2009-11-17 15:35:00 UTC
I don't like this either, but I don't think this is the point. This is a hidden preference, which is *false* by default. If you don't like it don't enable it, simple as that. It could be argued that the whole thing is pointless and that we shouldn't even support it, but that's a different issue and a different bug.
Comment 7 Dan Winship 2009-11-17 15:41:22 UTC
(In reply to comment #6)
> This is a hidden preference, which is *false* by default.

Oh! Was wondering why I hadn't tripped over it yet :-)
Comment 8 Josselin Mouette 2009-11-17 17:00:48 UTC
(In reply to comment #6)
> This is a hidden preference, which is *false* by default. 

D’uh, I wondered how that could happen. You are right of course. I had enabled it at a time when it was safe to do so (invalid URIs were rejected) but apparently it is no longer the case.

Well, my request to disable the “middle click to search engine” behavior stands, but since it’s not the default it’s a less important issue.
Comment 9 Xan Lopez 2009-11-17 17:11:22 UTC
(In reply to comment #8)
> (In reply to comment #6)
> > This is a hidden preference, which is *false* by default. 
> 
> D’uh, I wondered how that could happen. You are right of course. I had enabled
> it at a time when it was safe to do so (invalid URIs were rejected) but
> apparently it is no longer the case.
> 
> Well, my request to disable the “middle click to search engine” behavior
> stands, but since it’s not the default it’s a less important issue.

I agree that since ephy used to reject anything different than an URI in this feature we should, in theory, get back to that behaviour. I won't reject a patch that does that, although as I said my personal opinion is that this feature is basically silly.
Comment 10 Reinout van Schouwen 2009-12-08 16:15:30 UTC
About the Google issue, see bug 604087.
Comment 11 Kelvin Li 2012-04-29 08:24:02 UTC
It looks like there's no more issue. Can this bug be closed?