* Both functions need to wait when encountering a busy vnode, in order
not to create a race condition in combination with get_vnode().
* This should fix both #5262, and #9839.
* I did not implement Ingo's suggestion to burden the file systems with
solving the problem for the following reasons:
- I think the file system interface should be as simple as possible.
- I can't think of a possible locking issue at least for BFS.
- The solution on the file system side would look pretty similar to
what the VFS already does or has to do (minus controlling the locking
directly), so it would cause quite a functional duplication.
As pointed by Marcus, FindString returns a pointer to the internal data
of the BMessage, so the message must stay in scope as long as we use the
pointer.
Thanks for reviewing!
ffmpeg, especially older versions, is not very good at detecting file
types and relies on us giving it a file extension for a better guess.
Use our MIME sniffing to have a first try at detecting the file we are
about to play, and pass this information to ffmpeg.
Fixes#13150.
*Bug fix- list item colors use Appearance settings
*Problem Solver window buttons right aligned, window now app modal
*The main window's stripe and icon now use the same scaling as BAlert
*A pop-up menu displays an option to view more details (replaces the
tooltips)
*New prompt at application start to choose between doing an Update or
a Full Sync. Also lays the ground work for implimenting silent
check-only feature.
Needed a quick new release:
* Fix disappearing favorites.
* Add context menu with quick access to favorites and QuickLaunch's
"About" to Deskbar replicant.
* Updated localization and documentation.
* Tiny tweaks to the QuickLaunch icon.
fActiveSoundFont wasn't initialized when there's no midi_settings file.
Don't allocate it on the heap, the object get implicitely initialized to
an emtpy string.
Tiny style fix.
Thanks to AnEvilYak for his consultation.
Fixes#13402.
* The previous xhci endian fixes performed some double
conversions.
* Tighten down conversions to only where the driver interacts
with raw trb3 data in the ring buffer
* No impact to x86*
Connecting to wifi is very slow, there are several issues.
Easiest to find and fis is that on my wifi DHCP always took
4s, as first request fails. Still investigating why..
Reducing timeout to 0.25s makes DHCP process fast.
Moved timout handling into its own struct, and changed state
timeout from max(remaing, 60s) to min(remaining, MAX_TIMEOUT)
Not sure about that change, but why would you want a max value
that is at least 60s?
* BView incorrectly passed the current width and height to
FrameResized(), not the one from the message. Since there is such
a call for each size change, a Draw() might have been called for
each of those in return.
* This should also fix such problems in BMenuBar, as it used the size
from FrameResized() as it should have.
* BBox now correctly takes the update events into account, and no
longer ignores the size passed to FrameResized().
* This fixes bug #3037.
* Solves #13388 and likely a bunch of other bug reports.
* While this *is* an issue that can potentially be serious,
there really isn't a risk of data loss or a system
instability, only broken PCI devices. Show warning and
keep going.
Bug fix- show correct version on package uninstall list item tooltip
Correct if else formatting
Use the BPackageManager FullSync() function instead of Update
Change warning strings in BPackageManager to sentence casing
Show list item tooltip on a mouse down (helps with touch devices)
Bug fix- use a proper sized 32 pixel application icon for notifications
Bug fix- sorting list of packages to install and uninstall added
Removed use of FinalWindow class now that resizing issues are resolved
Diaplsy alert if no network connections are found at application start
* Reintroduce change in last 2016 acpica upgrade
* Our semaphores are int's. This is the only change needed
to the acpica source code to get it to function under
Haiku.
* We could also disable this int-conversion warning,
however lets go with the "more correct" behaviour in
something that pretty much controls how our kernel finds
PCI devices, CPU's, etc :-|
Notable changes since 20160729:
* Will be dual licensed going forward. We choose the Intel License.
* Fixed two issues with auto-repair double-delete an internal object
* New hook for sleep support (AcpiOsEnterSleep)
* Fixed a regression introduced in 20160729, "FADT support cleanup"
around early 64-bit address support.
* Fixed a problem where runtime implicit conversion was
incorrectly disabled for the ASL operators below. This brings the
behavior into compliance with the ACPI specification:
FromBCD,ToBCD,ToDecimalString,ToHexString,ToInteger,ToBuffer
* Fixed a regression in the internal AcpiTbFindTable function where
a non AE_OK exception could inadvertently be returned even if the
function did not fail. This problem affects the following operators:
DataTableRegion,LoadTable
It was licensed with a "no commercial usage allowed" license, which if
we were using it would be a gray area at best. We weren't using it at all,
and AGG seems to have disabled building it by default, so just remove it
from the tree entirely.
Briefly discussed on IRC.
Changes:
Fixed window resizing bug that happens after packages start downloading
Application icon and left stripe are sized relative to font size
Added alert window dialog to display BPackageManager warnings
Added tooltips to the packages in the list view
Found item has data offset higher than block size. Move the stream to get the
data. If the data does not fit in the block, fetch it directly from the device.
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
Fixes#12788.
