Age | Commit message (Collapse) | Author |
|
Review #97
This hopefully fixes the incorrect behavior of using the external browser when the target is specified.
|
|
Now LLUI::getLanguage() returns a result that is consistent on Mac,
Windows, and Linux. Also, cleaned up some legacy language code.
|
|
is external or pref is set
Review #86
This changes the logic to bring up the notification "Do you want to open your Web browser to view this content?" only if either the target is _external, or the user pref says to. Otherwise it avoids the notification.
|
|
Support href links in web views that have a target attribute set to
"_blank". This will open the link in the user's preferred browser
(internal or external, as defined in Prefs). This change relies on the
new llqtwebkit package that has just been pushed.
Also updated the hardcoded Qt version number in the About floater.
|
|
search/help windows. This should address EXT-4097.
Added a notion of "background color" to LLViewerMediaImpl and LLPluginClassMedia.
Added background color parameters to the size_change message.
Webkit plugin now sets the background color of the instance from the supplied background color, and navigates to a data: url with that background color instead of about:blank as its initial navigate.
Webkit plugin now no longer waits for the first onPageChanged event
LLViewerMediaImpl now clears the texture to the background color when initializing it.
Made LLMediaCtrl fill with its opaque background color when the media impl isn't set up yet.
Removed the initial data URL from the search and help floaters, since what it was doing is now handled internally by the new background color code.
Reviewed by callum and rick.
|
|
The url from the cmd line was being overridden by the LoginLocation setting. (my bad!)
Removed unnecessary llurlsimstring.h includes.
|
|
the login screen.
Added clearCache() functions to LLViewerMediaImpl and LLMediaCtrl to facilitate this.
|
|
Cuts number of includes of llnotifications.h from 300+ to 40.
|
|
reviewed by James
|
|
|
|
Reinstated the hide_loading parameter to LLMediaCtrl, and made it cause LLMediaCtrl::draw() to early-exit between creating the LLViewerMediaImpl and receiving the initial MEDIA_EVENT_NAVIGATE_COMPLETE.
|
|
|
|
to most user events and using new scroll wheel event.
|
|
into the password field).
Moved some special-case handling from LLMediaCtrl::handleKeyHere/handleUnicodeCharHere to LLViewerMediaImpl::handleKeyHere/handleUnicodeCharHere so that the key handling for LLViewerMediaFocus gets the benefit of it as well.
|
|
MEDIA_EVENT_NAME_CHANGED.
|
|
|
|
|
|
fails) causing an error message loop:
Made LLPluginProcessParent differentiate between failures launching/loading the plugin and failures after the plugin has been loaded. This allows us to handle launch failures differently, since retrying is unlikely to fix them.
Added new media event MEDIA_EVENT_PLUGIN_FAILED_LAUNCH to indicate a launch failure.
Added a case for the new event to LLViewerMediaImpl::handleMediaEvent() that sets the "failed init" flag to prevent retries.
|
|
svn+ssh://svn.lindenlab.com/svn/linden/branches/media-on-a-prim/moap-7
Merging branches/media-on-a-prim/moap-7 down to viewer-2.0.
|
|
svn merge -r134800:134805 svn+ssh://svn.lindenlab.com/svn/linden/branches/viewer2help-3
|
|
following SLAPP URL types in an untrusted browser:
secondlife:///app/agent/...
secondlife:///app/group/...
secondlife:///app/parcel/...
In order to find a compromise between supporting these commands and
security concerns over potential griefing vectors, we use a throttling
solution when these commands are issued by untrusted web browsers.
That is, we only process one command per 15 seconds.
This applies to external browsers, like Firefox, as well as the
internal SL browser.
Notably, we continue to block secondlife:///app/teleport URLs.
Reviewed by james.
|
|
svn+ssh://svn.lindenlab.com/svn/linden/branches/moss/pluginapi_05-merge@129910
svn merge -r 129913:131718 svn+ssh://svn.lindenlab.com/svn/linden/branches/pluginapi/pluginapi_05
Some branch shenannigans in the pluginapi_05 branch caused this to become a two-part merge.
|