/*VERSION 0.3.2*/ *Added support for displays with 16bpp color depth. *A bug has been fixed, that caused the Xfixes cursor to appear distorted on amd64. *A lot of code refactoring and cleanups of duplicated parts. /*VERSION 0.3.1*/ *Added Added custom X error handler, to avoid BadWindow on XGetWindowAttributes, with XQueryTree. *Fixed sound_buffer initialization, which could cause segfault in startup. *Corrected buffer size in sound recoding. (Added new option to specify buffer size). *Corrected interleaving of audio and video streams. *Fixed missing end-of-stream in generated files. *Applied patch submitted by Andreas Zwinkau, correcting program crash/segfault, when run under window managers that don't conform to the FreeDesktop specification. /*VERSION 0.3.0r2*/ *Minor corrections on the inclusion of the include/ directory in the project. /*VERSION 0.3.0r1*/ *Fixed paused function, which was broken in 0.3.0 /*VERSION 0.3.0*/ *Implemented cache functionality (Encoding by default happens right after the recording) *Better damage event handling. *Saving under the same file-name,no longer deletes the previous one, instead it saves under a number postfixed file. *Opening the sound device doesn't block anymore. *A few new options were added related to caching. *Program now checks if it is run uder a known 3d compositing window manager and set behavior accordingly, to reduce the need for extra options at launch. *documented all fuctions in header. *Removed unneeded casts that bloated the code. *Updated man page. *Corrected X11 library checking during configuration. *Removed uneeded bitmap screenshot function. /*VERSION 0.2.6*/ *Behavior on stream desync has been changed. Now, instead of dropping sound buffers, image capturing speeds up. That is, waiting for timer to expire is disabled until sync is achieved. When not recording sound, this does not apply. *Theora return values are now properly checked, to avoid inserting corrupt packages into the stream. *Major memory leaks have been fixed. These were related to the rectangle insertion algorithm and discarding of used sound buffers, *Segmentation fault on seemingly random occasions has been found and is now fixed. Reason was lack of thread safety over libogg calls. /*VERSION 0.2.5*/ *Drawing of the correct cursor is now supported through the Xfixes extension. *Change of behavior on lack of any Xserver extension and failure to open sound device. The program now exits and informs instead of continuing with a different configuration. *Exit status also indicates the nature of any error that occured. *Documentation has been updated with exit status explanations. *Subsampling of the chroma planes is made by averaging, while pixel discarding is an option(used to be otherwise). /*VERSION 0.2.4*/ *Behavior on lack of $DISPLAY evironment variable has been corrected. *Size adjustment is now quantized at 16 pixels, in order to avoid complete image distortion, present at certain resolutions, when using shared memory (conditional or not). *Some other minor bugfixes, related mostly to default values of options. /*VERSION 0.2.3*/ *Problems on big-endian architectures, corrected (thanks to Marcel Unbehaun for bringing up the problem and providing the neccessary feedback to resolve it). *Failure to respect change in channels number, while opening soundcard, which resulted in crashes, segfaults and repeated error-messages has been fixed. *other minor bugfixes /*VERSION 0.2.2*/ *Changes to the configure script. *BGR to YUV convertion corrected. /*VERSION 0.2.1*/ *Unupdated lines bug fixed. *Fuzzy cursor at border bug fixed. *Incorrect display of windows that extended beyond recording height fixed. *Size adjustement that resulted in extra or missing pixels at borders has been corrected. New adjustment is 8 pixels max,split evenly if possible. *The option to drop frames has been added. *The option to do better subsampling has been added. *Default behavior now is to use shared memory when size of image gets over 75% of total (or a user set threshold). *Many other minor bugfixes.