...while the setting "Text files only" is active.
Fixes#12539.
Once users are able to set supported MIME types on packaged apps
(which are read-only), it might be a good idea to have TextSearch
check its supported MIME types and examine the target files accordingly.
BNode::ReadAttrString does not check for the attribute to actually be a
string. It will return B_OK (and useless data) if called on a
B_INT32_TYPE attribute. Check the attribute type before reading.
Also remove various debug ouput leftovers.
May fix#12976.
* Helps prevent mainboard explosions and other bad stuff.
* ...maybe not explosions, but bad things.
* The RX 480 reference design can pull as much as 90W
from its PCIe slot at load. PCIe is rated for 75W.
* This change overloads the PCIe power cables vs the PCIe
slot. The PCIe power cables can handle going over spec.
* Untested since we really can't come close to loading a
RX 480 without hardware OpenGL ^_^
* Should be able to test on real hardware within a week.
Based on a patch by Duggan.
A timeout of 0 means "no timeout". Avoid an overflow case. Translate
B_INFINITE_TIMEOUT to a NULL timeout for select().
Fixes#7870
BeRometer works now.
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
This is a private API, but some apps used it anyway. We don't need a
complete implementation to at least make BeRoMeter run.
Fixes#5146.
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
This spent enough years waiting on the bugtracker for someone willing to
test it. The style looks mostly ok, so let's have it in trunk and see if
it breaks someone's tablet.
The patch restores the older behavior found in early versions of the
driver published for BeOS. It does so only for the "intuos" tablets, so
newer devices won't be affected.
ticket : #4847
As usual we ask ICU to do the actual work. The TextEncoding constructor
is fed with a sample of the text to identify (ICU docs recommend a few
hundred bytes). The text is analyzed in various ways (bytes patterns
such as UTF-8 escaping schemes, common letter sequences from known
languages, byte order marks) and an encoding is determined.
Replace code in StyledEdit by this new implementation.
Note that ICU seems to always return some valid encoding, even with fed
with obviously non-text data. This makes StyledEdit open the files no
matter what, where it would error out before.
Fixes#9395.
This reverts commit 8cf0752f1b.
This reverts commit af3057423c.
This change fixed one package, and broke everything else. Please find a
proper fix and actually test it before committing.
Fixes#12909
These were left out because the same RegExp.cpp file is also used by
collectcatkeys. This needed some jam file changes to get collectcatkeys
to build on both the host and target.
Fixes the last remaining part of #6374.
Tracker can create query templates when they don't exist yet. This was
done with an hardcoded archive of BColumns for the pose view, which did
not allow for an easy way to replace the strings. Now the archive is
created on the fly when the function is run, which makes it possible to
build it with localized strings (and in the system's endianness, which
will make it load faster).
Based on patch by mt, but with a lot of rework.
Fixes part of #6374.
- Remove hardcoded decorator border size, and get it from the system.
- Move the window at the right place when using the "zoom" button
(if the auto-raise setting is enabled).
- Make the auto-raise work even when the mouse is above the window
borders, not the active area. This works similarly to LaunchBox
auto-raise feature.
fixes#8188, #8194.
Manual rework of patches by Daniel Devine which were never merged and
didn't apply cleanly anymore.
Activated SSL encrption for POP3 and changed server name for SMTP to
mail.gmx.net. The first change is vital to get it to work, the latter
seems to be only cosmetic.
I still get a "503 Bad sequence of commands" when sending a mail, but
it is sent nonetheless. See ticket: #12960.
This patch reverts http://cgit.haiku-os.org/haiku/diff/?id=57ab0395ad31761e27ef6d5aa3af68cc3e4d71b2
It may fix#661. Thanks to kcg369 for pointing it out.
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
The initial issue was the storage of a pointer in an int32 for
input method cookies. This was fixed in two ways:
- The int32 was initally replaced with a void*, which allowed to store a
complete pointer, but changed the format of the BMessages used for
communication, breaking Canna.
- Then, the code was changed to not use a pointer and instead use an
int32 for the cookie (see #8831). However, the int32 was still cast
into a void* when putting it into a BMessage, so the problem was still
there for Canna.
This commit reverts the remaining parts of the initial solution and
restores the ABI to use int32 everywhere.
Fixes#661.
Select the text to find as the Find window is opened.
This makes the behavior more consistent with other apps.
Fixes the main subject of #10844.
Also removes trailing whitespace.
Signed-off-by: Axel Dörfler <axeld@pinc-software.de>
The code was resetting the pointer to NULL, instead of setting the
pointed char to null-terminate the string. The result was resolving the
current directory later on, instead of the parent as requested.
This only happened when calling open_from with O_CREAT, and a name with
sub-directories inside it. The boot loader never does that, so the code
wasn't used until now.
Fixes#12941.
Thanks to mt for spotting the problem.
- Add new interface TeamFunctionSourceInformation. Currently this
exposes a single function allowing one to query for the currently
active source code given a FunctionDebugInfo instance.
- Implement TeamFunctionSourceInformation on TeamDebugInfo.
- Pass TeamFunctionSourceInformation to Dwarf{Team,Image}DebugInfo.
In turn, make use of it in DwarfImageDebugInfo::GetStatement() in
order to determine whether to return the corresponding assembly
or source statement.
With this piece of information, the debugger is now correctly able to
determine that the user is currently looking at disassembly despite debug
info being available, and consequently adjust its stepping behavior based on
that. Previously, the source code statement was always used, leading to it
not being possible to single step assembly lines in such a circumstance
without manually using run to cursor.
Other related cleanups:
- TeamDebugInfo now inherits BReferenceable directly, rather than relying on
indirectly inheriting it from TeamTypeInformation.
- Remove BReferenceable from TeamTypeInformation. The latter is only an
interface anyways, and inheriting that base class from multiple locations
was causing GCC5 trouble when resolving BReference<TeamDebugInfo>, even
when virtual inheritance was used.
* BReferenceable deleted itself when the latest
reference was released causing a double free.
* The code now avoid to call the function when
we are about to delete the object.
* Cleanup object management.
* Remove Close() and use the destructor in place.
* Release listener sem on exit.
* Return when there's a request mismatch.
* Add some debugging.
* MediaPlayer crashes are fixed as result of the patchset.
