| Age | Commit message (Collapse) | Author |
|---|
|
tested
|
|
|
|
|
|
|
|
|
|
repository.
|
|
added buffer merging and copying
|
|
created buffer type for storing trace data
added merging logic for different trace types
|
|
further development of lltrace (accumulator and storage classes)
|
|
lltrace cleanup
|
|
|
|
|
|
|
|
|
|
|
|
|
|
InspectObject.TakeFreeCopy
|
|
LLUI::setupPaths() went away with DRTVWR-210.
|
|
|
|
|
|
|
|
|
|
reference.
|
|
HUD groups. clearRebuildGroups() is called during the teleport process. In the previous state, HUD objects were being incorrectly removed from the rebuild groups during the teleport process. This change will preserve HUD groups and should resolve this issue.
|
|
|
|
|
|
the sUpdateDelay value and simply updating all flexi-prims. This change most likely reduces frame rate.
|
|
method. The callstacks indicated that the std library was throwing a length exception on the std::vector::resize() call. Most likely cause was that the recent changes for flexi-prims were causing the mRenderRes to become negative. And thus, the 1<<mRenderRes calculation was generating a really large value for num_render_sections.
|
|
|
|
|
|
|
|
Two fairly simple conflicts: dead stats sending code in the
texture fetch code (new llcorehttp library) and the cleanup
code in llappviewer was moved around in 3.4.x.
|
|
|
|
|
|
|
|
The tricky thing about this fix is that the "Log In" button used to be in the
same layout_panel as the username and password fields. Now it's not, so the
fact that it's the default button for its layout_panel doesn't matter because
that layout_panel doesn't have focus. Richard pointed out that we can make the
commit action for the username and password fields initiate login -- as long
as neither field implicitly runs its commit action when it loses focus!
|
|
|
|
Original LLDir::findSkinnedFilenames() implementation used a tricky rule: a
given skin directory was only considered if it provided a default-language
override for the sought filename, regardless of whether it also provided
localizations for that filename. Discussion with Richard clarifies that we
want to allow the user to override neither, either or both. Change
findSkinnedFilenames() accordingly; update unit tests to verify new semantics.
|
|
|
|
|
|
Previously we borrowed the sourceid= param value from create_account_url,
which we "happened to know" was overridden with the sourceid of interest. Now
that we have a settings variable that directly captures sourceid, though, much
more straightforward to use that.
|
|
join.secondlife.com needs to know the sourceid as well as the user's language
of choice. Ensure that sourceid gets passed with the URL.
|
|
The whole point of --skin steam was to override (specifically)
"create_account_url" in strings.xml, adding to each URL value a sourceid= URL
parameter. Now we can do that more simply with '--set sourceid blah'. Less
overhead, less maintenance, scales better to potential future sourceid values.
Remove the steam skin.
|
|
|
|
Calling LLTrans::setDefaultArg() after LLTransUtil::parseStrings() is almost
good enough -- but it fails to address the case in which one or more of the
default_trans_args strings (e.g. "create_account_url") embeds a reference to
the new substitution. So after the setDefaultArg() call, go back through
default_trans_args, refetching each string to perform the substitution and
updating it with a setDefaultArg() call of its own. All this is way too much
logic to replicate in both LLAppViewer::initConfiguration() and init(), so
break out new LLAppViewer::initStrings() method and call it from both places.
|
|
Add logic after (both!) LLTransUtil::parseStrings() calls to ensure that
"[sourceid]" embedded in (e.g.) strings.xml content will be replaced.
|
|
Richard points out that LLUICtrlFactory::findSkinnedFilename() adds little
value. It was called from exactly one place, and that one place could easily
obtain the information another way. The concern is that it could confuse a
reader of the code with regard to the other findSkinnedFilename[s]() methods
in LLDir. Clarifying the code base is a Good Thing. Removing.
|
|
Per code review:
Previous refactoring of LLUI::locateSkin() preserved odd failure behavior: it
would return last-considered pathname, whether or not it exists. Changed to
emit LL_WARNS log message and return empty string.
Use Boost.Assign to simplify initialization of a couple static containers in
lldir.cpp.
|
|
At this point, LLUICtrlFactory::getLayeredXMLNode() is a pretty thin wrapper
around LLDir::findSkinnedFilenames() and LLXMLNode::getLayeredXMLNode().
Until now, LLUICtrlFactory::getLayeredXMLNode() passed (by default)
LLDir::CURRENT_SKIN to LLDir::findSkinnedFilenames(). But that meant that a
caller such as LLTransUtil::parseStrings() that wants almost the same
functionality, but with LLDir::ALL_SKINS instead, had to clone the logic from
LLUICtrlFactory::getLayeredXMLNode(). Allowing its caller to pass the desired
LLDir::ESkinConstraint enum value eliminates the need to clone its logic.
Remove cloned logic from LLTransUtil::parseStrings().
|
|
At Richard's suggestion, changed the bool merge parameter to new enum
ESkinConstraint with values CURRENT_SKIN and ALL_SKINS. This clarifies what
we're requesting at the point of the call.
|
