| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Linux startup crash appears to be due to static/global C++ init of LLAtomic
types. The initializer with explicit value makes some runtime calls and it
looks like these assume, at least on Linux, that apr_initialize() has been
called. So move the static POST count to a member and provide accessors
and increment/decrement. Command queue was built on a pointer to a class
in anonymous namespace and that's not quite valid. Made it a nested
class (really a nested forward declaration) while keeping the derived
classes in anonymous.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
an outfit folder now creates links to these items.
|
|
Legacy of the LLSD::Map-to-LLSD::Array conversion, this ended up
performing an erase on the array rather than the map taking out
all the regions. So, there *was* a metrics report, it was just empty
of regions. Fixed and scanned for more array/map problems and
corrected the data type for duration sorts (should have been Real).
|
|
|
|
inventory
applied Kitty's code changes listed in STROM-288 as it is a related issue and this
actualy fixed both issues!
|
|
|
|
|
|
Caused by an improper fix of STORM-690.
Made the way we switch between profile view panels (profile view / pick_info / pick_edit)
more robust. It now suits both My Profile and Profile View panels.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
after decreasing panel height
- Set height of scroll container the same as bounding panel's height
|
|
- Added localized tooltips for icons and info button in LLPanelTopInfoBar
|
|
- As the class LLToastNotifyPanel is deprecated, made the class LLToastScriptTextbox derived directly from LLToastPanel.
- Added callback for ignore button.
Now LLToastScriptTextbox has its own XML, therefore it's not needed to dynamically create toast panel.
Since LLToastNotifyPanel is deprecated all new notification toasts should be created this way.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
correct but looked a bit dodgy with pointer ownership.
|
|
of the same object using external editor.
The bug was caused by using the object ID as temporary file name for editing script,
which of course didn't work for multiple scripts in the same object inventory.
The fix is to use MD5("object id" + "script inventory item id") for the file name.
|
|
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.
|
|
LLSD serialization for viewer metrics.
|
|
First, introduced a compact payload format that allows blocks of
metrics to be dropped from the viewer->collector payload compressing
1200 bytes of LLSD into about 300, give-or-take. Then converted
to using LLSD arrays in the payload to enumerate the regions
encountered. This simplifies much data handling from the viewer
all the way into the final formatter of the metrics on the grid.
|
|
honor floater opacity settings.
|
|
Make sure older toasts don't appear after newer ones fade out.
|
|
|
|
representation transferring ownership, doing data aggregation
in a very pedantic way. That's just adding unneeded cost and
complication. Used the same objects to transport data as are
collecting it and everything got simpler, faster, easier to
read with fewer gotchas. Bit myself *again* doing the min/max/mean
merges but the unittests where there to pick me up again. Added
a per-region FPS metric while I was at it. This is much asked
for and there was a convenient place to sample the value.
|
|
on the mac script.
Should be safer if the user is installing to a path with spaces in it.
|
|
|
|
--args which is 10.6 only. Also fix erroneous check in process launcher which was mistakenly reporting a failed execution of the new updater script.
|
|
|
|
Wanted to avoid computing metrics for duplicate requests as much as
possible, they artificially depress averages but missed an opportunity
and was including them in the counts. The non-texture case is solid.
Textures are.... confounding still. Do a better job of trying to send
one last packet to the grid when quitting. It is succeeding now, at
least sometimes. Put a comment in base llassetstorage.cpp pointing
to cut-n-paste derivation in llviewerassetstorage.cpp so that changes
can be replicated. Hate doing this but current design forces it.
|
