If PeerAddress is cleared before invoking GetLocalAddress() we make sure
it won't contain any garbage data in a very unlike case when getsockname()
fails.
repo_internalize() apparently reorganizes the storage for Solvable
objects so that our fSolvablePackages and fPackageSolvables maps could
contain invalid pointers. Now the maps use the solvables' ID instead.
Also use the IDs instead of the Solvable objects in most other places,
which in some cases even simplifies the code a bit.
Position the text control vertically in the middle of the ramp in
the case that the text controls are pushed down so that the red, green,
and blue text controls are next to the red, green, and blue ramps.
In a few cases we were doing the work of _LayoutView() twice because
we called _LayoutView() and then called ResizeToPreferred() which called
_LayoutView() again.
Now only call ResizeToPreferred() which calls _LayoutView().
* Also change kMinCellSize from a uint32 to a float so that it can be used
with std::min() and std::max() instead of min_c() and max_c().
* Set the text controls sizes and margins based on the font size. Also rework
_TextRectOffset() so that it will get the right spacing from by dividing the
palette frame by 3.
* Replace bare numbers and refactor with calculation or magic constant.
* Create a private method _TextRectOffset() which calculates and
returns the vertical text rect offset to use based on the font size.
* Replace 2.0 with new kBevelSpacing constant where appropriate.
* fPaletteFrame calculation in _LayoutView() was refactored but should
not have changed.
Since hrev45725 we no longer resize fixed sized menu fields to their preferred
height in the constructor so as to be backwards compatible with the behavior on
BeOS R5. As a consequence though, the menu fields in Mail are now a bit too
tall so this commit tweaks the menu fields to be just a bit shorter matching
the height of the text controls.
That commit switched to a single string column for representing
breakpoint locations, but neglected to update the case of a breakpoint
in a non-debug function to actually format the address as a string
appropriately, leading to those simply showing no data since then.
Since some time now, we construct minimal dwarf image debug infos
for release images if they contain frame unwind information, in order
to allow using that to unwind the stack in as many cases as possible.
As such, it's entirely possible that such an image may be asked for
statement information regarding a function that isn't in fact compiled
with debugging. As such, we need to simply fall back to disassembly
rather than failing entirely. Fixes setting breakpoints in such
functions (i.e. anything in a release version of libstdc++).
We have to use actual targets that cause the respective download and
extract the packages. Otherwise the build fails when the packages
haven't been extracted yet.
Missed that when adding the script. Therefore it would be created in the
current directory and when building multiple packages concurrently the
script would be overwritten.
* Make grist for subtype source files unique.
* Our MakeLocate only appends to LOCATE, so we have to call it before
ResComp, if we want a different location.
Fixes #9816
It is no longer necessary, or even desirable for us to set the max
content width of the menu bar of a BMenuField now that BMenuItem
truncation and BMenuField sizing are working.
The user may, however, wish to set the max content width of the menu
bar of a BMenuField themselves like so:
menuField->MenuBar()->SetMaxContentWidth(width);
and the Interface Kit will automatically deduct the left and right
margins from the width including the space used by the drop down arrow.
The new configure option "--use-xattr-ref" enables an xattr assisted
variant of the generic attribute emulation. Instead of using the inode
ID of a node to identify its attribute directory, we use a reasonably
unique random 128 bit number, which we generate and attach as an
attribute to the node. This way, when a node changes its inode ID
(defragmentation?) or the inode ID of a removed node with a left-over
attribute directory is reused, attributes won't get mixed up.
The old method is still used for symlinks (since on Linux only
priviledged users can write attributes on symlinks), but those usually
only have a rather boring BEOS:TYPE attribute, so mix-ups wouldn't be
that problematic anyway.
When a filter is active, only expand parent nodes if either a) there is
only one matching parent, or 2) the match actually hit a function
contained in it. This allows the case where the intent of the filter is
to find a particular set of files or subdirectories to be handled more
efficiently.
ImageFunctionsView's filtering field now allows shell-style glob matches
rather than just simple direct string matches. Implements remaining
part of #7955.
Store the filter match index on the SourcePathComponentNode so we can
retrieve it directly when drawing instead of having to recompute the
position of the match every time.
- ImageFunctionsView now contains a text input allowing one to specify a
filter for its contained functions.
- When in filtered mode, the previous flattened view is used rather than
the hierarchical tree.
- The matching portion of the string is highlighted. However, currently
only simple string matches are supported.
* For all identifiers: Rename global settings file to global writable
file. We want to use the respective attribute also for other writable
files, not only settings files.
* User settings file/global writable file info/attribute: Add
isDirectory property/child attribute. This allows declaring global/
user settings directories associated with the package.
* when an executable with a different ABI is being loaded and some
needed image isn't found, don't retry using the standard ABI folders
of the system - those are now considered incompatible
