* The areas allocated for BBitmaps were never deleted, even though the
app_server deleted its part when the memory got freed.
* This resulted in a constant memory increase if the application in question
would operate on many changing large bitmaps, like photos.
* Since the bitmaps are reference counted, we don't actually know when to delete
the areas, so that the app_server now notifies the client whenever that is
possible.
* This might fix#6824.
* Since bitmaps are reference counted, it might not be easy (and already not
the case) that holding the ServerApp lock can be enforced.
* To be on the safe side, allocations and freeing memory now performs its own
locking.
* Brought the documentation to the status quo.
- SetTopTap() -> SetTopTab().
- When reloading the decor, if the window in question is borderless,
short circuit. Previously, we wouldn't allocate a decorator for such
a window, but would then go through the remaining steps of attempting
to set the focus/top tab, which would obviously crash. Fixes#8500 and
possibly others.
- Add error check that adding tabs to the decorator actually succeeded.
* This removes the fVisibleToolTip member from BView, and fixes bug #5669;
BToolTipManager::ShowTip() now gets the owner of the tool tip as an extra
parameter.
* Removed the work-around to hide that bug.
* Improved ToolTipTest application to include more test cases like a view that
periodically update its tool tip via SetToolTip(const char*), and one that
sets a new tool tip every second.
* Furthermore, added a test that shows that inner views inherit the tool tip
of their parents.
* Fixed another bug in BToolTipManager::ShowTip() that would release an
extra reference to the tool tip currently shown.
When asked to build via a relative path (i.e. from within a generated subdir),
gcc2 was generating a compilation directory attribute pointing to the generated
subdirectory, as well as entering the actual filename attribute as the full
absolute path to the file. This wasn't checked for, leading to the issue
described in ticket #8498. gcc4 does not appear to have the same behavior,
it generates a relative path + filename in all cases.
* Use the ServerConnection class in POP3 too. This means no #ifdef USE_SSL for the connection anymore.
* I remember from the mailing list that there were problems with to many open connection using POP3. Thus, if there is an error or nothing to retrieve close the connection.
Optimization, only update the tooltip if not the same as the cached
last moused over item. That way the same application name won't
keep getting set over and over again.
If Deskbar is set to hide application names show the application
name as a tooltip when you hover your mouse over the item.
* Style fixes.
* Save CountItems() into a variable outside the loop so that it only
gets called once (micro-optimization/best-practice).
* Convert a for loop to a while loop that wasn't really being used
used as a for loop anyway, the variables are declared and used
outside the loop.
Siarzhuk Zharski informed me that KeymapSwitcher depends on this
view being named "Status" to work, so, renamed it back to that.
Instead of sending the message by name to the replicant tray view,
since the view name "Status" is not very unique and I fear that
there could be collisions, resend the message to BarView which
then passes the message on to the replicant tray which updates the
clock and reflows the icons.
This being said, KeymapSwitcher should probably not send a message
hardcoded to the name of the replicant tray view and instead send
the message to Deskbar and allow it to pass the message on to the
appropriate view.
* Resolves blocker #8322.
* Should resolve#8151, #8163, #8282, #8457
* Digital output on DVII was getting hard set
in the encoder to analog / CRT mode.
* We now check the edid "type" flag
* Flag gets set to 1 if edid data from digital
transmitter, 0 if analog transmitter.
If CalcView is embedded as a replicant on the desktop it's parent
window is not CalcWindow. Furthermore, it can't switch between
compact, basic and scientific keypad modes, you are locked in tox
the one you picked at the time you embeded the view into the desktop.
So, I have updated the right-click message to reflect this reality
by taking away the keypad mode switching options. You can still
set auto num-lock and turn beep on and off (theoretically someday
anyway) but the CalcView receives these messages instead of
CalcWindow. So, these 2 options now work in the case of a CalcView
embedded in the desktop at least.
* Create my own kIndentSpacing parameter from ControlLook which sets
the indent spacing width to match the previously used spacing. This
lines the indented checkbox up with the text of the unindented
checkbox above it. This matches what it looked like before I
updated the dialog to use font-aware spacing, while keeping the
same relative spacing at different font sizes.
* Nobody thought time zone was a particularly useful option to have
in the clock.
* You can still open Time preferences from Deskbar by right clicking
on the clock, but, since there isn't a big connection between
the clock settings in Deskbar and Time preferences anymore
this button isn't needed here.
* There are still 2 clock options, Show seconds, and Show day of week
and day of week is localized as well so this wasn't a total wash.
instead of "_replicant_tray_". That follows the convention a bit
better and is less likely to cause name collisions. While I am at
it put the BView line below the : as is our usual convention.
* 12/24 hour clock is set from the Locale prefs only.
* Redirect B_LOCALE_CHANGED message to the replicant tray instead
of the time view and tell the replicant tray to update the time
instead of updating the time view directly. This is because there
was a bug where if the AM/PM marker grew the time view over the
replicant tray icons when switching from 24 hour to 12 hour clock
it wouldn't reflow the replicant tray icons like it should.
* Rename the replicant tray view from "Status" to "_replicant_tray_"
because BarApp uses FindView() to get the view by name and I'm
assuming that this method can have conflicts. Following the example
of the "_deskbar_tv_" naming convention.
* First steps at getting card command processor wired
up to the ring buffers.
* Code doesn't run yet as I have *no* idea what happens
when these rings are in an invalid state.
