Removed the skipping of a working test as this was causeing a the test to fail
    building and thus not actualy running all the tests.

To use ld.gold configure with:
  -DCMAKE_EXE_LINKER_FLAGS:STRING="-Wl,-use-gold".
ld.gold links the viewer on my machine in 8 seconds, as
opposed to 19 seconds with ld.bfd. Moreover, it uses a
LOT less memory during linking (about 750 MB instead of
2.5 GB!).

VWR-24254: Don't link with fontconfig on non-linux.

While we already added fontconfig in the above patch,
that code turned out to also be used by Windows and
Darwin (contrary to the comments in the code).
After looking at the history of commits and a
discussion on IRC it was decided that the original
coder (Kyle Ambroff <ambroff@lindenlab.com>) really
meant (LINUX AND VIEWER) instead of (NOT LINUX OR VIEWER).

Add non-broken FindZLIB.cmake.

If tut/tut.hpp isn't installed in a standard include directory all tests
fail because the found include directory for tut isn't passed to the compiler.

This patch fixes this by passing it.
Note that using include_directories() in a Find*.cmake file is bad practise.
The correct way is to set an include dir variable and call
include_directories() once. It certainly doesn't work for the tests anyway
because the tests are all over the place and include_directories is on a
per folder basis.  What is needed is to set it for each (test) target.

However, there is no TARGET_INCLUDE_DIRECTORIES. The closest thing that we
have is to set the COMPILE_FLAGS property for a target.

Fortunately, standalone is only used for linux, so we can just use
-I${TUT_INCLUDE_DIR} to get the effect we want.

The old one just assumed that libndof is installed,
even on standalone.

This patch looks for it on standalone, unless configured
with -DNDOF:BOOL=OFF, and when not found prints an error
message (including telling about -DNDOF:BOOL=OFF).

SNOW-240 was actually about adding the whole search.
While this has been done, there is one issue left:
On linux (and remember, all of this is about standalone)
the libjson packages of distributions don't have this
complex compiler version baked into their name.

See also the last comment on SNOW-240:
https://jira.secondlife.com/browse/SNOW-240?focusedCommentId=205305&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-205305

This patch fixes this issue by first searching for
libjson_linux-gcc-${_gcc_COMPILER_VERSION}_libmt.so
and when that fails search for the system package
library file libjson.so.

If Qt and/or llqtwebkit is found in a non-standard place, you still
have to set LD_LIBRARY_PATH yourself (to $QTDIR/lib) before running
the viewer of course (or the webkit plugin will silently fail).

The viewer code includes "tut/tut.hpp" everywhere, so that is what we
should search for when determining an include path for this header.
This just searches in the usual places, including the paths provided
through CMAKE_INCLUDE_PATH, and will find tut/tut.hpp: exactly what
we need.

This fixes the problem that when some package contains symbolic
links, they were never removed anymore (after an uninstall of
that package).

   applied Kitty's code changes listed in STROM-288 as it is a related issue and this
   actualy fixed both issues!

STORM-229 FIXED Fixed long loading times and stalling of Viewer while loading big scripts or pasting a lot of text into script.

The bug was fixed by Satomi Ahn. Here is the description of what causes the problem from her comment in ticket:
"Disabling the loading of syntax keywords in LLScriptEdCore::postBuild() removes the freeze (and with it: syntax highlighting).
So it obviously comes from the parsing of the text.

I also noticed something else: by adding a llwarn in LLTextEditor::updateSegments(), I saw that this function was called a lot of times when loading a script, roughly once for each line in the script (naively I would have thought only necessary to update when finished... or to only update the new line).
My llwarn was in the "if (mReflowIndex < S32_MAX && mKeywords.isLoaded())", which means that, at each call, the text is actually parsed for all keywords... so the parsing of the script becomes quadratic instead of linear!!!"

- To fix this, Satomi added a flag depending on which parsing is disabled when it is not necessary.

Modified code in indicated changeset to correct problem after consulting with
 Leyla Linden who sugested the orginal change. This fix will convert the Legacy
 name to a user name id not useing Display Names.

 after loss during cleanup of preferences.

Reason: The old shortcut (Ctrl+Alt+D) was eaten by some window managers.

Changes:
- Changed the shortcut to Ctrl+Alt+Shift+D.
- Moved the appropriate menu item from "Advanced > Shortcuts" to "World > Show"
  (so that it's not in the menu it triggers) and made it visible.
  The old shortcut is still available but marked as legacy.

Submitting on behalf of Boroondas Gupte.

 matter when it was generated or if it has the date stamp in the name of the log or not.

--HG--
branch : storm-102

--HG--
branch : storm-102

This is a patch originally written by Robin Cornelius.
I made it work with Google Breakpad.

STORM-420 : allow teleport when double clicking the minimap if doubleclick teleport is on, allow beacon setting if not

LLTextBase::setCursor() sometimes failed to work properly if line wrapping was enabled.
This is a slightly optimized version of the patch made by Satomi Ahn.

 Chat, Group and IM Logs.

--HG--
branch : storm-102

The point of this patch is to make the Mac updater code a bit more flexible
and reliable than it is right now. The issue is double:
 * reliability: the string comparison code on the bundle identifier is not UTF8 compliant
 * flexibility: the bundle identifier is hard coded to match the bundle identifier of LL viewer
   (i.e. com.secondlife.indra.viewer) so it can't work for another viewer
   (in particular, it didn't work for Snowglobe).
   The "bundle identifier" is one of those Mac only thing stored in the Info.plist of a "bundle"
   (the ".app" folder that's bundling an executable and all its resources and is seen
   as an application when browsing with the Mac OS X Finder).

The patch fixes both issues:
 * compare correctly UTF8 encoded strings
 * allow the bundle ID to be passed as a parameter to the updater

The patch has really no consequence on LL viewer. It's more a matter of having cleaner, better code.

Author: Cypren Christenson
Ported and reviewed by: Merov Linden

Author: Robin Cornelius
Ported by: Techwolf Lupindo
Reviewed by: Merov Linden