Fixed HID device class driver still using PrevReportINBuffer for GetReport control requests even when it has been set to NULL by the user application (thanks to Axel Rohde).

Minor tweaks to the library documentation.

Spell check code and manual pages. Remove custom Doxygen CSS stylesheet, as the new 1.7 Doxygen's default stylesheet is much better.

Fix HID parser documentation for the default maximum report/path/etc. elements so that it matches the current code values.

Add new ReportType parameter to the HID class driver device callback and host report sending routines.

Renamed internal Host mode Class driver descriptor comparator callback routines so that they all start with a uniform DCOMP_{Class Abbreviation}_ prefix.

Add svn:eol-style property to source files, so that the line endings are correctly converted to the target system's native end of line style.

Device mode class driver callbacks are now fired before the control request status stage is sent to prevent the host from timing out if another request is immediately fired and the device has a lengthy callback routine.

Inlined currently unused MIDI ProcessControlRequest routine.

Added new EVENT_CDC_Device_BreakSent() event and CDC_Host_SendBreak() function to the Device and Host CDC Class drivers.

Ensure that the CDC class drivers return 0 if the number of unread bytes is queried and the bank has become empty.

Fixed Host mode HID class driver not sending the correct report type when HID_Host_SendReportByID() was called and the HID_HOST_BOOT_PROTOCOL_ONLY compile time option is set.

The Audio_Device_IsSampleReceived() and Audio_Device_IsReadyForNextSample() functions are now inline, to reduce overhead.

Use puts_P() and printf_P() instead of the normal variants where possible in the Host mode Class Driver demos.

Fixed HID Device Class Driver not allocating a temporary buffer when the host requests a report via the control endpoint and the user has set the PrevReportINBuffer driver configuration element to NULL (thanks to Lars Noschinski).

Clean up incomplete BluetoothHost debugging commands to use GCC extension to avoid NULL parameters when no formatting is required.

Rename FunctionAttributes.h to Attributes.h, as some attributes are applicable to variables also. Add new ATTR_NOINIT attribute for global variables.

Add the beginnings of a SDP implentation to the incomplete BluetoothHost demo.

Add const attribute to the Mass Storage Host driver functions where it was applicable, but missing.

Removed two-step endpoint/pipe bank clear and switch sequence for smaller, faster endpoint/pipe code.

Added spacing between multiple paragraphs in Doxygen \note sections.

Removed call to the clock prescaler reset function in the Teensy bootloader to save space - the user application is accessed via a watchdog reset anyway, so the prescale reset would only affect the speed of the bootloader itself.

Remove accidentally duplicated model-specific peripheral driver files.

All Class Driver Host mode demos now correctly set the board LEDs to READY once the enumeration process has completed.

Fixed Still Image Host class driver not resetting the transaction ID when a new session is opened, fixed driver not sending a valid session ID to the device.

Fixed compilation error in the AudioInput demos when MICROPHONE_BIASED_TO_HALF_RAIL is defined (thanks to C. Scott Ananian).

Minor documentation cleanups.

Make sure HID class driver uses properly cast pointers when writing to the report buffer.

Turn OFF double banking in the Webserver demo - leaving it on seems to destroy the system throughput.

Fix HID Host Class driver sending incorrect ReportType values to the device when issuing requests via the control pipe.

Add more include protection macros to give the user warnings when they try to manually include private driver header files, instead of the public driver headers.

Moved the USB_GetHIDReportItemInfo() calls into the main report item passing loop in the *HostWithParser demos - it is fast enough not to effect performance, and avoids duplicate code.

Make Webserver project report the LUFA version as part of the HTTP header.

Fix AVRISP-MKII clone project's TPI Chip Erase command processing - ensure erase location is the high byte in the given address space, check NVMBUSY for completion rather than the NVM Bus Enable bit.

Change If-Else chains over to switch statements in XPROGProtocol.c for clarity.

Re-added Pipe_IsEndpointBound() calls to the CDC and RNDIS host class drivers, not that the function has the correct behaviour for devices with bidirectional endpoints.

Be doubly-certain that the incomming CDC class driver's endpoint/pipe is flushed only once when the bank is empty.